An Introduction To Environmental Chemistry And Pollution

User Manual: An introduction to environmental chemistry and pollution

Open the PDF directly: View PDF .
Page Count: 485

Download
Open PDF In Browser	View PDF

Understanding Our Environment
An Introduction to Environmental
Chemistry and Pollution
Third Edition

Edited by
Roy M. Harrison
The University of Birmingham, UK

R S « C
ROYAL SOCE
ITY OF CHEMS
ITRY

ISBN 0-85404-584-8
A catalogue record for this book is available from the British Library.
© The Royal Society of Chemistry 1999
All rights reserved.
Apart from any fair dealing for the purposes of research or private study, or criticism or review as
permitted under the terms of the UK Copyright, Designs and Patents Act, 1988, this publication may
not be reproduced, stored or transmitted, in any form or by any means, without the prior permission in
writing of The Royal Society of Chemistry, or in the case of reprographic reproduction only in
accordance with the terms of the licences issued by the Copyright Licensing Agency in the UK, or in
accordance with the terms of the licences issued by the appropriate Reproduction Rights Organization
outside the UK. Enquiries concerning reproduction outside the terms stated here should be sent to The
Royal Society of Chemistry at the address printed on this page.
Published by The Royal Society of Chemistry,
Thomas Graham House, Science Park, Milton Road, Cambridge CB4 OWF, UK
For further information see our web site at www.rsc.org
Typeset by Paston PrePress Ltd, Beccles, Suffolk
Printed by Redwood Books Ltd, Trowbridge, Wiltshire

Preface
The field of environmental chemistry goes from strength to strength.
Twenty-five years ago it existed in the UK in the form of a few isolated
research groups in Universities, Polytechnics, and Research Institutes,
but was very definitely a minority interest. It was not taught appreciably
in academic institutions and few books dealt with any aspect of the
subject. The awakening of environmental awareness, first in a few
specialists and subsequently in the general public has led to massive
changes. Environmental chemistry is now a component (optional or
otherwise) of many chemistry degree courses, it is taught in environmental science courses as an element of increasing substance, and there
are even a few degree courses in the subject. Research opportunities in
environmental chemistry are a growth area as new programmes open up
to tackle local, national, regional, or global problems of environmental
chemistry at both fundamental and applied levels. Industry is facing ever
tougher regulations regarding the safety and environmental acceptability
of its products.
When invited to edit the second edition of 'Understanding Our
Environment', I was delighted to take on the task. The first edition had
sold well, but had never really met its original very difficult objective of
providing an introduction to environmental science for the layman. It
has, however, found widespread use as a textbook for both undergraduate and postgraduate-level courses and deserved further development with this in mind. I therefore endeavoured to produce a book
giving a rounded introduction to environmental chemistry and pollution,
accessible to any reader with some background in the chemical sciences.
Most of the book was at a level comprehensible by others such as
biologists and physicians who have a modest acquaintance with basic
chemistry and physics. The book was intended for those requiring a
grounding in the basic concepts of environmental chemistry and pollution. The third edition follows very much the same ethos as the second,
but I have tried to encourage chapter authors to develop a more

international approach through the use of case studies, and to make the
book more easily useable for teaching in a wide range of contexts by the
incorporation of worked examples where appropriate and of student
questions. The book is a companion volume to 'Pollution: Causes,
Effects and Control' (also published by the Royal Society of Chemistry)
which is both more diverse in the subjects covered, and in some aspects
appreciably more advanced.
Mindful of the quality and success of the second edition, it is fortunate
that many of the original authors have contributed revised chapters to
this book (A. G. Clarke, R. M. Harrison, B. J. Alloway, S. J. de Mora,
C. N. Hewitt, R. Allott, and S. Smith). I am pleased also to welcome new
authors who have produced a new view on topics covered in the earlier
book (A. S. Tomlin, J. G. Farmer, M. C. Graham, and A. Skinner). The
coverage is broadly the same, with some changes in emphasis and much
updating. The authors have been chosen for their deep knowledge of the
subject and ability to write at the level of a teaching text, and I must
express my gratitude to all of them for their hard work and willingness to
tolerate my editorial quibbles. The outcome of their work, I believe, is a
book of great value as an introductory text which will prove of widespread appeal.
Roy M. Harrison
Birmingham

Contributors
R. Allott, AEA Technology, Risley, Warrington, WA3 6AT, UK
B. J. Alloway, Department of Soil Science, University of Reading, Whiteknights, Reading, RG6 6DW, UK
A. G. Clarke, Department of Fuel and Energy, Leeds University, Leeds,
LS2 9JT, UK
S. J. de Mora, Departement d'Oceanographie, Universite du Quebec a
Rimouski, 300, allee des Ursulines, Rimouski, Quebec, G5L 3Al,
Canada
J. G. Farmer, Environmental Chemistry Unit, Department of Chemistry,
The University of Edinburgh, King's Buildings, West Main Road,
Edinburgh, EH9 3JJ, UK
M. C. Graham, Environmental Chemistry Unit, Department of Chemistry,
The University of Edinburgh, King's Buildings, West Main Road,
Edinburgh, EH9 3JJ, UK
R. M. Harrison, Institute of Public and Environmental Health, The
University of Birmingham, Edgbaston, Birmingham, B15 2TT, UK
C. N. Hewitt, Institute of Environmental and Natural Sciences, Lancaster
University, Lancaster, LAl 4YQ, UK
A. Skinner, Environment Agency, Olton Court, 10 Warwick Road,
Solihull, B92 7HX, UK
S. Smith, Division of Biosphere Sciences, King's College, University of
London, Campden Hill Road, London, W8 7AH, UK
A. S. Tomlin, Department of Fuel and Energy, Leeds University, Leeds,
LS2 9JT, UK

Contents

Preface ................................................................................

Contributors .........................................................................

xvi

1. Introduction ..................................................................

The Environmental Sciences ...........................................

The Chemicals of Interest ................................................

Units of Concentration .....................................................

The Environment as a Whole ..........................................

Bibliography .....................................................................

2. The Atmosphere ...........................................................

The Global Atmosphere ...................................................

1.1

The Structure of the Atmosphere ..........................
1.1.1 Troposphere and Stratosphere ..................
1.1.2 Atmospheric Circulation .............................
1.1.3 The Boundary Layer ..................................

9
9
10
11

1.2

Greenhouse Gases and the Global Climate ..........
1.2.1 The Global Energy Balance .......................
1.2.2 The Carbon Dioxide Cycle .........................
1.2.3 Global Warming .........................................
1.2.4 Climate Change .........................................
1.2.5 International Response ..............................

12
12
14
14
17
18

1.3

Depletion of Stratospheric Ozone .........................
1.3.1 The Ozone Layer .......................................
1.3.2 Ozone Depletion ........................................
1.3.3 The Antarctic Ozone ‘Hole’ ........................
1.3.4 Effects of International Control
Measures ...................................................

19
19
20
21

This page has been reformatted by Knovel to provide easier navigation.

vii

viii

Contents
2.

Atmospheric Transport and Dispersion of Pollutants ......

2.1

Wind Speed and Direction .....................................

2.2

Atmospheric Stability .............................................
2.2.1 The Lapse Rate .........................................
2.2.2 Temperature Inversions .............................

28
28
30

2.3

Dispersion from Chimneys ....................................
2.3.1 Ground-level Concentrations .....................
2.3.2 Plume Rise ................................................
2.3.3 Time Dependence of Average
Concentrations ..........................................

31
31
32

Mathematical Modeling of Dispersion ...................

Emissions to Atmosphere and Air Quality .......................

3.1

Natural Emissions .................................................
3.1.1 Introduction ................................................
3.1.2 Sulfur Species ...........................................
3.1.3 Nitrogen Species .......................................
3.1.4 Hydrocarbons ............................................

35
35
36
37
38

3.2

Emissions of Primary Pollutants ............................
3.2.1 Carbon Monoxide and Hydrocarbons ........
3.2.2 Nitrogen Oxides .........................................
3.2.3 Sulfur Dioxide ............................................
3.2.4 Particulate Matter ......................................
3.2.5 Emissions Limits ........................................
3.2.6 Emissions Inventories ................................

38
38
40
41
41
43
43

3.3

Air Quality ..............................................................
3.3.1 Air Quality Standards .................................
3.3.2 Air Quality Monitoring ................................
3.3.3 Air Quality Trends ......................................
3.3.4 Vehicular Emissions – CO and
Hydrocarbons ............................................
3.3.5 Nitrogen Oxides .........................................
3.3.6 Sulfur Oxides .............................................
3.3.7 Vehicular Particulates ................................

44
44
44
47

2.4
3.

This page has been reformatted by Knovel to provide easier navigation.

47
48
50
51

Contents
3.3.8
3.3.9

Heavy Metals .............................................
Toxic Organic Micropollutants
(TOMPS) ...................................................

Gas Phase Reactions and Photochemical Ozone ...........

4.1

Gas Phase Chemistry in the Troposphere ............
4.1.1 Atmospheric Photochemistry and
Oxidation ...................................................
4.1.2 Ozone ........................................................

Trends in Ozone Levels ........................................

Particles and Acid Deposition ..........................................

5.1

Particle Formation and Properties .........................
5.1.1 Particle Formation .....................................
5.1.2 Particle Composition ..................................
5.1.3 Deliquescent Behaviour .............................
5.1.4 Optical Properties ......................................

59
59
60
60
61

5.2

Droplets and Aqueous Phase Chemistry ..............

5.3

Deposition Mechanisms ........................................
5.3.1 Dry Deposition of Gases ............................
5.3.2 Wet Deposition ..........................................
5.3.3 Deposition of Particles ...............................

63
63
64
65

5.4

Acid Rain ...............................................................
5.4.1 Rainwater Composition .............................
5.4.2 The Effects ................................................
5.4.3 Patterns of Deposition and Critical
Loads Assessment ....................................

66
66
67

Questions ................................................................................

4.2
5.

53
56

Color Plates ............................................................................ 70a

3. Freshwaters ..................................................................

Introduction ......................................................................

Fundamentals of Aquatic Chemistry ................................

2.1

74
74

Introduction ...........................................................
2.1.1 Concentration and Activity .........................

This page has been reformatted by Knovel to provide easier navigation.

Contents
2.1.2
2.1.3

Ionic Strength ............................................
Equilibria and Equilibrium Constants .........

75
77

Dissolution/Precipitation Reactions .......................
2.2.1 Physical and Chemical Weathering
Processes ..................................................
2.2.2 Solubility ....................................................
2.2.3 Influence of Organic Matter .......................

2.3

Complexation Reactions in Freshwaters ...............
2.3.1 Outer and Inner Sphere Complexes ..........
2.3.2 Hydrolysis ..................................................
2.3.3 Inorganic Complexes .................................
2.3.4 Surface Complex Formation ......................
2.3.5 Organic Complexes ...................................

82
82
82
83
84
84

2.4

Species Distribution in Freshwaters ...................... 85
2.4.1 pH as a Master Variable ............................ 85
2.4.2 pε as a Master Variable ............................. 97
2.4.3 pε – pH Relationships ................................ 100

2.5

Modeling Aquatic Systems .................................... 106

2.2

79
80
81

Case Studies ................................................................... 106
3.1

3.2

Acidification ...........................................................
3.1.1 Diatom Records .........................................
3.1.2 Aluminium ..................................................
3.1.3 Acid Mine Drainage and Ochreous
Deposits .....................................................
3.1.4 Acid Mine Drainage and the Release of
Heavy Metals .............................................

106
106
107

Metals in Water .....................................................
3.2.1 Arsenic in Groundwater .............................
3.2.2 Lead in Drinking Water ..............................
3.2.3 Cadmium in Irrigation Water ......................
3.2.4 Selenium in Irrigation Water ......................
3.2.5 Aquatic Contamination by Gold Ore
Extractants .................................................

112
112
113
114
115

This page has been reformatted by Knovel to provide easier navigation.

108
109

117

Contents
3.3

Historical Pollution Records and Perturbatory
Processes in Lakes ...............................................
3.3.1 Records – Lead in Lake Sediments ...........
3.3.2 Perturbatory Processes in Lake
Sediments ..................................................
3.3.3 Onondaga Lake .........................................

xi
119
119
119
123

3.4

Nutrients in Water and Sediments ......................... 125
3.4.1 Phosphorus and Eutrophication ................ 125
3.4.2 Nitrate in Groundwater .............................. 129

3.5

Organic Matter and Organic Chemicals in
Water ..................................................................... 130
3.5.1 BOD and COD ........................................... 130
3.5.2 Synthetic Organic Chemicals .................... 131

Treatment ........................................................................ 134
4.1

Purification of Water Supplies ............................... 134

4.2

Waste Treatment ................................................... 135

Questions ................................................................................ 136
Further Reading ...................................................................... 138

4. The Oceanic Environment ........................................... 139
1.

Introduction ...................................................................... 139
1.1

The Ocean as a Biogeochemical Environment ..... 139

1.2

Properties of Water and Seawater ........................ 142

1.3

Salinity Concepts ................................................... 146

1.4

Oceanic Circulation ............................................... 148

Seawater Composition and Chemistry ............................ 150
2.1

Major Constituents ................................................ 150

2.2

Dissolved Gases ...................................................
2.2.1 Gas Solubility and Air-sea Exchange
Processes ..................................................
2.2.2 Oxygen ......................................................
2.2.3 Carbon Dioxide and Alkalinity ....................

This page has been reformatted by Knovel to provide easier navigation.

153
153
155
158

xii

Contents
2.2.4

Dimethyl Sulfide and Climatic
Implications ................................................ 165

2.3

Nutrients ................................................................ 167

2.4

Trace Elements ..................................................... 169

2.5

Physico-chemical Speciation ................................. 171

Suspended Particles and Marine Sediments ................... 177
3.1

Description of Sediments and Sedimentary
Components .......................................................... 177

3.2

Surface Chemistry of Particles ..............................
3.2.1 Surface Charge .........................................
3.2.2 Adsorption Processes ................................
3.2.3 Ion Exchange Reactions ............................
3.2.4 Role of Surface Chemistry in
Biogeochemical Cycling ............................

3.3

181
181
182
183
184

Diagenesis ............................................................ 185

Physical and Chemical Processes in Estuaries ............... 186

Marine Contamination and Pollution ................................ 190
5.1

Oil Slicks ............................................................... 191

5.2

Plastic Debris ........................................................ 193

5.3

Tributyltin ............................................................... 194

Questions ................................................................................ 197

5. Land Contamination and Reclamation ....................... 199
1.

Introduction ...................................................................... 199

Soil: Its Formation, Constituents, and Properties ............ 201
2.1

Soil Formation ....................................................... 202

2.2

Soil Constituents ................................................... 203
2.2.1 The Mineral Fraction .................................. 204
2.2.2 Soil Organic Matter .................................... 205

2.3

Soil Properties ....................................................... 206
2.3.1 Soil Permeability ........................................ 206
2.3.2 Soil Chemical Properties ........................... 207

This page has been reformatted by Knovel to provide easier navigation.

Contents
2.3.3

xiii

Adsorption and Decomposition of
Organic Contaminants ............................... 210

Sources of Land Contaminants ....................................... 212

Characteristics of Some Major Groups of Land
Contaminants .................................................................. 214
4.1

Heavy Metals ........................................................ 214

4.2

Organic Contaminants ........................................... 215

4.3

Sewage Sludge ..................................................... 218

Possible Hazards from Contaminated Land .................... 219

Methods of Site Investigation .......................................... 220

Interpretation of Site Investigation Data .......................... 223

Reclamation of Contaminated Land ................................ 226

8.1

Ex Situ Methods .................................................... 226
8.1.1 ‘Dig and Dump’ .......................................... 226
8.1.2 Soil Cleaning ............................................. 226

8.2

In Situ Methods ..................................................... 227
8.2.1 Physico-chemical Methods ........................ 227
8.2.2 Biological Methods .................................... 229

8.3

Specific Techniques for Gasworks Sites ............... 230

Case Studies ................................................................... 230
9.1

Gasworks Sites ..................................................... 230

9.2

Soil Contamination by Landfilling and Waste
Disposal ................................................................ 232

9.3

Heavy Metal Contamination from Metalliferous
Mining and Smelting .............................................. 233

9.4

Heavy Metal Contamination of Domestic Garden
Soils in Urban Areas .............................................. 234

9.5

Land Contamination by Solvents, PCBs, and
Dioxins Following a Fire at an Industrial Plant ...... 235

10. Conclusions ..................................................................... 235
Questions ................................................................................ 236

This page has been reformatted by Knovel to provide easier navigation.

xiv

Contents

6. Environmental Cycling of Pollutants .......................... 237
1.

Introduction: Biogeochemical Cycling .............................. 237
1.1

Environmental Reservoirs ..................................... 239

1.2

Lifetimes ................................................................ 240
1.2.1 Influence of Lifetime on Environmental
Behaviour .................................................. 243

Rates of Transfer between Environmental
Compartments ................................................................. 244
244

2.1

Air-land Exchange ................................................

2.2

Air-sea Exchange ................................................. 247

Transfers in Aquatic Systems .......................................... 254

Biogeochemical Cycles ................................................... 257

4.1

Case Study 1: the Biogeochemical Cycle of
Nitrogen ................................................................. 259

4.2

Case Study 2: Aspects of the Biogeochemical
Cycle of Lead ........................................................ 260

Environmental Partitioning of Long-lived Species ........... 264

Questions ................................................................................ 265

7. Environmental Monitoring Strategies ........................ 267
1.

Objectives of Monitoring .................................................. 267

Types of Monitoring ......................................................... 269
2.1

2.2

Source Monitoring .................................................
2.1.1 General Objectives ....................................
2.1.2 Stationary Source Sampling for Gaseous
Emissions ..................................................
2.1.3 Mobile Source Sampling for Gaseous
Effluents .....................................................
2.1.4 Source Monitoring for Liquid Effluents .......
2.1.5 Source Monitoring for Solid Effluents ........

271
271
271
271
272
272

Ambient Environment Monitoring .......................... 274
2.2.1 General Objectives .................................... 274
2.2.2 Ambient Air Monitoring .............................. 274

This page has been reformatted by Knovel to provide easier navigation.

Contents
2.2.3
2.2.4
3.

Environmental Water Monitoring ............... 279
Sediment, Soil, and Biological
Monitoring .................................................. 285

Sampling Methods ........................................................... 291
3.1

Air Sampling Methods ...........................................
3.1.1 Intake Design .............................................
3.1.2 Sample Collection ......................................
3.1.3 Flow Measurement and Air Moving
Devices ......................................................

291
291
293
300

3.2

Water Sampling Methods ...................................... 300

3.3

Soil and Sediment Sampling Methods .................. 302

Modeling of Environmental Dispersion ............................ 303
4.1

Atmospheric Dispersal .......................................... 305

4.2

Aquatic Mixing ....................................................... 309

4.3

Variability in Soil and Sediment Pollutant
Levels .................................................................... 311

Duration and Extent of Survey ......................................... 311
5.1

Duration of Survey and Frequency of
Sampling ............................................................... 311

5.2

Methods of Reducing Sampling Frequency .......... 315

5.3

Number of Sampling Sites ..................................... 316

Prerequisites for Monitoring ............................................. 316
6.1

Monitoring Protocol ............................................... 317

6.2

Meteorological Data .............................................. 318

6.3

Source Inventory ................................................... 319

6.4

Suitability of Analytical Techniques ....................... 320

6.5

Environmental Quality Standards .......................... 322

Remote Sensing of Pollutant ........................................... 324

Presentation of Data ........................................................ 326

Questions ................................................................................ 328

This page has been reformatted by Knovel to provide easier navigation.

xvi

Contents

8. Ecological and Health Effects of Chemical
Pollution ........................................................................ 331
1.

Introduction ...................................................................... 331

Toxicity: Exposure-response Relationships ...................

Exposure ......................................................................... 336

Absorption ....................................................................... 339

Internal Pathways ............................................................ 342

Ecological Risk Assessment ............................................ 347

Individuals, Populations, and Communities and the
Role of Biomarkers .......................................................... 349

Health Effects of the Major Air Pollutants ........................ 358

Effect of Air Pollution on Plants ....................................... 362

333

10. Ecological Effects of Acid Deposition .............................. 366
11. Forest Decline ................................................................. 373
12. Effects of Pollutants on Reproduction and Development:
Evidence of Endocrine Disruption ................................... 374
12.1 Eggshell Thinning .................................................. 375
12.2 GLEMEDS ............................................................. 377
12.3 Marine Mammals ................................................... 378
12.4 Imposex in Gastropods ......................................... 379
12.5 Endocrine Disruptors ............................................. 380
13. Hydrocarbons in the Marine Environment ....................... 383
14. Health Effects of Metal Pollution ...................................... 388
14.1 Mercury ................................................................. 388
14.2 Lead ...................................................................... 391
15. Conclusion ....................................................................... 394
Questions ................................................................................ 395

9. Managing Environmental Quality ............................... 397
1.

Introduction ...................................................................... 397

Objectives, Standards, and Limits ................................... 400
2.1

Environmental Objectives ...................................... 400

This page has been reformatted by Knovel to provide easier navigation.

Contents

xvii

2.2

Environmental Standards ...................................... 401

2.3

Emission Limits ..................................................... 402

2.4

Integrating Limit Values and Quality Standards ....
2.4.1 Use-related Approach ................................
2.4.2 Uniform Emission Standards .....................
2.4.3 Sectoral Approach .....................................

2.5

Specifying Standards ............................................ 405

2.6

Remediation Targets ............................................. 411

2.7

The Principles of No Deterioration and
Precaution ............................................................. 413

404
404
405
405

Legislation to Control and Prevent Pollution .................... 413
3.1

Origins of Pollution Control Legislation ................. 414

3.2

Trends in European Environmental Legislation ..... 415

3.3

Reporting Environmental Performance ................. 417

3.4

Pollution Control and Land Use Planning .............. 418

Pollution Control Agencies .............................................. 420
4.1

Structure and Organization of Pollution Control
Agencies ............................................................... 420

4.2

Forestalling Pollution ............................................. 422

4.3

Other Regulatory Action ........................................ 426

Economic Instruments for Managing Pollution ................ 427
5.1

Alternatives to Pollution Regulation by Permit ...... 427

5.2

Tradeable Permits ................................................. 430

Public and Commercial Pressures to Improve the
Environment .................................................................... 432
6.1

Environmental Management Systems ................... 432

6.2

Public Opinion and the Environment ..................... 434

Questions ................................................................................ 435

Index ................................................................................... 437

This page has been reformatted by Knovel to provide easier navigation.

CHAPTER 1

Introduction
ROY M. HARRISON

1 THE ENVIRONMENTAL SCIENCES
It may surprise the student of today to learn that 'the environment' has
not always been topical and indeed that environmental issues have
become a matter of widespread public concern only over the past
twenty years or so. Nonetheless, basic environmental science has existed
as a facet of human scientific endeavour since the earliest days of
scientific investigation. In the physical sciences, disciplines such as
geology, geophysics, meteorology, oceanography, and hydrology, and
in the life sciences, ecology, have a long and proud scientific tradition.
These fundamental environmental sciences underpin our understanding
of the natural world, and its current-day counterpart perturbed by
human activity, in which we all live.
The environmental physical sciences have traditionally been concerned with individual environmental compartments. Thus, geology is
centred primarily on the solid earth, meteorology on the atmosphere,
oceanography upon the salt water basins, and hydrology upon the
behaviour of fresh waters. In general (but not exclusively) it has been the
physical behaviour of these media which has been traditionally perceived
as important. Accordingly, dynamic meteorology is concerned primarily
with the physical processes responsible for atmospheric motion, and
climatology with temporal and spatial patterns in physical properties of
the atmosphere (temperature, rainfall, etc.). It is only more recently that
chemical behaviour has been perceived as being important in many of
these areas. Thus, while atmospheric chemical processes are at least as
important as physical processes in many environmental problems such as
stratospheric ozone depletion, the lack of chemical knowledge has been
extremely acute as atmospheric chemistry (beyond major component
ratios) only became a matter of serious scientific study in the 1950s.

There are two major reasons why environmental chemistry has
flourished as a discipline only rather recently. Firstly, it was not
previously perceived as important. If environmental chemical composition is relatively invariant in time, as it was believed to be, there is little
obvious relevance to continuing research. Once, however, it is perceived
that composition is changing (e.g. CO2 in the atmosphere; 137Cs in the
Irish Sea) and that such changes may have consequences for humankind,
the relevance becomes obvious. The idea that using an aerosol spray in
your home might damage the stratosphere, although obvious to us
today, would stretch the credulity of someone unaccustomed to the
concept. Secondly, the rate of advance has in many instances been
limited by the available technology. Thus, for example, it was only in
the 1960s that sensitive reliable instrumentation became widely available
for measurement of trace concentrations of metals in the environment.
This led to a massive expansion in research in this field and a substantial
downward revision of agreed typical concentration levels due to
improved methodology in analysis. It was only as a result of James
Lovelock's invention of the electron capture detector that CFCs were
recognized as minor atmospheric constituents and it became possible to
monitor increases in their concentrations (see Table 1). The table
exemplifies the sensitivity of analysis required since concentrations are
at the ppt level (1 ppt is one part in 1012 by volume in the atmosphere) as
well as the substantial increasing trends in atmospheric halocarbon
concentrations, as measured up to 1990. The implementation of the
Montreal Protocol, which requires controls on production of CFCs and
Table 1 Atmospheric halocarbon concentrations and trends*
Concentration
(PPt)

Annual Change
(PPt)

Halocarbon

Pre-industrial

1992

To 1990

Since 1992

CCl 3 F (CFC-Il)
CCl 2 F 2 (CFC-12)
CClF 3 (CFC-13)
C 2 Cl 3 F 3 (CFC-113)
C 2 Cl 2 F 4 (CFC-IM)
C 2 ClF 5 (CFC-115)
CCl 4
CH 3 CCl 3

0
0
0
0
0
0
0
0

268
503
<10
82
20
<10
132
135

+ 9.5
+ 16.5

0
+7

+ 4-5

+ 2.0
+ 6.0

-0.5
-10

Lifetime
(years)
50
102
400
85
300
1700
42
4.9

*Data from: Intergovernmental Panel on Climate Change, 'Climate Change—The IPCC Scientific
Assessment', ed. J. T. Houghton, G. J. Jenkins, and J. J. Ephraums, Cambridge University Press,
Cambridge, 1990; and 'Climate Change 1995, The Science of Climate Change', ed. J. T. Houghton,
L. G. Meira Filho, B. A. Callendar, N. Harris, A. Kattenberg, and K. Maskell, Cambridge
University Press, Cambridge, 1996.

some other halocarbons, has led to a slowing and even a reversal of
annual concentration trends since 1992 (see Table 1).
2

THE CHEMICALS OF INTEREST

A very wide range of chemical substances are considered in this book.
They fall into three main categories:
(a) Chemicals of concern because of their human toxicity. Some metals
such as lead, cadmium and mercury are well known for their
adverse effects on human health at high levels of exposure. These
metals have no known essential role in the human body and
therefore exposures can be divided into two categories (see Figure
1). For these non-essential elements, at very low exposures the
metals are tolerated with little, if any, adverse effect, but at higher
exposures their toxicity is exerted and health consequences are
seen. In the case of the so-called essential trace elements (see also
Figure 1) the human body requires a certain level of the element,
and if intakes are too low then deficiency syndrome diseases will
result. These can have consequences as severe as the ones which
result from excessive intakes. In between, there is an acceptable
range of exposures within which the body is able to regulate an
optimum level of the element.
Environmental exposure to chemical carcinogens is very topical
despite the minuscule risks associated with many such exposures
at typical environmental concentrations. Examples are benzene
(largely from vehicle emissions) and polynuclear aromatic hydrocarbons (generated by combustion of fossil fuels). Figure 2 shows
the structures of benzene, benzo(a)pyrene (the best known of the
carcinogenic polycyclic aromatic hydrocarbons), and 2,3,7,8tetrachlorodibenzodioxin (the most toxic of the chlorinated
dioxin group of compounds). Despite great public concern over
emissions of the last compound, the evidence for carcinogenicity
in humans is quite limited.
(b) Chemicals which cause damage to non-human biota but are not
believed to harm humans at current levels of exposure. Many
elements and compounds come into this category. For example,
copper and zinc are essential trace elements for humans and
environmental exposures very rarely present a risk to health.
These elements are, however, toxic to growing plants and there
are regulations limiting their addition to soil in materials such as
sewage sludge which is disposed of to the land. Another category
of substance for which there is ample evidence of harm to biota,

ESSENTIAL
TRACE
ELEMENT

Beneficial

Uptake
Deleterious

INESSENTIAL
TRACE
ELEMENT

Beneficial

Figure 1 Comparison of the consequences of exposure to essential and inessential trace
elements. For the essential trace elements, Region A represents the deficiency
syndrome when intakes are insufficient, Area B is the optimum exposure window
and in Area C, excessive intake leads to toxic consequences. In the case of the
inessential trace element at low exposures (Zone E) the element is tolerated and
little if any adverse effect occurs. In Zone F toxic symptoms are developed

(a)
Figure 2

(b)

(C)

Some molecules believed to have human carcinogenic potential: (a) benzene;
(b) benzofa.Jpyrene; (c) 2,3,7,8-tetrachlorodibenzodioxin

but as yet little, if any, hard evidence of impacts on human
populations, are the endocrine-disrupting chemicals (also termed
oestrogenics). These synthetic chemicals mimic natural hormones
and can disrupt the reproduction and growth of wildlife species.
Thus, for example, bis-tributyl tin oxide (TBTO) interferes with
the sexual development of oysters and its use as an anti-fouling
paint for inshore vessels is now banned in most parts of the world.
A wide range of other chemicals including polychlorinated biphenyls (PCBs), dioxins, and many chlorinated pesticides are also
believed to have oestrogenic potential, although the level of
evidence for adverse effects is variable.
(c) Chemicals not directly toxic to humans or other biota at current
environmental concentrations, but capable of causing environmenta
damage. The prime example is the CFCs which found widespread
use precisely because of their stability and low toxicity to humans,
but which at parts per trillion levels of concentration are capable
of causing major disruption to the chemistry of the stratosphere.
3 UNITS OF CONCENTRATION
The concentration units used in environmental chemistry are often
confusing to the newcomer. Concentrations of pollutants in soils are
most usually expressed in mass per unit mass, for example, milligrams of
lead per kilogram of soil. Similarly, concentrations in vegetation are also
expressed in mg kg~ l or fig kg~ l . In the case of vegetation and soils, it is
important to distinguish between wet weight and dry weight concentrations, in other words, whether the kilogram of vegetation or soil is
determined before or after drying. Since the moisture content of vegetation can easily exceed 50%, the data can be very sensitive to this
correction.
In aquatic systems, concentrations can also be expressed as mass per
unit mass and in the oceans some trace constituents are present at
concentrations of ng kg" 1 or fig kg""1. More often, however, sample
sizes are measured by volume and concentrations expressed as ng 1~1 or
/ng 1"l. In the case of freshwaters, especially, concentrations expressed as
mass per litre will be almost identical to those expressed as mass per
kilogram. As a kind of shorthand, however, water chemists sometimes
refer to concentrations as if they were ratios by weight, thus, mg I" 1 are
expressed as parts per million (ppm), jag 1~l as parts per billion (ppb) and
ng I""1 as parts per trillion (ppt). This is unfortunate as it leads to
confusion with the same units used in atmospheric chemistry with a quite
different meaning.
Concentrations of trace gases and particles in the atmosphere can be

expressed also as mass per unit volume, typically /ig m 3. The difficulty
with this unit is that it is not independent of temperature and pressure.
Thus, as an airmass becomes warmer or colder or changes in pressure so
its volume will change, but the mass of the trace gas will not. Therefore,
air containing 1 fig m~ 3 of sulfur dioxide in air at 0 0C will contain less
than 1 fig m~ 3 of sulfur dioxide in air if heated to 25 0C. For gases (but
not particles) this difficulty is overcome by expressing the concentration
of a trace gas as a volume mixing ratio. Thus, 1 cm3 of pure sulfur
dioxide dispersed in 1 m3 of polluted air would be described as a
concentration of 1 part per million (ppm). Reference to the gas laws
tells us that not only is this one part per 106 by volume, it is also one
molecule in 106 molecules and one mole in 106 moles, as well as a partial
pressure of 10~6 atmospheres. Additionally, if the temperature and
pressure of the airmass change, this affects the trace gas in the same way
as the air in which it is contained and the volume mixing ratio does not
change. Thus, ozone in the stratosphere is present in the air at
considerably higher mixing ratios than in the lower atmosphere (troposphere), but if the concentrations are expressed in fig m~ 3 they are little
different because of the much lower density of air at stratospheric
attitudes. Chemical kineticists often express atmospheric concentrations
in molecules per cubic centimetre (molec cm" 3 ), which has the same
problem as the mass per unit volume units.
Worked Example

The concentration of nitrogen dioxide in polluted air is 85 ppb. Express
this concentration in units of fig m~~3 and molec cm""3 if the air
temperature is 200C and the pressure 1005 mb (1.005 x 105 Pa).
Relative molecular mass of NO2 is 46; Avogadro number is 6.022 x
1023.
The concentration OfNO2 is 85 fA m~3. At 20 0C and 1005 mb,
8 5 x

1 0

2 7 3

n
'UAz
~
85 ,1i M
NO
2 weigh46 x - ^ ^ x —
Q<

1 0 0 5

x ~

= 161 x 10~ 6 g
NO2 concentration = 1 6 1 ^ m " 3
This is equivalent to 161 pg cm" 3 , and
161 x 10~12
23
161 pg NO2 contain 6.022 x 10 x
46
12
= 2.1 x 10 molecules
and NO2 concentration = 2.1 x 1012 molec cm" 3 .

4 THE ENVIRONMENT AS A WHOLE
A facet of the chemically centred study of the environment is a greater
integration of the treatment of environmental media. Traditional
boundaries between atmosphere and waters, for example, are not a
deterrent to the transfer of chemicals (in either direction), and indeed
many important and interesting processes occur at these phase boundaries.
In this book, the treatment first follows traditional compartments
(Chapters 2, 3, 4, and 5) although some exchanges with other compartments are considered. Fundamental aspects of the science of the atmosphere, waters, and soils are described, together with current
environmental questions, exemplified by case studies. Subsequently,
quantitative aspects of transfer across phase boundaries are described
and examples given of biogeochemical cycles (Chapter 6). Monitoring
considerations are covered in Chapter 7, with the effects of chemical
pollution in Chapter 8, and finally the regulatory aspects in Chapter 9.
5 BIBLIOGRAPHY
For readers requiring knowledge of basic chemical principles:
R.M. Harrison and SJ. de Mora, 'Introductory Chemistry for the Environmental
Sciences', Cambridge University Press, Cambridge, 2nd Edn., 1996.
For more detailed information upon pollution phenomena:
'Pollution: Causes, Effects and Control', ed. R.M. Harrison, Royal Society of Chemistry,
Cambridge, 3rd Edn., 1996.

CHAPTER 2

The Atmosphere
A. G. CLARKE AND A. S. TOMLIN

1
1.1

THE GLOBAL ATMOSPHERE
The Structure of the Atmosphere

1.1.1 Troposphere and Stratosphere. The vertical structure of the
atmosphere, showing the features that are most relevant to the problems
covered in this chapter, is illustrated in Figure 1. The figure shows the
stratosphere, troposphere and boundary layer (that closest to the earth's
surface). The difference between the layers is characterized by changes in
temperature with height, and with changes in structure of the layers such
as cloud cover and turbulence. The depth of the troposphere is 8-15 km,
the lowest values occurring at the poles and the highest at the equator
with some seasonal variations. Within this layer occurs most of the
variability of conditions which leads to 'the weather' as the layman
experiences it. The stratosphere is relatively cloud-free and considerably
less turbulent—hence long distance passenger jets fly at stratospheric
altitudes. Within the troposphere temperature decreases with height
owing to the decreasing influence of radiation from the earth's surface,
but as we enter the stratosphere the temperature starts to increase again.
The turning point is called the tropopause. This situation of a layer of
warmer, less dense air over a layer of cooler, denser air is quite stable.
Consequently air is mixed across the tropopause very slowly unless
special events such as tropospheric folding occur.
We normally think of 'air pollution' in terms of the troposphere,
within which most pollutants have a fairly limited lifetime before they are
washed out by rain, removed by reaction, or deposited to the ground.
However, if pollutants are injected directly into the stratosphere they can
remain there for long periods because of slow downward mixing,
resulting in noticeable effects over the whole globe. Thus major volcanic

Stratopause
STRATOSPHERE
Pressure
mb
Intercontinental
airliner

Altitude

High level cloud
(Cirrus)

Tropopause

Storm
clouds
(Cumulonimbus)

TROPOSPHERE

Low level cloud
(Stratus)
Boundary layer
Temperature K
Figure 1

The vertical structure of the atmosphere. The temperature profile would be
typical for latitude 60° N in summer. Note the change of scale used for the upper
half of the figure

eruptions injecting fine dust into the stratosphere can lead to a reduction
in the amount of solar energy reaching the ground for more than a year
after the event. Other global problems relating to events in the stratosphere such as the possibility of damage to the ozone layer are discussed
later.
7.7.2 Atmospheric Circulation. To understand both global and local
environmental problems we must first understand how pollutants
circulate throughout the atmosphere. The main driving forces for the
circulation of the atmosphere are the incident solar radiation and the
earth's rotation. Because of the sun's angle, the amount of solar energy
falling on a given area varies with latitude so that the poles are cold and
the equatorial regions warm. Warm air rises at the equator and cold air
flows inwards from both North and South. A similar situation occurs at
the poles where warm air flows towards them and falls in the cold regions
there. The rotation of the earth affects the circulation patterns in a

fundamental way due to an effect called the Coriolis force. For example,
air moving south towards the equator gives the impression of being
influenced by a force in the westerly direction. The net result is the
tendency for air to circulate in large-scale eddies around the 'low' and
'high' pressure regions on synoptic weather charts.
The proportions of incident energy reflected back to space, absorbed
by the land or sea, and re-radiated at a longer wavelength all vary from
place to place and affect the temperature distribution and circulation
patterns. This energy balance is crucial to the determination of the global
climate and is considered in more detail in Section 1.2. The processes of
evaporation of water, cloud formation, and precipitation also affect the
energy balance and circulation patterns. The presence of the ground has
only a small effect on the overall pattern of atmospheric circulation and
at most altitudes air movements approximate to those of a non-viscous
fluid. The theoretical wind speed can be calculated from the pressure
gradient and the rotational velocity of the earth—the so-called geostrophic wind speed. The pressure gradient is reflected on a weather chart
by the closeness of the isobars, lines of constant pressure. If the isobars
are close together the wind speed will be high.
1.1.3 The Boundary Layer. Near to the ground the situation is more
complicated due to the effects of frictional and buoyancy forces.
Turbulence is generated by mechanical forces as air flows over uneven
ground features such as hills, buildings, or trees. The ground may also
warm or cool the air next to it resulting in upcurrents and downcurrents.
In the language of fluid mechanics we have turbulent transport of
momentum and energy with corresponding velocity and temperature
gradients in the vertical direction. Consider the variation of wind speed
with height over the lowest few hundred metres of the atmosphere. This
variation is greatest over rough surfaces {e.g. a city) where the effect
could be a reduction of 40% of the wind speed aloft, that is, the
geostrophic wind. Over smooth surfaces {e.g. sea, ice sheets) the effect is
less and the reduction may be only 20%. The changing effect of friction
with height also causes a variation of wind direction as we move away
from the earth's surface, i.e. 'wind shear'. A plume from a tall chimney
may therefore appear to be travelling at an angle to the ground level
wind.
Within the troposphere we therefore define a boundary layer within
which surface effects are important. This is of the order of 1 km in depth
but varies significantly with meteorological conditions (Figure 1). Vertical mixing of pollutants within the boundary layer is largely determined
by the atmospheric stability which relates to the intensity of the buoyancy effects previously mentioned. This is the subject of a later section.

Table 1

Time and distance scales for atmospheric dispersion of emissions

Time of travel

Typical distances

Area affected

Hours

Tens of km

Throughout the boundary layer

Days

Thousands of km

Pollutant escaping from boundary layer into free
troposphere

Weeks

Round the earth

The whole troposphere in one hemisphere
Transport to other hemisphere beginning

Months

Round the earth

Whole global troposphere. Some penetration into
lower stratosphere

As a generalization, mixing within the boundary layer is relatively rapid
whereas mixing through the remainder of the troposphere is slower. This
gives rise to the idea of a mixing depth within which pollutants are
retained and may be transported long distances. So, for example, models
of pollutant transport from the UK to the rest of Europe involve a
distance scale of about 1000 km and often assume vertical mixing depth
of perhaps 1 km with the pollutants uniformly distributed within this
layer. Table 1 indicates the time and distance scales involved in the
dispersion of pollutants emitted from the ground. No account is taken in
this table of the rates of removal of any pollutant by reaction, deposition
to the ground, etc.
1.2

Greenhouse Gases and the Global Climate 16

1.2.1 The Global Energy Balance. The energy that reaches the earth
comes from the sun, and the absorption and loss of radiation from the
earth and its atmosphere determine our climate. If the earth had no
atmosphere, the mean surface temperature would be 255 K, well below
the freezing point of water. The atmosphere serves to retain heat near the
surface and the earth is thereby made habitable. This accounting for
incoming and outgoing energy is called the global energy balance and
1

'The Greenhouse Effect, Climate Change and Ecosystems' (SCOPE 29), ed. B. Bolin, B. R. Doos, J.
Jager, and R. A. Warwick, John Wiley & Sons, Chichester, 1986.
'Climate Change 1995: The Science of Climate Change', ed. J. T. Houghton, L. G. Meira Filho,
B. A. Callander, N. Harris, A. Kattenberg, and K. Maskell, Cambridge University Press,
Cambridge, 1996.
3
'The Greenhouse Effect and Terrestrial Ecosystems of the UK', ed. M. G. R. Cannell and M. D.
Hooper, ITTE Research Publication No. 4, HMSO, London, 1990.
4
'Climate Change, The IPCC Scientific Assessment', ed. F. T. Houghton, G. J. Jenkins, and J. J.
Ephraums, Cambridge University Press, Cambridge, 1990.
5
http://www.ipcc.ch/ 'The Intergovernmental Panel on Climate Change (IPCC) home page',
December 1997.
6
http://www.bna.com/prodhome/ens/text4.htm, 'The Kyoto Protocol', December 1997.
2

Reflected Solar
Radiation2
107Wm"

Reflected by Clauds,
Aerosol and
Atmosphere
77

Incoming

Solar
Radiation
342 W m"2

Emitted by
Atmosphere
Absorbed by
Atmosphere

Oy^oiitg
Longwave
Radiation
235Wm 2

40
Atmospheric
Window
Greenhouse
Gases

Latent
78 Heat
Reflected by
Surface
30 '

Figure 2

Back
Radiation

The earth's radiation and energy balance for a net incoming solar radiation of
342 Wm'2
(Reproduced with permission from the Intergovernmental Panel on Climate
Change2)

could potentially be upset by any significant changes to the earth's
atmosphere.
Most of the radiant energy from the sun lies in or near the visible
region of the spectrum {i.e. at short wavelength ca. 0.6 /mi) with some in
the UV region. The stratosphere absorbs UV radiation primarily due to
the ozone present and this results in the warming shown in Figure 1. The
lower atmosphere is transparent to visible light so it gains relatively little
energy from incoming radiation. Some of the transmitted radiant energy
in the visible region penetrates to the ground and is absorbed. Some light
is reflected unchanged from clouds or from the ground (especially by
snow or ice). The fraction of reflected light is termed the albedo and is
over 0.5 for clouds but below 0.1 for the oceans. The global average
albedo is about 0.3. Figure 2 shows the amounts of radiation for different
components of the overall energy balance. The radiation emitted from
the ground lies in the infrared region of the spectrum (long wavelength,
ca. 10-15/im) and several atmospheric constituents absorb in this
region. Carbon dioxide, water vapour, and ozone are the most important
of these. Methane, nitrous oxide, and chlorofluorocarbons (CFCs) are
also significant. Some of the absorbed energy will still be re-radiated
back to space but a part will be returned to the ground or retained in the
atmosphere. The final factors that result in surface to atmosphere
transfer of energy are direct warming of the air nearest the ground

together with evaporation/condensation processes. The net effect is that
more energy is retained near the surface of the earth and the mean
temperature is therefore higher (global average 288 K). This is described
as the 'greenhouse effect' by analogy with the properties of glass. Glass is
largely transparent to solar radiation while absorbing completely radiation in the infrared at wavelengths greater than 3 /mi. In fact the most
important function of a greenhouse is to prevent the circulation of air,
inhibiting the normal cooling processes, but the term 'greenhouse effect'
has none the less been retained.
7.2.2 The Carbon Dioxide Cycle. Carbon dioxide is of major concern
as a greenhouse gas because there is no doubt that human activities are
leading to a gradual increase in the atmospheric carbon dioxide level.
This suggests that we may eventually modify the global climate. Fossil
fuel burning is the main contributor to the global annual emissions,
which have increased by a factor of about 10 since 1900 to an enormous
6.1 x 109 tonnes in 1994. Deforestation adds about another 1.6 x 109
tonnes per annum.2 This must be considered in relation to the total
atmospheric content of CO2 which is about 750 x 109 tonnes, corresponding to a concentration of around 358 ppmv in 1994 as opposed to
280 ppmv in pre-industrial times. The various components of the overall
global balance of carbon dioxide are generally understood but not easily
quantified. Figure 3 shows the global carbon cycle and carbon reservoirs.
CO2 is removed from the atmosphere by photosynthesis in plants thus
fixing CO2 into a biomass reservoir. CO2 is released in the processes of
respiration and decay and these processes are naturally in balance unless
we destroy the biomass reservoir or burn fixed forms of carbon. The
oceans contain vast amounts of CO2 in inorganic form as well as in
association with living organisms such as plankton. Exchange of gas
between the atmosphere and the upper layers of the ocean is rapid and
subsequent transfer to deep ocean regions slow. In some areas there may
be net release of CO2 and in other areas net removal but overall the
oceans represent a net sink for CO2 although on a slow time-scale. It is
estimated that the time taken for the atmosphere to adjust to changes in
sources and sinks of CO2 is between 50 and 200 years although this is
difficult to quantify because each part of the carbon cycle has its own
time-scale.
1.2.3 Global Warming. The rate of concentration change of CO2 and
other greenhouse gases is shown in Table 2. What is important however
is not just the rate of increase but the effect each species could have on
global warming. Molecule for molecule changes in CH4, N2O and the
CFCs have more effect than changes in CO2 although their overall
concentrations are lower. The Global Warming Potential (GWP) is a

Atmosphere
750

Fossti fuels and
cement production
Vegetation 610
Soils and detritus 1580
2190
Surface ocean
1020
Marine biota
3

Dissolved
organic carbon
<700

Intermediate and
deep ocean
38,100

Surface sediment
1SO
Figure 3

The carbon dioxide cycle, showing the reservoirs (in GtC) and fluxes (GtC/yr)
relevant to the anthropogenic perturbation averages over the period 1980 to 1989

(Reproduced with permission from the Intergovernmental Panel on Climate
Change2)

quantified measure of the relative effect of each species on radiative
forcing of the atmosphere, including both direct and indirect effects. The
index is defined as a cumulative radiative forcing between the present
time and some specified time in the future caused by a unit mass of gas
emitted at present, relative to CO2.2 The total global warming effect of
each gas is then determined by multiplying by the amount of gas emitted.
A typical uncertainty in the figures is about 35%. Table 2 demonstrates
the GWPs for a number of greenhouse gases. Although CO2 is the most
important contributor, the other gases taken together contribute about
half the overall radiative forcing. Some of the species represent CFCs
and their replacements following the Montreal Protocol. Although these
species have high GWPs their concentrations are small and their total
impact less than 3%. We will return to these species in Section 1.3. The
situation with ozone is quite complex and depends on its vertical
distribution with particular sensitivity around the tropopause. The
reduction in lower stratospheric ozone over the last 15-20 years has
been shown to have a slight negative effect on global warming. In the
troposphere there appears to be a gradual increase in the level of ozone

Table 2

Concentration changes, lifetimes and global warming potentials of greenhouse gases2

Species
CO2
CH4
N2O
CFC-Il
CF4
HCFC-22
(a CFC substitute)

Pre-industrial
concentration

Concentration
in 1994

Rate of
concentration
change

Atmospheric
lifetime
(years)

GWP (time
horizon 20 years)

GWP (time
horizon 100 years)

280 ppmv
700 ppbv
275 ppbv
0
0
0

358 ppmv
1720 ppbv
312ppbv
268 pptv
72 pptv
110 pptv

0.4% yr" 1
0.6% yr" 1
0.25% yr" 1
0%yr~1
12% yr" 1
5%yr~1

50-200
12
120
50
50000
12

56
280
4500
4400
1500

21
310
3500
6500
510

due to emissions of NO x and hydrocarbons, and this will have a positive
effect on global warming. The net effect of changing ozone levels is
predicted to be positive although small and may vary from region to
region. Aerosols (including particles, small droplets and soot) may also
affect global warming by either scattering and absorbing radiation or
through their effects on clouds. Although the effect of aerosols shows a
complex dependency on their size and distribution there have been
significant advances in quantifying their contribution to global
warming, and current models predict they have a negative (i.e. a
cooling) overall effect. Aerosols are very short-lived species and hence
their radiative forcing will respond quickly to changes in emissions. An
example of this is the cooling effect caused by the 1991 volcanic eruption
of Mount Pinatubo.
1.2.4 Climate Change. There now seems to be some consensus that
mean surface temperatures have been increasing since the late 19th
century at a rate over and above natural variability. However, the
climatological consequences of global warming are still not well understood. Modelling the effect of increased greenhouse gas levels on the
global climate is an enormously complex problem requiring high performance computers. Three-dimensional global circulation models describe
the vertical, latitudinal, and longitudinal variations in conditions and
attempt to compare the present situation with various scenarios for future
emissions based on predicted population and economic growth, energy
availability, etc.2 The feedback processes described above are being better
and better represented in the models as is the coupling between atmosphere and ocean. Predictions show that changes in surface and atmospheric temperatures, cloud cover, evaporation and precipitation, etc. are
all affected by the changed radiation balance but the effects are not
equally distributed over the globe. For a medium emissions scenario,
predicting CO2 doubling by 2100, most models now suggest an increase of
about 2 0C in mean global temperatures relative to 1900,2'5 with the
largest increases of 8-100C over high Northern latitudes as shown in
Plate 1*, 2-3 0C in Europe and N. Asia (> 50 N) in winter, and increases of
up to 4 0C in Antarctica. Although the global precipitation might increase
by 5-10%, the studies suggest that the tropics and areas bordering the
eastern coasts of continents would become generally wetter and the subtropical regions become drier. This might be critical for some central
African regions which already suffer severe drought conditions. A rise in
sea level resulting from the melting of ice-caps and thermal expansion of
the oceans is predicted at 15-50 cm by 2100. A significant thinning of the
Arctic ice seems to have occurred already and is currently being studied.
* Plates are between pages 48 and 49.

1.2.5 International Response. International discussions have been
taking place for some years with a view to limiting the emissions of
greenhouse gases. The second World Climate Conference met in Geneva
in 1990. It had as its technical basis a report4 from the UN Intergovernmental Panel on Climate Change, an international body of 300 scientists.
A follow-up meeting in Berlin (1995) agreed that targets suggested by the
Rio summit were inadequate and industrialized nations should act
within a shorter time. The third session of the Conference of the Parties
(COP)5 to the Climate Change Convention took place in Kyoto in
December 1997 where agreements were finally made and a Protocol
established.6 The Parties to the Convention, shown in Table 3, have
agreed individually or jointly to ensure that their aggregate anthropogenic carbon dioxide equivalent emissions of the greenhouse gases (CO2,
CH4, N2O, hydrofluorocarbons, perfluorocarbons, and sulfur hexafluoride) do not exceed their assigned amounts by the year 2010. The
countries who have signed the Protocol are listed below along with their
allowed emissions as a percentage of their emissions in the base year,
1990. Some countries (marked with an *) are agreed to be undergoing the
transition to a market economy and therefore their base year is different.
Reductions are expected to be achieved in a number of ways including:
the enhancement of energy efficiency; sustainable forest management
and agricultural practices; the development and increased use of new and
renewable forms of energy and carbon dioxide sequestration technologies; the progressive reduction or phasing out of tax exemptions and
subsidies that are counter to the objective of the Convention; measures to
limit and/or reduce emissions of greenhouse gases in the transport sector;
and the limitation and/or reduction of methane emissions through
Table 3

Annex B of the Kyoto Protocol—Agreed reductions or limitations in
emissions of greenhouse gases not covered by the Montreal Protocol6

Party

Agreed emission
limitation or reduction
(% base year)

Iceland
Australia
Norway
New Zealand, Russian Federation*, Ukraine*
Croatia*
Canada, Hungary*, Japan, Poland*
United States of America
Bulgaria*, Czech Republic*, Estonia*, European Community,
Monaco, Switzerland, Latvia*, Liechtenstein, Lithuania*,
Romania*, Slovakia*, Slovenia*

110
108
101
100
95
94
93
92

recovery and use in waste management as well as in the production,
transport and distribution of energy.
Developing countries remain exempt for the present although there is
a 'clean development mechanism' aimed at promoting sustainable
development through technology transfer. Without this it is likely that
emissions from developing nations will rise steeply and may account for
60% of emissions over the next two decades.
It is probable that the wider use of natural gas as a substitute for coal
will result in some benefit since the mass of CO2 emitted per unit of heat
released is less: gas 0.43, oil 0.62, coal 0.75 ktonne (MWyr)~ ! . This
trend will occur in the UK as natural gas is increasingly used for power
production and the amount of coal used is reduced. However, many
developing nations such as India and China have large coal reserves. The
other alternatives are the use of renewable energy sources such as wind,
solar, wave, and tidal power or the further development of nuclear
energy. This seems unlikely in the short term for both economic and
environmental reasons. The use of biomass is also a possibility on a local
scale since the replanting of biomass fuels makes it a sustainable energy
source. National emissions of CO2 for several countries are listed below.
Figures for China and former USSR are calculated estimates.
UK (1994) 551 Tg7
USA (1995) 4786Tg8

1.3

Germany (1994)
874 Tg7
Former USSR (1995) 3804Tg9

France (1994) 308 Tg7
China (1995) 2389 Tg9

Depletion of Stratospheric Ozone 10 " 20

13.1 The Ozone Layer. Although ozone occurs in the troposphere
and plays an important role in air pollution chemistry, about 90% of the
7

http://www.aeat.co.uk/netcen/corinair/94/corin94.html 'European topic centre on air emissions',
December 1997.
http://www.epa.gov/globalwarming/inventory/ 'US EPA global warming emissions inventories'.
9
http://cesimo.ing.ula.ve/GAIA/Countries/co2_total.html
10
UK Review Group on Stratospheric Ozone, 'Stratospheric Ozone', HMSO, London, 1987.
11
UK Review Group on Stratospheric Ozone, 'Stratospheric Ozone 1988', HMSO, London, 1988.
12
UK Review Group on Stratospheric Ozone, 'Stratospheric Ozone 1990', HMSO, London, 1990.
13
http://www.unep.ch/ozone/home.htm 'The Ozone Secretariat WWW Home Page: UNEF,
December 1997.
14
http://www.al.noaa.gov/WWWHD/pubdocs/WMOUNEP94.html 'Executive summary of the
WMO/UNEP Scientific Assessment of Ozone Depletion: 1994 document', December 1997
15
'Scientific Assessment of Ozone Depletion: 1994', World Meteorological Organization, Global
Ozone research and Monitoring Project, Report No. 37, WMO, Geneva, 1995.
16
J. C. Farman, B. J. Gardiner, and J. D. Shanklin, Nature, 1985, 315, 207.
17
http://www.epa.gov/ozone/mbr/mbrqa.html 'The US EPA Methyl Bromide Phase Out Web Site',
December 1997.
18
http://jwocky.gsfc.nasa.gov/ 'The TOMS home page', December 1997.
19
S. Solomon and D. L. Albritton, Nature, 1992, 357, 33.
20
http://www.he.net/~afeas/index.html 'AFEAS, Alternative Fluorocarbons Environmental
Acceptability Study', December 1997.
8

total ozone content of the atmosphere occurs in the stratosphere at
altitudes between 15 and 50 km. The ozone layer acts as a filter for
ultraviolet radiation from the sun, removing most of the radiation below
300 nm. This serves to protect humans from the adverse effects of UV
which become significant below 320nm since decreasing wavelength
corresponds to higher energy photons which can cause sunburn and
types of skin cancer. Any depletion of stratospheric ozone would therefore lead to a larger amount of UV radiation incident on the earth's
surface and an increased risk of the induction of cancers.
Concern was first expressed about this risk in the early 1970s in
connection with emissions of nitrogen oxides from supersonic aircraft
such as Concorde, which fly in the lower stratosphere. Nitrogen oxides
are potential catalysts for the destruction of ozone. This particular effect
is now thought to be relatively minor and attention switched in the 1980s
to halogen compounds, especially CFCs or freons. Freons are a group of
chlorofluorocarbons which have been used as aerosol propellants,
refrigerants and as gases for the production of foamed plastics. Their
attraction lies in the fact that they are non-toxic, non-flammable and
chemically inert. Global production of the two commonest gases, CFC
11 (CFCl3) and CFC 12 (CF2Cl2), rose rapidly from below 50,000 tonnes
per annum in 1950 to 725,000 tonnes per annum by 1976 decreasing
slightly to 650,000 tonnes in 1985. About 90% was released directly to
the atmosphere while the remainder, representing refrigerant use, will be
released when the equipment is eventually discarded. The actual concentration of CFCs in the atmosphere is extremely small, (less than lppb, see
Table 2), but has risen dramatically this century at a rate which correlates
well with known emissions. This rise in CFCs in the 1980s has clearly
affected stratospheric ozone levels via the processes described below and
has been the subject of control measures in the 1990s as will be discussed
later.
Because they are chemically inert CFCs are resistant to attack by
molecules, radicals, or the UV radiation present in the troposphere and
are not subject to significant dry deposition or rain-out. The higher
energy UV radiation in the stratosphere can, however, lead to photodissociation forming chlorine atoms which can in turn lead to the
destruction of ozone. Despite the slow exchange of air between the
troposphere and the stratosphere this effect is now known to be
significant.
1.3.2 Ozone Depletion. The chemistry of ozone depletion is complex
but a basic outline of the important processes is as follows. Ozone is
formed from the dissociation of molecular oxygen by short wavelength
UV radiation in the upper stratosphere:

(1)
(M = inert third body)

(2)

However, ozone itself is rapidly photodissociated:
(3)
and the so-called 'odd oxygen' species and O 3 may interconvert many
times before they destroy one another by:
(4)
In fact, measurements of the ozone profile in the atmosphere suggest that
ozone destruction must be considerably faster than could be achieved by
reaction (4) alone and that other reactions must be involved. These other
mechanisms can be represented by
X + O3 -> XO + O2
XO + O' -> X + O2
Net effect

(5)
(6)

O" + O3 -> O2 + O2

X may represent a range of species including Cl", Br*, NO, OH* and H*
and is not consumed in the overall ozone destruction process. If X =
NO, the reactions form and destroy NO 2 ; if X = Cl*, the reactions form
and destroy ClO, but because this is a catalytic cycle small concentrations of X can have a significant effect on ozone levels. Other sets of
reactions involving NO and Cl* simply achieve the interconversion of O 3
and O* and therefore have no effect on the net ozone levels. The reactive
NO x and Cl* species can be removed by the formation of the relatively
stable 'reservoir' molecules HNO 3 and HCl or the somewhat shorter
lived chlorine nitrate ClONO 2 . About half the stratospheric content of
NO x is stored as HNO 3 and about 70% of the chlorine as HCl. Although
these may be reactivated by conversion back to NO x and Cl*, they may
eventually be transferred back to the troposphere and removed to the
ground by rain-out.
7 J J The Antarctic Ozone 'Hole'}°~X6 The above was the general
picture (although a highly simplified account) of the homogeneous
stratospheric chemistry as understood before 1985. In that year Farman
et al}6 published the results of ground-based measurements in Antarctica showing very significant depletions, of the order of 50%, in the total
column ozone content of the atmosphere. The Antarctic ozone 'holes' of

1992 and 1993 were the most severe on record, with ozone being locally
depleted by more than 99% between about 14 and 19 km in October,
1992 and 1993. Subsequent aerial surveys and analysis of satelite data
confirmed this phenomenon and led to a complete reappraisal of the
chemistry and meteorology involved.
During the dark, cold Antarctic winter upper stratospheric air moving
from low to high latitudes subsides and as it does so develops a strong
westerly circulation pattern. This produces a vortex which effectively
isolates the air in the lower stratosphere over the Antarctic continent
from the air at lower latitudes. Within the vortex the temperature falls
progressively until below about — 80 0 C polar stratospheric clouds
(PSCs) may form. These are composed of very small particles (1 /im) of
nitric acid trihydrate (HNO3.3H2O). A further drop in temperature of
about 5 0C may result in water ice crystals being formed. These are rather
larger (10/im). It is the heterogeneous reactions involving these cloud
crystals which dramatically alter the chemistry of the stratosphere.
Basically these reactions convert chlorine from its inactive, reservoir
forms (HCl, ClONO2) into forms which are active ozone depletors (Cl\
ClO). HCl is readily incorporated into ice crystals and can undergo
reaction with ClONO2 :
(7)
and
(8)
The nitric acid is left in the ice phase. The chlorine remains in the gas
phase until the polar spring when the sun reappears and photodissociates
it to chlorine atoms:
(9)
The Cl* atoms rapidly react with ozone generating ClO:
(10)
In the winter the stratosphere is thus chemically 'preconditioned' by
heterogeneous reactions so that in the spring very rapid ozone depletion
occurs. In addition to the ozone destruction cycle represented by
reactions (5) and (6) with X = Cl* it is now recognized that chlorine
monoxide dimers are also important. This was realized because the
oxygen atom concentrations in the lower stratosphere are too low to
account for the observed ozone destruction rates.

(H)
(12)
(13)
These reactions by-pass the ClO + O* reaction as a route for reconversion of ClO back to Cf.
The importance of ClO within the Antarctic stratosphere is illustrated
in Figure 4. Reactions of bromine atoms in addition to those of chlorine
atoms are now thought to account for about 20% of the ozone depletion.
Bromine emissions occur in the form of methyl bromide which has
natural and synthetic sources such as soil fumigation, biomass burning,
and the exhaust of automobiles using leaded gasoline, plus another
family of halocarbons the Halons—Halon 1301 is CF3Br, Halon 1211 is
CF2BrCl (Table 4). Methyl bromide is currently being targeted for phase
out and is covered under amendments to the Montreal Protocol.17
Current models of ozone depletion also consider the effects of CO2.
Increased CO2 levels would lead to a lowering of temperatures in the
stratosphere which may serve to slow down the destruction reactions:
O + O 3 and NO + O3. The role of sulphate aerosols also has to be
considered in the lower stratosphere.15

3.5 ppmv

Chemically perturbed
region

10 ppbv
0.75 ppbv
250 ppbv

5pptv
Polewards
Figure 4

Profiles of ClO and other species in the Antarctic stratosphere, 18 km altitude,
near the boundary of the chemically perturbed region. The decreases in water
vapour and NOx are due to condensation of water and nitric acid in polar
stratospheric clouds followed by gravitational settling

(Reproduced with permission from UK Review Group on Stratospheric
Ozone11)

Table 4

Atmospheric lifetimes and ozone depletion potentials for halogenated
compounds

Compound name

Chemical formula

Atmospheric
lifetime2'12 (yr)

Ozone depletion
potential 3

CFCIl
CFC 12
CFC 113
Carbon tetrachloride
Methyl chloroform
HCFC 22
HFC 134a
Halonl301
Halonl211
Methyl bromide

CFCl3
CF2Cl2
C2F3Cl3
CCl4
CH3CCl3
CH3CCl3
C2H2F4
CFBr
CF2BrCl
CH3Br

50
102
85
42
4.9
12.1
14.6
65
20
1.2

1.0
1.0
1.1
1.08
0.12
0.05
0
12.5
3.0
0.6

But is the 'hole' in the Antarctic ozone layer significant for the rest of
the globe? Several facts suggest that it is. As the Antarctic spring
progresses the vortex breaks up and the ozone depleted air can then be
transported to lower latitudes. Such an event has been observed in
Australia. Large increases of surface UV are observed in Antarctica and
the southern part of South America during the period of the seasonal
ozone 'hole'. In the Northern Hemisphere airborne studies of the Arctic
winter were carried out in 1988/9. Although the temperatures are not as
low as in the Antarctic and the occurrence of stratospheric clouds not as
common, nevertheless they are formed and the existence of a similar
chemistry with high ClO concentrations has been demonstrated. The
extent of ozone depletion is less marked—perhaps 15-20% in the range
20-25 km altitude representing a reduction of some 3% in total column
ozone, with the worst years again 1992/1993 as in the Antarctic. The link
with chlorine and bromine levels has been proven although is more
difficult to quantify in the Arctic due to a larger uncertainty in the
dynamics of the Arctic vortex. Data from the network of ground level
monitors for column ozone show a clearly decreasing trend in the
Northern Hemisphere since 1970, although the considerable variability
in the data makes the precise percentage decrease sensitive to the start
date assumed. For Europe and North America the decrease is 2.5 to
3.5% per decade with an indication that the trend has accelerated in the
last decade, in parallel with the worsening conditions in the Antarctic.
Data from the Total Ozone Mapping Spectrometer (TOMS) shown in
Plate 2 illustrates the ozone depletion in the Northern Hemisphere.18
1.3.4 Effects of International Control Measures. The UN Convention
on the Protection of the Ozone Layer (the 'Vienna Convention') was

agreed in 1985 and subsequently measures to reduce the emissions of
various halocarbons were incorporated into the 'Montreal Protocol' in
September 1987. Meetings in London in 1990 and Copenhagen in 1992
further tightened the restrictions under the Protocol which has as its final
objective the elimination of ozone depleting substances. More than 160
countries are now Parties to the Convention and the Protocol13 with
some special agreements for developing countries. The use of both CFCs
and Halons will be phased out by the year 2000 as will the use of carbon
tetrachloride CCl4. Methyl chloroform will be phased out by 2005.
Replacement chemicals such as the HCFCs are now being used and the
change in reported emissions20 is illustrated in Figure 5. Because these
contain hydrogen atoms as well as halogens they are more reactive in the
troposphere and have shorter lifetimes as illustrated in Table 4. Their
potential impact on the stratosphere is therefore much reduced.
Evidence is now emerging which indicates that the atmospheric
growth rates of the main ozone depleting substances are slowing,
showing that the Montreal Protocol is having an effect. Total tropospheric organic chlorine increased by only about 60 ppt/year (1.6%) in
1992, compared to 110 ppt/year (2.9%) in 1989.14 Even so, the expectation is that the total stratospheric chlorine loading will peak after 2000 at
about 4 ppb and only decline slowly through the remainder of the 21st
century.12 Figure 6 shows the predicted fall in stratospheric chlorine in
the 21st century but also demonstrates the sources of this chlorine.20
CFCs are predicted to be a major source well into the next century
because of their long lifetimes. How long it takes for the chlorine loading
to fall to the pre-war level of below 1 ppb depends on the extent of global
participation in implementing restrictions and the extent of future use of
alternative chlorine containing compounds such as the HCFCs (Table 4).
Global ozone losses and the Antarctic ozone 'hole' are predicted to
recover in about the year 2045 as long as the Montreal Protocol and
amendments are closely adhered to.

2 ATMOSPHERIC TRANSPORT AND DISPERSION OF
POLLUTANTS
2.1 Wind Speed and Direction
In the previous section we dealt with global pollution problems. We now
move on to local pollution problems which we think of as directly
affecting the air quality around us. Local air quality is significantly
influenced by the rate of mixing of pollutants which are emitted and
therefore we need to have an understanding of issues such as wind speed

Year
Figure 5

Annual Production of Fluorocarbons Reported to AFEAS 1980-1995

(Reproduced with permission from the Alternative Fluorocarbons Environmental Acceptability Study20)

pptv

Year
Figure 6

Predicted chlorine loading for the next century by source. Bromine from all
sources is shown as its equivalent in chlorine

(Reproduced with permission from the Alternative Fluorocarbons
Environmental Acceptability Study20)

and atmospheric stability. In general, low wind speeds result in high
pollutant concentrations and vice versa. If we imagine the wind blowing
across the top of a chimney emitting smoke at a constant rate, the
volume of air into which the smoke is emitted is directly proportional to
the wind speed. The concentration of smoke in the air is thus inversely
proportional to the wind speed. A similar description can be applied to a
source distributed uniformly over a wide area, e.g. domestic emissions
from a city. The concentration of pollutants in the hypothetical box of
air into which the pollutants are mixed is proportional to the emissions
rate and inversely proportional to the wind speed. In practice this picture
is grossly oversimplified and the concentration of pollutants measured in
urban areas rarely decreases with wind speed as rapidly as predicted.
For reasons which will be discussed later, the wind speed at ground
level tends to drop overnight and rise again during the morning,
especially during cloud-free conditions. Of course, emissions also tend
to drop overnight—fewer fires, boilers and furnaces are alight, fewer cars
are on the roads. So some of the highest pollution levels occur in the
morning when emissions increase rapidly before the wind speed picks up
and dispersion conditions improve.

The people most affected by air pollution are those who are situated
downwind of the major sources. A knowledge of the prevailing wind
direction is therefore important in predicting the likely impact of these
sources. The wind direction at a point is not sufficient to identify high
pollution levels with the effect of distant sources. Over the scale of
hundreds to thousands of kilometres the overall path of the particular
mass of air must be computed for periods of perhaps 1 to 3 days. Such
trajectories can be quite curved as illustrated in Plate 3, which relates to
high smog episodes over the UK. The origin of the smog precursors is
clearly the pollutant emissions from the southern UK and/or continental
Europe. Corresponding trajectories from the Atlantic generally bring
much less polluted air.
2.2

Atmospheric Stability

In addition to wind direction, the extent of vertical mixing in the
atmosphere also affects pollutant concentrations. This is related to the
stability of the atmosphere which is dependent on many factors such as
the time of day, the synoptic weather conditions, the nature of the earth's
surface, etc.
2.2.1 The Lapse Rate. The roughness of the ground produces a
certain amount of turbulence in the lowest layer (boundary layer) of the
atmosphere which promotes the mixing and dispersion of pollutants.
This effect increases with the scale of the surface roughness and is greater
for a city with large buildings than for open ground with few obstructions. However the major factor affecting atmospheric stability and
turbulence is thermal buoyancy. The pressure in the atmosphere
decreases exponentially with height. Ascending air expands as the
pressure decreases and as it expands it cools. A simple calculation based
on the properties of gases leads to the conclusion that we should expect a
decrease in temperature with height of 9.86 0C km" 1 or about 10C for
every 100 m of dry air. The figure is somewhat lower for moist air. The
variation of temperature with height is called the lapse rate and the
calculation for the ideal case leads to what is known as the adiabatic
lapse rate (a.l.r.). In the real atmosphere the lapse rate can be greater
than, smaller than, or close to the adiabatic lapse rate. This fundamentally affects the extent of vertical mixing of air as the following two
examples show:
Case (a). Temperature decreases with height more rapidly than the
a.l.r., Figure 7a. Air that is slightly warmer than its
surroundings starts to rise and to cool at the a.l.r. The

Height

Temperature

Height

Temperature
Figure 7

Schematic illustration of the atmospheric lapse rate for various stability
categories. Full lines: actual temperature profile; dashed lines: adiabatic lapse
rate (a.Lr.)

temperature difference between the rising parcel of air and
its surroundings increases with height, thus the upward
movement due to thermal buoyancy continues and is
magnified. The atmosphere is unstable. Upcurrents in one
location are balanced by downcurrents elsewhere and rapid
vertical mixing of air occurs, promoting rapid dispersion of
pollutants.
Case (b). Temperature decreases with height less rapidly than the
a.Lr. or actually increases with height, Figure 7b. Air that is
slightly warmer than its surroundings starts to rise and cool
at the a.l.r. and the temperature difference between the
rising parcel of air and its surroundings soon decreases to
zero. Upward movement due to thermal buoyancy ceases.
The atmosphere is stable since any vertical movement of air
tends to be damped out. Lower polluted layers stay near the
ground and pollutant concentrations will be high.

Unstable situations occur with bright sunlight warming the ground to
a temperature above that of the air. The air adjacent to the ground is
subsequently warmed and rises due to its buoyancy. Such situations are
common during daytime in summer, especially when the wind speed is
low. High wind speeds tend to lead to neutral conditions with the lapse
rate close to the adiabatic value (Figure 7c).
2.2.2 Temperature Inversions. Stable situations occur when the lowest
layer of air is cooled by the ground beneath. The most common cause is
overnight radiative cooling of the ground which often leads to low level
temperature inversions on clear nights. The temperature profile, in
Figure 7c which may represent late afternoon, gradually changes into
profile 7d overnight. The effect is that ground level emissions become
trapped in the stable inversion layer which may not be more than 100200 m deep. Emissions from tall industrial chimneys however may be
above the inversion layer and be vertically dispersed by relatively good
mixing conditions aloft. The following day the inversion layer is
gradually eroded by the warming effect of the sun until, by midmorning, the temperature profile has returned to that of neutral conditions (Figure 7c) and the trapped pollutants are effectively released to be
dispersed to higher levels.
Another factor contributing to high pollution levels during inversion
conditions is the lowered wind speed. Since high level, faster moving air
does not mix with low level air, there is no mechanism for the downward
transport of momentum. The lowest level air therefore becomes stagnant, In the low temperature conditions, dew, frost, or fog formation
may occur. Fog adds to the problem by slowing down the break-up of
the overnight inversion layer because the sun's energy is reflected by its
upper surface and does not reach the ground. The ground therefore stays
cool rather than warming. In extreme cases fog may persist for several
days as happened in the London 1952 smog. Polluted fogs are more
persistent than clean fogs because the chemicals dissolved in the water
droplets prevent complete evaporation even when the relative humidity
drops well below 100%.
The other main type of inversion occurs during anticyclonic conditions and is described as a subsidence inversion. Within an anticyclone,
air is diverging from a high pressure region and at the centre subsides
from a high level to lower levels of the atmosphere. As it subsides it
progressively warms with decreasing height resulting in the development
of an elevated inversion layer as illustrated in Figure 7e. Below the
inversion layer the air may be neutral or even unstable so that good
mixing occurs but only up to the inversion height. Local urban pollution
levels are rarely as great under such conditions as during ground level

inversions. However, subsidence inversions are often associated with
warm, dry weather and they provide the ideal conditions for a long range
transport of pollution. Summer haze conditions, in which the UK
receives already polluted air from Europe before the addition of our
own emissions, can lead to particulate pollution levels as high as those on
the most polluted winter days. At the same time the levels of photochemical oxidants such as ozone are also high as shown in Plate 3.
In the discussion so far, no reference has been made to the geographical situation in which air pollution levels are being considered. Towns
situated in valleys are particularly susceptible to pollution problems.
Cool air will tend to flow downhill into the valley so aggravating the
problem of low level inversions. Mixing between the air in the valley and
the air above is reduced. Fogs will persist longer. Often in winter, a layer
of polluted air over the town with cleaner air above can be clearly seen
from the higher ground. Towns situated by the sea may be subject to sea
breezes. The proximity of relatively warm ground to the cool water
surface results in a circulation of air from sea to land. The sea breeze will
be cooler than the overlying air, another example of inversion conditions, so in polluted areas conditions will be worse than for a corresponding inland site. Sea mists may blow inland aggravating the general
discomfort. Los Angeles is one example of a city whose geographical
location exagerates pollution problems caused by high emissions. It is
situated in a basin area surrounded by large hills which inhibit mixing
out of the city. Inversions are also common due to the presence of
subsidence conditions over the Pacific. In fact, five of the worst polluted
cities of the world are situated in the Pacific basin region.21

2.3

Dispersion from Chimneys22

2.3.1 Ground-level Concentrations. The effluent gases leaving a
chimney gradually entrain more and more air, and the plume width both
in the vertical and in the horizontal directions increases with downwind
distance. In the cross wind direction there is a rapid fall-off in concentration as we move away from the plume centre line which is defined as
being along the average wind direction. The ground-level concentration
very close to the chimney will be zero because it takes some time for
vertical mixing to drive the plume to the ground. Some distance downwind the dispersing plume reaches ground level and the concentration
21

D. Mage , G. Ozolins, P. Peterson, A. Webster, R. Orthofer, V. Vandeweerd, and M. Gwynne.
Atmos. Environ., 1996, 30, 681. This is a follow up report to the WHO/UNEP report 'Urban Air
Pollution in Megacities of the World' published in 1992 in which the detailed data are included.
22
'Recommended Guide for the Prediction of the Dispersion of Airborne Effluents', ed. J. R.
Martin, American Society of Mechanical Engineers, New York, 3rd Edn., 1979.

Unstable
Concentration

Neutral
Stable

Downwind distance
Figure 8

The variation in ground-level concentration along the plume centre line
downwind of a chimney for various stability classes

rises rapidly to a maximum value. For example, for Eggborough Power
Station in northern England this occurs up to about 8 km from the
200 m chimney. Thereafter the concentration falls off with distance as
the plume becomes more and more dilute, Figure 8. The distance at
which maximum concentrations occur depends on stability conditions.
Unstable conditions (significant vertical mixing) bring the plume down
to ground rapidly, that is closer to the chimney, but the rate of dilution is
large so the concentration falls rapidly from its maximum value as we go
to greater distances. Stable conditions result in the plume dispersing very
slowly. It may remain visible for a considerable downwind distance. The
point of maximum ground level concentration is a long way from the
chimney. Above the chimney top, the vertical dispersion of the plume
can be hindered by an inversion layer. The pollutants are then trapped
between the inversion height and the ground. At sufficient distance
downwind, the concentration may be virtually constant at all heights as
the plume becomes well dispersed within the mixing layer.
2.3.2 Plume Rise. The basic models of plume dispersion suggest that
the maximum ground level concentration depends inversely on the
square of the chimney height. A 40% increase in chimney height should
roughly halve the ground level impact. However, the height of the plume
depends on both the chimney height and the plume rise caused by the exit
conditions of the release. The major factor is the thermal buoyancy of the
plume. The hot flue gases will rise relative to the cooler surrounding air
and the hotter the release the greater the plume rise. A lesser factor is the
vertical momentum of the gases due to their efflux velocity out of the

chimney top. Plume rise can mean that the effective chimney height is
double the physical chimney height under low wind speed conditions and
so is a considerable asset in achieving effective dispersion. Conversely,
any control technology for pollutant reduction which also reduces the
flue gas temperature (such as a scrubber) results in poorer dispersion of
the plume. If we take plume rise into account then the effect of raising the
stack height to reduce ground level concentrations is less than would
otherwise be predicted.
2.3.3 Time Dependence of Average Concentrations. The above
description of plumes is really a simplification since atmospheric turbulence is random, making the dispersion process irregular. Visible plumes
will often look ragged and, from the point of view of an observer making
spot measurements on the ground, one minute the concentration may be
close to zero and the next very high. The longer averaging time taken, the
less the variability of the results. This can be illustrated by the results of
measurements made by the Central Electricity Generating Board near
Eggborough, a 2000 MW power station.23 At the radius of maximum
effect, which is about 8 km from the chimney, the peak 3 minute
concentrations of sulfur dioxide due to the power station were approaching 1000 fig m~3. Such peaks were extremely rare and 95% of the 3
minute averages were below 10 fig m~ 3 . The highest daily averages were
around lOOjugm"3 and the highest monthly averages around
10 jug m~3. The overall annual average contribution of the power
station to the ambient SO2 concentration was 2-3 fig m~ 3 in an area
where the prevailing concentration from other sources was about
40 /igm~ 3 .
2.4

Mathematical Modelling of Dispersion 24 " 37

There are a number of techniques for modelling the dispersion and
reaction of atmospheric pollutants. Space does not allow us a full study
23

A . J. Clarke, 'Environmental Effects of Utilising M o r e Coal', ed. F . A . R o b i n s o n , R o y a l Society of
Chemistry, L o n d o n , 1980, p . 55.
24
J o h n H . Seinfeld a n d Spyros N . Pandi, 'Atmospheric Chemistry a n d Physics, from Air Pollution
to Climate Change', J o h n Wiley & Sons, N e w York, 1998, p p . 916-919, 1193-1285.
25
D . G . Atkinson, D . T . Bailey, J. S. Irwin, a n d J. S. T o u m a , J. Appl. MeteoroL, 1997, 36, 1088.
26
D . J. Carruthers, H . A. E d m u n d s , K . L. Ellis, C. A. M c H u g h , B. M . Davies, a n d D . J. T h o m s o n ,
Int. J. Environ. Poll, 1995, 5, 382.
27
Z. Zlatev, J. Christensen, and A. Eliassen, Atmos. Environ., 1993, 27, 845.
28
C. Pilinis and J. H. Seinfeld, Atmos. Environ., 1988, 22, 1985.
29
J. S. Chang, R. A. Brost, I. S. A. Isaksen, S. Madronich, P. Middleton, W. R. Stockwell, and C. J.
Walcek, J. Geophys. Res.—Atmospheres, 1987, 92, 14681.
30
W. J. Collins, D. S. Stevenson, C. E. Johnson, and R. G. Derwent, J. Atmos. Chem., 1997, 26, 223.
31
R. I. Sykes and R. S. Gabruk, J. Appl. MeteoroL, 1997, 36, 1038.
32
J. C. Weil, L. A. Corio, and R. P. Brower, / . Appl. MeteoroL, 1997, 36, 982.
33
D. M. Lewis and P. C. Chatwin, J. Appl. MeteoroL, 1997, 36, 1064.

here but the main methods will be introduced. Choosing a particular type
of model depends on many factors such as the distance scale in which we
are interested, e.g. urban roadside, plume dispersion, regional scale smog
modelling, global circulation modelling, etc. We may also need to take
into account whether we need to include chemical reactions and deposition, and the computer resources available. Available methods include
Gaussian formulations,22'24^26 Eulerian grid modelling,27"29 Lagrangian
trajectory modelling30 and turbulence or statistical models which try to
account for random fluctuations.31"34
The simplest approach is the Gaussian model which uses empirical
formulae to describe the distribution of pollutants downwind of a single
source or a group of sources. Fast screening type calculations of the
dispersion of a pollutant from a point source such as a large chimney are
usually based on the Gaussian Plume Model22 which is discussed in
Chapter 7. This treats steady state emissions and situations which can be
described as a superposition of a number of different steady states. For
example long term average distributions around a source can be
calculated by using statistically averaged meteorological data. The
model can be adapted to treat area sources such as wind-blown dust
from a stockpile or odours from a sewage works. Urban areas can be
modelled as a sum of area sources representing domestic and commercial
emissions plus larger point sources such as factories or power stations.
Another application is dispersion from sources distributed along a line
such as a motorway. Some current urban models use Gaussian formulations to predict the dispersion from a range of source types but at present
the validation of such models for such a complex situation is limited.
Because it is a steady-state model the Gaussian plume approach has
limitations in that it cannot represent diurnal variations of pollutants.
On its own it also has limitations in that it can only represent first-order
reaction or deposition processes although it can be coupled with
chemical box models. This is particularly important for secondary
pollutants such as ozone whose concentrations (as we shall see in a later
section) depend on a series of non-linear chemical reactions involving
NO x and hydrocarbon species, and levels of sunlight. Also because it is
steady state the Gaussian model cannot represent the random fluctuations present in a real plume since for this a turbulence model would be
required. The Gaussian model is best suited to averaged concentrations
34

F. Pasquill and F. B. Smith, 'Atmospheric Diffusion', 3rd Edn., Ellis Horwood, Chichester, 1983.
J. Saltbone and H. Dovland, 'Emissions of sulphur dioxide in Europe in 1980 and 1983'. EMEP/
CCC Report 1/86. Norwegian Institute for Air Research, Lillestrom, Norway, 1986.
36
H. S. Eggleston and G. Mclnnes, 'Methods for the compilation of UK air pollutant emission
inventories', Report LR 634(AP), Warren Spring Laboratory, Stevenage, 1987.
37
http://www.aeat.co.uk/netcen/airqual/emissions/den-naei.html 'National Atmospheric Emissions
Inventory for the UK', December 1997.
35

of non-reactive pollutants but has the advantage that it uses a small
amount of computer resources and is therefore commonly used as an air
quality management tool.
The more complex processes involved in long range transport and the
reactions producing smog and acid rain, which also involve a number of
source types, require numerical simulation on large computers. Here we
may choose to use either Lagrangian or Eulerian models. Lagrangian
models follow the trajectories of single or multiple air parcels as they are
transported by the winds over long distances. The Eulerian approach
uses a grid which covers the whole of the domain and therefore equations
have to be solved at each grid point. There are advantages and
disadvantages to both methods. The Lagrangian approach, because it
can simulate only a few air parcels, can have lower computer simulation
times and can therefore accommodate large chemical reaction mechanisms, describing the formation of smog for example. However, if limited
trajectories are used we may not get a full picture and will only simulate
smog concentrations at a few points in the region. Models of regional
transport of SO2 and O3 are often based on large numbers of backtrajectories (see Plate 3) for specified receptor points taking account of
emissions and loss processes along the path. The Eulerian approach gives
much better coverage of the domain but at the expense of computer
simulation time. Often large parallel computers are used for Eulerian
simulations,27"29 e.g. smog models, global climate models, etc.
We can see that the choice of model type depends on the required
outcome and accuracy needed. The accuracy of all types of model,
however, depends heavily on the input data. Detailed knowledge of
emissions data for a range of sources is needed along with an accurate
representation of meteorological data such as wind speed and direction,
stability class, and precipitation. The resolution of the input data
significantly affects the results. Detailed modelling of urban pollution
requires an emissions inventory on a grid basis, for example 1 km
squares, along with point source information. Emissions inventories for
the most common pollutants are now available for the whole of Europe
on a national basis,7 on the 150 km x 150 km grid used by the European
Monitoring and Evaluation Programme (EMEP)35 and on grid sizes
down to 10 km x 10 km for the UK.36'37
3 EMISSIONS TO ATMOSPHERE AND AIR QUALITY
3.1 Natural Emissions
3.1.1 Introduction. The primary components of pure dry air are
nitrogen N2 (78.1%), oxygen O2 (20.9%), argon Ar (0.9%), and carbon

Table 5 Natural emissions ofS3S and N39 compounds
Source

Tg Syr~l

Source

Tg Nyr~l

Volcanoes
Biomass burning
Marine biosphere
Terrestrial biosphere

9.3
2.2
15.4
0.35

Natural total
Anthropogenic

27.25
77

Lightning
Biomass burning
Soil—biogenic
NH 3 oxidation
From stratosphere
Natural total
Anthropogenic

8
5
7
0.9
0.6
22
22

dioxide CO2 (0.035% or 350 ppm). Water vapour is present in amounts
which typically range from 0.5 to 3% at ground level, depending on
temperature and relative humidity. Analysis of air samples reveals the
presence of hundreds of other substances in trace amounts. Some of
these can be explained directly in terms of their emissions either from
natural sources or human activity. Others must have arisen indirectly by
chemical processes in the atmosphere. We distinguish these as primary
and secondary pollutants. The secondary pollutants, including gases like
ozone and particulate compounds such as sulfates, are dealt with in
Section 4. Here we concentrate on the primary pollutants.
Within densely populated areas of land, the levels of most pollutants
are dominated by the contributions for which humans themselves are
responsible. However, nature is also a generator of what we call
'pollutants' and on a global scale the natural emissions may be comparable to human emissions. This is illustrated in Table 5.
3.1.2 Sulfur Species. Sulfur in the form of SO2 and some H2S is
emitted most dramatically from volcanoes (Table 5). Significant
amounts are emitted from biological processes. In the absence of air,
biological decay results in emissions of hydrogen sulfide, H2S, and
organic compounds such as dimethyl sulfide (DMS). Carbonyl sulfide
(OCS) emissions also occur together with small amounts of carbon
disulfide (CS2) and dimethyl disulfide (DMDS). From the oceans the
emissions related to phytoplankton are primarily of DMS. The grouping together of all sulfur compounds is appropriate since H2S and
organic sulfur compounds are converted to SO2 in the atmosphere.
Various estimates place the total emissions at 20-30 Tg sulfur per year
(1 Tg = 1 million tonne). Since the global SO2 from combustion
emissions is 70-100 Tg S yr~\ the natural sulfur emissions are of the
order of one third or one quarter of the anthropogenic emissions.
Anthropogenic emissions are greatest in the northern hemisphere and

for the southern hemisphere natural and anthropogenic emissions are of
a similar order.
This assessment omits the largest single component of the atmospheric
sulfur budget which is represented by the sulfate content of sea-salt
aerosols. There is obviously a sharp vertical gradient of these aerosols
and a large proportion return to the sea quite quickly. Emission figures
of the order of 200 Tg yr ~ ] are quoted for aerosols in the 10-20 m height
range but the particle number concentrations in the free troposphere
(> 2 km) may be 3 orders of magnitude lower than in the oceanic
boundary layer.38
3.1.3 Nitrogen Species. Biological processes in soil lead to the release
of all of the common nitrogen oxides, NO, NO2, and N2O. The amounts
involved are very uncertain but for NO and NO2 are of the order of 510 Tg N yr" 1 . Lightning and biomass burning are other major sources.
Oxidation OfNH3 to NO occurs in the troposphere and some nitrogen in
the form of HNO3 is transferred to the troposphere from the stratosphere. These sources total 20-30 Tg N yr" 1 (Table 5). In comparison
the anthropogenic emissions of NO + NO2 from combustion are about
20 Tg N yr""1. Lee et al.39 estimate the total global NO^ emissions to be
44 Tg N yr ~ l with an uncertainty range of 23-81 Tg N yr ~ l .
The major source of nitrous oxide, N2O, is the release from soil
especially in situations where fertiliser has been added. Smaller releases
occur from the oceans and from combustion processes. The sources are
not as well understood as the main loss mechanism which is decomposition in the stratosphere (6-10 TgN 2 O yr" 1 ). Industrial emissions
occur during the production of nitric acid and adipic acid (an intermediate in the production of nylon). Total UK emissions were
estimated at 94 000 tonnes in 199440 as compared with 470 000 tonnes
in the US in 1995.8
The other important nitrogen species is ammonia, NH3. Its sources
may be considered partly natural and partly anthropogenic since they are
dominated by animal excreta. But there are also significant releases from
biomass burning, crops, and the oceans. Several estimates41 of global
emissions place the total at around 5 0 T g N y r - 1 . The sources of
ammonia are difficult to quantify since the earth's surface can act as a
source and a sink and the overall emission rate depends on soil and
38

T . S. Bates, B. K. L a m b , A. G u e n t h e r , J. D i g n o n , and R. E. Stoiber, J. Atmos. Chem., 1992, 14,
315.
39
D . S. Lee, I. Kohler, E. Grobler, F . R o h r e r , R. Sausen, O. Gallard, L. Klenner, J. G. J. Oliver, and
F . J. Dentener, Atmos. Environ., 1997, 3 1 , 1735.
40
Digest of Environmental Statistics N o . 18, H M S O , L o n d o n , 1996.
41
A. F . B o u w m a n , D . S. Lee, et al., A global high resolution emissions inventory for a m m o n i a ,
Global Biogeochemical Cycles, 1997, 11, 561.

climatic conditions. The official figures for the UK for 1993 are, in
ktonnes yr" 1 :
Cattle
Pigs
Sheep
Poultry
Fertilizers and crops
Agricultural Total
Non-Agricultural sources
Overall total

130
25
15
25
_30
225
_35
260

Currently ammonia emissions in Europe are estimated at 2.7 Mtonnes
yr ~ l for the countries of the European Union and 6.4 Mtonnes yr~ l for
all the European countries covered by the EMEP area (European
Monitoring and Evaluation Programme).42 Currently there are no
international targets for ammonia emissions but future proposals by the
EU are likely to cover ammonia.
3.1.4 Hydrocarbons. The importance of methane as a greenhouse gas
has been discussed earlier. The largest natural sources of methane are
anaerobic fermentation of organic material in rice paddies and in northern wetlands and tundra, plus enteric fermentation in the digestive
systems of ruminants {e.g. cows). Methane is also released from insects,
from coal mining, gas extraction, and biomass burning. Total emissions
are 300-550 Tg yr ~~l and appear to be increasing at a rate of 50 Tg yr ~ l .
Within the UK, animals (29%), mining (8%), landfill gas (46%), and gas
leakage (9%) account for most of the 3.9Tgyr~~1 emissions (1994
estimate ).
Heavier hydrocarbons, such as isoprene, a and jS-pinene and other
terpenes, are released directly to the atmosphere from trees and can
contribute to the formation of photochemical smog in high emission
areas (see Section 4). Global emissions of these Biogenic Volatile Organic
Compounds (BVOCs) are estimated to be 1150 Tg C yr" 1 of which 88%
is from trees and shrubs and 10% from crops.43
3.2

Emissions of Primary Pollutants

3.2.1 Carbon Monoxide and Hydrocarbons. Efficient combustion can
be achieved in most stationary combustion appliances providing they are
42

'Acid Deposition in the United Kingdom 1992-94', Fourth Report of the Review Group on Acid
Rain, Department of the Environment, Transport and the Regions, London, 1997.
43
A. Guenther, C. N. Hewitt, et at., J. Geophys. Res. 1995, 100, D5, 8873.

Table 6

Typical urban emissions from current European technology cars
(g k m ~ l ) assessed in on-road tests44

Vehicle type

Hydrocarbons

NOx

Standard petrol without catalyst
Petrol with catalyst
Diesel

27
2.0
0.9

12.8
0.2
0.3

1.7
0.4
0.8

0.4

Table 7

Present European Standards for passenger cars applicable to all models
from 1.1.97 (Directive 94/'12/'EC)

Vehicle type

CO
gkm"1

HC + NOx
gkm"1

PM
gkm"1

Evaporative losses
g/test

Petrol
IDI Diesel
DI Diesel*

2.2
1.0
1.0

0.5
0.7
0.9

0.08
0.10

2.0

*DI diesel limits to be equal to those for IDI engines from 1.10.99.

properly adjusted, and CO and hydrocarbon concentrations do not give
serious cause for concern. Faulty or improperly adjusted appliances can
produce dangerous amounts of CO (several per cent in the flue gas)
usually due to some abnormal limitations of the air supply.
CO emissions from internal combustion engines are more of a
problem. In petrol engines lacking any control devices, incomplete
burnout of the fuel leads to high carbon monoxide and significant
hydrocarbon emissions, especially during idling and deceleration. Installation of catalytic converters results in reductions of 90% for CO and
hydrocarbons although significant emissions may take place under cold
start conditions before the catalyst 'lights off'. Diesel engines have much
lower carbon monoxide and hydrocarbon emissions (Table 6).44
Permitted emissions from European vehicles have been progressively
reduced and the current standards are indicated in Table 7. Further
reductions have been proposed for the years 2000 and 2005.
The term Volatile Organic Compounds (VOCs) is used to describe
organic material in the vapour phase excluding methane. There are many
non-combustion sources of VOC emission of which the most important
is the use of solvents, including those released from paints. Evaporative
losses of gasoline during storage and distribution are also significant (see
Table 8).
44

Diesel Vehicle Emissions and Urban Air Quality. Second Report of the Quality of Urban Air
Review Group, Department of the Environment, London, 1993.

Table 8

Estimated UK emissions of primary pollutants by UNECE source
category for 1994 (ktonne) 40
NOx*

SO2

VOCb

Black
smoke

21
272
5
6
26

526
69
34
45
48

1759
90
74
138
89

6
37
2
2
4

19
93
3
3
2

40
37
5
5
20

128

422

7
658
411

109
653
442
151
4
3

2
19
44
57
4
9

2218

2718

CO
Power stations
Domestic
Commercial/public service
Refineries
Iron and steel
Other industrial
combustion
Solvent use
Non-combustion processes
Extraction/distillation of
fossil fuelsd
Road transport—petrol
Road transport—diesel
Other transport
Waste treatment and disposal
Agriculture
Forests
Total

47
4117
203
40
48
11
4833

243
545
76
25
21

PM10c

15
232
3
41
1

21 e
45 e
7
1

80f
2117

426

263

NOx is expressed as NO2 equivalent.
Volatile organic compounds excluding methane.
1993 data. Secondary particulate pollutants such as sulfates and nitrates are not accounted for in
this total.
d
Includes offshore contribution from gas flaring and use of gas.
e
Authors' estimates from the road transport total of 66 ktonne.
f
Estimate of natural emissions.
b

VOCs are important in atmospheric chemistry for the formation of
photochemical smog which will be discussed in Section 4. Various
control measures are therefore being implemented so as to enable the
European Member States to meet their agreed obligation under a 1991
UNECE Protocol of reducing VOCs by 30% relative to 1988 by 2000.
Between 1990 and 1994 a 9% reduction in VOC emissions across Europe
was reported.45
3.2.2 Nitrogen Oxides. Although there are small emissions of NO x
from industrial processes such as nitric acid production, the main
emissions are from combustion. In stationary combustion plant, factors
such as the fuel type, fuel nitrogen content, burner design, the intensity of
combustion, the overall shape and size of the furnace, and the amount of
excess air all influence NO formation and can be modified to achieve a
certain measure of control. However this falls considerably short of
eliminating the emissions. Typical flue gas concentrations for industrial
45

http://tiger.eea.eu.int/, 'The European Environment Agency', December 1997.

coal-burning are about 550 ppm but with new designs of IoW-NOx
burners this is reduced by about 50%. Nitrogen dioxide, NO2, forms
only a small fraction of the waste gases from combustion (usually less
than 10%) so the description 'NO x emissions' actually refers mainly to
NO emissions. Once in the atmosphere, oxidation of NO to NO 2 occurs,
as described later, and the relative proportions of the two oxides may
then be comparable.
There is negligible nitrogen in gasoline or diesel fuels so the NO x arises
from the N2 and O2 in the air. Nitrogen oxide emissions are high due to
the high temperatures and pressures and are at a maximum during
acceleration and minimum during idling. Diesel engines have comparable NO x emissions to petrol engines as shown in Tables 6 and 7.
3.2.3 Sulfur Dioxide. Sulfur dioxide, SO2, arises from the sulfur
present in most fuels amounting to 1-2% wt in coal, 2-3% in heavy
fuel oils and decreasing amounts in lighter oil fractions. There are also
non-combustion emissions from sulfuric acid plant and from the roasting of sulfide ores during non-ferrous smelting.
The limit for the sulfur content of diesel fuels has been progressively
reduced and now stands at 0.05% in the USA and much of Europe.
Typical gasoline sulfur content is about half this. During combustion,
conversion to sulfur dioxide is virtually complete although about 5-10%
may be retained by coal ash. For the highest sulfur fuel oils (3% S) the
flue gas concentrations can reach 2000 ppm while for a typical power
station coal of 1.6% S the concentration is about 1200 ppm. No control
over SO2 emissions can be achieved by modification of combustion
conditions, and reductions must be sought by pre-treatment of the fuel
or desulfurisation of the flue gases after combustion.
3.2.4 Paniculate Matter. Dust can arise by the disturbing action of
outdoor industrial activity on the ground or on raw materials. Quarrying, open cast mining, tipping, building activities, the action of heavy
lorries, or simply a strong wind acting on stock piles can lead to grit and
dust blowing beyond the site and causing a nuisance to others in the area.
These are described as 'fugitive' emissions and the particles tend to be
quite coarse (>5 /mi diameter).
During combustion, formation of soot commonly accompanies
carbon monoxide formation and is generally due to inadequate air
supply. Soot particles produced from gas phase fuel/air reactions are
commonly sub-micrometre (< 1 /mi) and because they are of comparable
size to the wavelength of light they are effective both at light absorption
and light scattering. A relatively small mass concentration will therefore
render the exhaust of flue gases opaque and give a dark stain on a filter

paper. This is the basis of the reflectance technique used for measurement
of'smoke'.
In the combustion of coal, volatile material is released and burns as a
gas, the remaining char burns more slowly, and a final residue of ash is
left. Smoke can be produced from incomplete combustion of the volatile
matter. Anthracite, coke, and the other manufactured solid fuels (such as
the formed briquettes widely used for cooking stoves in Asia) contain
low levels of volatile material and so reduce the smoke problem.
Incomplete combustion of volatile matter is also the major problem
with wood smoke and smoke from bonfires or biomass burning. This can
lead to serious problems in developing countries where the use of
biomass is still common for domestic cooking and heating in cities
causing high emissions at low level.
Smoke formation is not a significant problem with spark ignition
engines because the petrol and air are well mixed before entering the
cylinders. Diesel engines suffer from the disadvantage of producing more
smoke, especially under heavy load or acceleration conditions, because
of the relatively poor mixing of air with the fuel injected as a spray into
the cylinder. This produces regions which are too rich in fuel for
complete combustion, leading to soot formation.
Four important terms relating to particulate matter in the atmosphere
are:
Smoke—the particulate material assessed in terms of its blackness or
reflectance when collected on a filter, as opposed to its mass. This is the
historical method of measurement of particulate pollution in the UK.
The size of particles collected into the sampler is below 10-15 /im.
TSP—Total Suspended Particulate matter. Mass concentration determined by filter weighing, usually using a 'Hi-VoI' sampler which
collects all particles up to about 20 fim depending on wind speed.
PMlO—Particulate matter below 10 ^m aerodynamic diameter. This
corresponds to the particles inhalable into the human respiratory
system, and its measurement uses a size selective inlet.
PM2.5—Particulate matter below 2.5 /mi. This is closer to, but slightly
finer than, the definitions of respirable dust that have been used for
many years in industrial hygiene to identify dusts which will penetrate
the lungs.
Particulate vehicle emissions, for example diesel smoke, are predominantly in the PM2.5 range. The total number of particles emitted is
dominated by ultrafine particles in the 30-80 nm (0.03-0.08 /an) range.
However, a high proportion of the total mass is contributed by larger
particles—one 0.5 /im particle has the same mass as 1000 particles of

50 nm. The health significance of these ultrafine particles is an area of
current research.
3.2.5 Emissions Limits. Emissions from industrial processes are controlled via emissions limits set by international bodies such as the
European Union or by national organisations such as the UK Environment Agency or US EPA. Processes subject to control receive an
authorization which includes the permitted levels of emission. For
example, large European municipal incinerators burning more than
3 tonne h " 1 of waste must comply with the following emissions standards in mg m~~3 of dry gas.
Particulates
SO2
HF
Pb + Cr + Cu + Mn
C d + Hg

30
300
2
5
0.2

CO
HCl
Volatile organics
Ni + As

100
50
20 (as total C)
1

3.2.6 Emissions Inventories. The use of spatially disaggregated emissions inventories in modelling has been mentioned above. Knowledge of
particular emission sources directs attention to the best targets for
control and emissions reductions. Table 8 shows the UK national
breakdown of emissions by sector for various pollutants while Table 9
illustrates national emissions for several European countries. These
tables will be referred to in following discussions.

Table 9

Emissions inventory data for European Countries from the CORINAIR
Project (ktonne per annum) 7

Country

Germany (all)
Germany West
Former GDR
France
UK
Italy
Spain
Poland
Former Czechoslovakia

NMVOC

NOx (as NO2)

SO2

1990

1994

1990

1994

1990

1994

3323
2484
839
2866
2682
2549
1894
1295
531

2531
2307
2430
2785
1970
-

2980
2424
556
1590
2773
2053
1257
1446
1010

2266
1682
2387
2157
1227
-

5255
912
4343
1300
3787
2253
2206
3273
2409

2998

1013
2696
3671
2060

3.3

Air Quality

3.3.1 Air Quality Standards. Air quality is assessed by reference to
air quality standards.46 Those applicable in Europe and the UK are
indicated in Table 10. The existing European standards are subject to
revision in the near future following a new Framework Directive on
Air Quality Assessment and Management (Directive 96/62/EC). The
values included in the first 'daughter' directive agreed in June 1998 are
indicated in the Table. Standards proposed by the UK Expert Panel
on Air Quality Standards (EPAQS) are also indicated. These were
promulgated to support the new responsibilities for air quality
management of the Local Authorities following the 1995 Environment
Act and are also used to define objectives within the government's
National Air Quality Strategy.47
3.3.2 Air Quality Monitoring. A critical activity in air quality management is the monitoring of the pollutant concentrations. Continuous
monitoring data for common pollutants such as CO, NO x , O3, SO2, and
particulate matter (TSP or PMlO) are recorded at many sites around the
world. Information from an increasing number of these sites is available
on the World Wide Web.48 Sample data from Leeds in the UK is
illustrated in Figure 9 and shows the diurnal cycles of the main urban
pollutants.
A global monitoring network, the GEMS/AIR programme, was
initiated by the WHO in 1973, and in 1975 became part of the Global
Environment Monitoring System GEMS,49 now jointly managed by
WHO and UNEP. GEMS/AIR was implemented to strengthen urban
air pollution monitoring and assessment capabilities, to improve validity and comparability of data among cities, and to provide global
assessments on levels and trends of urban air pollutants, and their
effects on human health. Approximately 170 monitoring stations in 80
cities in 47 countries participate in the programme measuring SO2 and
suspended particulate matter (55 cities in 33 countries) with other
stations measuring NO 2 , CO, O3, and Pb. An indication of the air
quality in major cities of the world from this programme is given in
Table 11.
46

Quality of Urban Air Review Group, 'Urban Air Quality in the United Kingdom', Department of
the Environment, London, 1993.
47
'The United Kingdom National Air Quality Strategy', Department of the Environment, The
Stationery Office Ltd, London, 1997.
48
U K A i r quality data is available a t http://www.aeat.co/uk/netcen/ A review of other sites is given
in J. Perks, Clean Air, 1997, 27, 2, 44.
49
http://www.gsf.de/UNEP/gemsair.html

Table 10 Air quality standards for European Community and UK
Pollutant

Measure

Concentration*

Source0

Smoke

Daily means—98%
Winter median
Annual median
Running 24 h means—99%
Stage 1:
24 h—3 5 times/yr
Annual mean
Stage 2:
24 h—7 times/yr
Annual mean

250/igm"3
130 /xg m ~ 3
80 /ig m ~ 3
50/xgm~ 3

EC(I)

Daily means—98%

EC(I)

15 min means—99.9%
1 h—24 times/yr
24 h—3 times/yr
1 yr, ecosystem protection

250/xgm" 3 (smoke < 150)
350/xgm~ 3 (smoke< 150)
130 /xg m ~ 3 (smoke > 60)
180/xg m~ 3 (smoke <60)
80 /xg m~ 3 (smoke >40)
180/igm" 3 (smoke <40)
100 ppb
350 /ig m " 3
125 /xg m ~ 3
20 /xg m~ 3

NO 2

Hourly means—98%
Hourly means—18/yr
Annual mean

200/xgm~ 3
200/xgm" 3
40 /xg m ~ 3

EC(2)
EC(5)

NO + NO 2

1 yr, vegetation protection

30 /xg m~ 3 (as NO 2 )

EC(5)

Running 8 h mean

lOppm

EPAQS

PMlO

SO2

Winter median
Annual median

Ozone

Lead
a

A. 8 h mean
B. 24 h mean
1 h mean
C. I h mean
D. 1 h mean
Running 8 h mean—97%
Annual mean
Annual mean

50 /xg m " 3
40 /xg m ~ 3
50 /xg m " 3
20 fig m ~ 3

110//gm~
65/xgm~ 3
200/igm"3
180/xgm~ 3
360/xgm~ 3
50 ppb
3

2/xgm~
0.5/xgm~ 3

EPAQS
EC(5)
EC(5)

EPAQS
EC(5)

EC(3)

EPAQS
EC(4)
EC(5)

Units: 100 ppb corresponds to 266.1 /xg m" 3 for SO2, 191.2 /ig m~3 for NO2, 116.4 /^g m~3 for CO,
and 199.5 /xg m~3 for O3 at a reference temperature of 200C.
b
Sources: EC(I) Directive 80/779/EEC; EC(2) Directive 85/203/EEC; EC(3) Directive 92/72/EEC.
Levels for ozone are reference levels not quality standards: A. Health protection standard,
B. Vegetation protection standards, C. Population information, D. Population warning; EC(4)
Directive 82/884/EEC; EC(5) Directive on limit values for sulfur oxide, oxides of nitrogen,
particulate matter and lead in ambient air, 1998; EPAQS—UK Expert Panel on Air Quality
Standards.

NO 2

PPb

Ozone

hour

ppm

PPb

hour

SO2

ppb

ngm"3

PM10

hour

Figure 9 Diurnal variation of pollutant concentrations for the Leeds A UN site.A% Each
point represents the monthly average for that hour of the day. • January 1997,
M July 1997

Table 11

Air quality data for megacities of the world. Annual average data for
1989 or 19902Uh

City

Population
1990
(W6)

Motor vehicle
registrations
(W6)

SO2
^m'

Tokyo
Mexico City
New York
London
Beijing
Moscow
Delhi

20.5
19.4
15.6
10.6
9.7
9.4
8.6

4.40
2.50
1.78
2.70
0.308
0.665
1.66

20-30
100-200
30-50
30-40
20-120
2(MO

a
b

NO2
fig m~ 3

TSP
/xgm" 3

60
60-80
50-100
40-50

50-60
200-500
50-70
20a
250-400
100
350-500

Smoke level. This will be significantly below the corresponding TSP level.
The data in the table represent the approximate ranges across several monitoring sites in each city
and are to be regarded as indicative.

3.3.3 Air Quality Trends. The most significant trends in air quality on
a national basis relate to a rather small number of factors such as:
•
•
•
•

changes in the patterns of fossil fuel usage
growth in vehicular transport
improved technology for vehicles giving lower emissions
tighter control of industrial emissions, especially those from large
scale combustion for power generation.

There is however a significant difference between industrialized and
developing nations with respect to the major contributors to air quality
problems. A UNEP report has shown that cities in developing countries
seem to be following the same historical trends as in other countries.
Before industrialization, pollution is mainly from domestic sources and
light industry. Concentrations are generally low but increase as the
population increases. As industrial development and energy usage
increase air pollution levels begin to rise dramatically and emission
controls have to be introduced.21 Emissions and air quality trends will
be illustrated in the next few sections by reference to several of the
commoner pollutants. Ozone, as a secondary pollutant, is dealt with
later.
3.3.4 Vehicular Emissions—CO and Hydrocarbons. A major feature
of the development of most countries in the last three decades has been
the growth in vehicular traffic. The number of cars in the UK doubled
from 10 to 20 million between 1970 and 1994, an average growth rate of
3% per annum. But in other countries the increase has been even more

rapid—China, for example, has seen growth of 13% per annum
representing a trebling of the vehicle numbers in 10 years, reaching 9.5
million by 1994.
The importance of vehicular emissions relative to other sources for the
UK is shown in Table 8. Carbon monoxide emissions are predominantly
from road vehicles whilst about 30% of VOCs also come from vehicles.
Historically, CO emissions increased as traffic density increased through
the 1980s but are now on a downward trend arising from the progressive
introduction of catalytic converters (Figure 10). The significance of
vehicular emissions is shown in the diurnal variation of the CO levels in
major cities as illustrated in Figure 9. The double humped curve
corresponding to the morning and evening rush hours is clearly seen.
Carbon monoxide levels beside busy roads can, under very adverse
conditions, reach approximately half the occupational exposure limit of
50 ppm but general urban levels in much of Europe and the US are
usually below 10 ppm, and annual averages reach only 1-2 ppm in the
UK and 5 ppm in the US.50 In major cities where traffic is heavy but
catalytic converters are not common, CO levels can be high; for example
in 1992 Mexico City exceeded WHO guidelines for CO by a factor of
more than 2.21
3.3.5 Nitrogen Oxides. Emissions of nitrogen oxides in the UK
gradually increased until the early 1990s due to increased traffic density
(Figure 10) and are only now coming down slowly as new vehicular
emissions regulations begin to take effect and the NO x emissions from
power stations also decrease. Countries which have high coal consumption and/or high vehicle populations tend to have high NO x emissions.
Figure 11 and Table 9 illustrate this for several European countries using
data from the CORINAIR inventory.7 Allowing for the populations, the
per capita emissions do not vary to the same extent as the total national
emissions. At the moment it is considered unlikely by the European
Environment Agency that targets for a 30% reduction by 2000 across the
EU will be met.45
Since nitrogen dioxide is the main health hazard, most attention has
been paid to monitoring urban levels of NO2 as opposed to total NOx.
Annual average concentrations in urban areas of the UK range from 25
to 45 ppb. A significant wintertime episode of NO 2 pollution was
experienced in London in December 1991 when the levels exceeded
300 ppb for 8 hours over one night and exceeded 100 ppb continuously
for 3 days. The current EU standard for hourly averages is 104.6 ppb
(200 fig m~~3). Throughout Europe there are a number of cities with long
50

http://www.epa.gov/oar/aqtrnd96/toc.html 'National Air Quality and Emissions Trends Report
1996'.

NOx

Transport

SO2

Domestic

Public Power

Other

Industry

Transport

Domestic

Figure 10

Public Power
Transport
Other
Domestic
Industry
UK annual emissions ofpollutants by category of emissions source**

Public Power

Industry

Other

Black smoke

Transport

Domestic

Public Power

Other

Industry

Total National Emissions and
per capita emissionsxiO
Figure 11 National NOx emissions (ktonne per annum) for various European countries
from the CORINAIR inventory7 for 1990 and the corresponding emissions per
capita

term average NO2 levels above the proposed EU guideline of 40 ^g m~ 3 .
Maximum concentrations do not show exceedences so often, however,
and NO2 episodes are not seen to cause a threat to those living outside
urban areas.45 Figure 9 shows that NO shows a similar diurnal variation
to CO because of the traffic contribution but that NO2, which can be
considered as a secondary pollutant, does not show as marked a trend.
This is discussed later.
3.3.6 Sulfur Oxides. Following the outcry over the London smog of
1952 and the Clean Air Act of 1956, a concerted effort was made in the
UK to reduce levels of smoke and sulfur dioxide in the air, which arose
mainly from coal combustion. The components of this effort were
control of visible smoke emissions from industrial chimneys, control of
chimney heights to ensure adequate dispersion of SO2, and control of
domestic smoke emissions by local authorities through the promotion of
the use of alternative fuels and the introduction of'smoke control zones'.
In practice, because of the widespread availability of natural gas from
the North Sea in the years following 1970, gas has captured most of the
domestic heating market in the UK and shares the commercial market
with oil. Domestic coal use is limited and the major users are now the
electricity supply industry and steel making. During the 1990s even this
pattern changed with the generating companies switching to gas for new
power stations for both economic and environmental reasons.
The urban concentrations of smoke declined dramatically from 1960
to 1980 but for sulfur dioxide the overall upward trend in emissions was

not reversed until much later—1970 was the peak year. This pattern is
mirrored in some but not all of the other European countries. Exceedences of the EU air quality standards for smoke and SO2 are now very
rare—none of the UK monitoring sites showed exceedences of the daily
standards in the period 1993-95. One of the cities with a remaining
smoke and SO2 problem is Belfast in Northern Ireland, where coal
burning in homes is still common since natural gas is only just becoming
available. Many cities in Eastern Europe where coal is still in heavy use,
have serious problems reminiscent of England in the 1960s. Concentrations reach 100-150 /ig m~ 3 in some areas of central and north-western
Europe during episodic 'winter smog' situations.45
A comparison of sulfur dioxide emissions across the countries of
Europe shows wide variations depending on the fuel mix in each country.
Table 9 shows data for 1990 and 1994 from the CORINAIR emissions
inventory. France has relatively low emissions owing to the dominance
of nuclear power for electricity generation. Many of the Eastern
European countries have high emissions due to dependence on relatively
poor quality coal. The former German Democratic Republic had
especially high emissions (over 80% of the 1990 figure for 'Germany'
shown in the table). These have been markedly reduced since reunification as shown by the 1994 figure.
3.3.7 Vehicular Particulates?1 Black smoke emissions were originally
dominated by coal smoke and as control efforts were targeted on this
source rapid improvement of air quality was achieved in many cities.
However, the more recent decline has not been as rapid because of the
rise of smoke derived from vehicles, particularly diesels. Diesel engine
soot is three times blacker than coal smoke and this is incorporated into
the estimation of the 'black smoke' emissions shown in Table 8. In
contrast, the small amount of smoke from a petrol engine is much lighter
than coal smoke because it consists largely of unburnt hydrocarbons
rather than elemental carbon. Figure 10 and Table 8 show how UK
smoke emissions, once dominated by coal, are now dominated by
vehicular emissions.
In terms of the mass of fine PMlO particles the picture is not so onesided (Table 8) and both petrol and diesel engines make a significant
contribution. Although for a given mass diesel smoke is much blacker
than petrol-derived smoke and gives a much greater influence on the
reflectance (and hence the 'smoke' measurement) this is irrelevant if a
direct mass measurement technique is used. Analysis of UK PMlO
measurements suggest that on average 40-50% of urban PMlO is
51

'Airborne particulate matter in the United Kingdom', Third Report of the Quality of Urban Air
Review Group, Department of the Environment, available from The University of Birmingham,
Institute of Public and Environmental Health, 1996.

contributed by vehicle exhausts in winter, with higher proportions when
the PMlO concentration itself is high and lower proportions in
summer.48 In addition to the primary emissions indicated in Table 8
account must also be taken of the secondary particulate pollutants such
as sulfates and nitrates. These are discussed in Section 5.1.
On a global scale, particulate matter represents the biggest problem
for the urban environment in terms of non-attainment of air quality
standards. Many of the world's largest cities persistently exceed WHO
limits, exposing large populations to a health risk. Although PMlO levels
are falling in developed areas such as the US and Europe, it is currently
estimated that in nearly three-quarters of major EU cities WHO Air
Quality Guidelines for SO2 and particulate matter (PM) are exceeded at
least once in a typical year.45
3.3.8 Heavy Metals. Emissions of lead, Pb, from automobile exhausts
arise from the use of lead tetra-alkyl anti-knock additives to improve the
octane rating of the petrol. About 70% of the lead is emitted from the tail
pipe, mainly as mixed halides. Increasing concern over the potentially
harmful effects of lead on health, particularly the health of children,
resulted in a gradual decrease in the maximum permitted amount which
can be added to petrol in many countries. This is currently 0.15 g Pb I" 1
in Europe. All new cars must now be able to run on unleaded gasoline.
The emissions of lead in 1995 were about 14% of what they were in 1975
and 70% of total gasoline sales in the UK are now of unleaded petrol.
The decrease in the use of lead has led to a marked reduction in the
atmospheric concentrations in many countries, with Europe following
some years behind the USA. For UK urban sites, long term averages
have come down from a typical 500 ngm" 3 to around 100 ng m~ 3 in
1998, with similar decreasing trends being mirrored in the US where
averages were around 50ngm~ 3 in 1995. These figures may be compared with the current EU air quality standard of 2 or the new standard
of 0.5 fig m~ 3 annual average. Although declining in Europe, airborne
lead remains an important air pollutant in cities where leaded petrol is
still a major transport fuel. UNEP reported that several of the world's
megacities still exceed WHO limits for lead, with Cairo and Karachi
showing the highest exceedences in 1992.21
3.3.9 Toxic Organic Micropollutants (TOMPS). This phrase is used
to describe several groups of organic pollutants which are present in the
atmosphere in very small concentrations but which, by inhalation or
ingestion of contaminated food products, could result in human health
effects. TOMPS may be present in the vapour phase or adsorbed onto
particles, the distribution between vapour and particulate phases
depending on temperature. They include polynuclear aromatic com-

pounds (described variously as PNA, PAH or PAC)—high molecular
weight hydrocarbons often found in association with soot, some of
which are carcinogenic. Benzo[#]pyrene is one example. Following the
US EPA, analyses are commonly made on a group of 16 compounds
with 2-6 rings.
There has been a long-standing interest in the presence of such
carcinogens in the general environment and in the levels emitted from
combustion processes. Attention has been switched from coal smoke to
diesel engine exhaust smoke as a source of these compounds but it
remains to be established whether there is a significant health hazard at
the very low concentrations to which the general public are exposed.
The other main group of TOMPS are polychlorinated aromatic
compounds. These include polychlorinated biphenyls (PCBs) which
have entered the environment following wide scale use as insulating
fluids in electrical transformers, and polychlorinated dibenzodioxins and
dibenzofurans ('dioxins', PCDDs; 'furans', PCDFs), which have combustion origins, especially from incinerators. There are numerous
isomers of each of these classes of compounds depending on the
number and locations of the chlorine atoms in the molecules.
Extensive monitoring of air, water and soil samples for TOMPS has
been undertaken in recent years. The average air concentrations in
London between 1991 and 1995, for example were as follows:40
PAHs 50-150 ng m~3; PCBs 1.0-1.5 ng m"3; Dioxins 100-200 fg TEQ m"3.
The US EPA carried out a thorough review of dioxins in the environment and their health effects, published in 1994.52 Airborne concentration data reported in the USA and Europe also indicated average air
concentrations of the order of 100 fg TEQ m~3.
Note: Ing = 10~9g, 1 fg = 10~15g. TEQ = Toxic Equivalent—in
summing the masses of different dioxins each is weighted by its relative
toxicity with 2,3,7,8 tetrachlorodibenzo-p-dioxin taken as the reference.
4 GAS PHASE REACTIONS AND PHOTOCHEMICAL OZONE
4.1 Gas Phase Chemistry in the Troposphere
4.1.1 Atmospheric Photochemistry and Oxidation. Although the emissions patterns and dispersion discussed in the previous sections give a
part of the picture of where high levels of pollutants can be found, we
52

'Estimating Exposure to Dioxin-like Compounds' (3 VoIs) and 'Health Assessment Document for
2,3,7,8-Tetrachlorodibenzo-/?-dioxin (TCDD) and Related Compounds' (3 VoIs), US Environmental Protection Agency, Office of Research and Development, Washington, 1994.

Mixing Height

Droplet
Reactions'
Raanout

wind

Dispersion

Emissions

Chemical
Transformation
(Gas phase )

Dry Deposition

Washout

Wet Deposition

Figure 12 Processes which may be involved between the emission of an air pollutant and its
ultimate deposition to the ground

must combine this with an understanding of other atmospheric processes
to see how secondary pollutants such as ozone and acid rain are formed.
These processes including dispersion, chemical reaction, and deposition
are illustrated in Figure 12. The chemical processes involved are complex
and changes both in the gas phase and in the aqueous droplet phase are
important. Not only do we have to consider the transformation of the
primary pollutants but also the formation of secondary pollutants such
as ozone which can have adverse effects on the environment and on
human health.
Of basic importance to an understanding of the gas phase chemistry is
the effect of the sun. Photons of ultraviolet light provide a means of
initiating chemical reactions which would otherwise not take place. In
addition to stable molecules, photochemical reactions involve free
radicals such as hydroxyl OH*, hydroperoxy HO*2 and methyl CH 3.
Free radicals are extremely reactive and have very short lifetimes. Their
concentrations in the atmosphere are small but none the less significant.
For example, OH* concentrations in polluted atmospheres may be in the
range 106-107 radicals cm" 3 , i.e. one radical for every 1013 nitrogen
molecules.
One of the most important overall processes we have to describe is
oxidation, that is, the combination of atmospheric oxygen with the
primary pollutants. For the three commonest inorganic pollutants the
overall results are
carbon monoxide
nitrogen oxides
sulphur dioxide

carbon dioxide
nitric acid
sulphuric acid.

Later we shall consider the involvement of the two acids in particles and
droplets but as far as the gas phase chemistry is concerned these species
mark the end point of the process. For organic hydrocarbon species there
may be a number of intermediate stable molecules formed, but the
overall process is rather like combustion with the end product being
carbon monoxide; for example:
methane CH4 -> formaldehyde HCHO -> carbon monoxide
although the time taken to complete this process can be very long. In all
these cases the main species initiating the sequence of reactions is the
hydroxyl radical:
(14)
(15)
(16)
(17)
In the case of nitric acid formation, there are no remaining free radicals
to continue the chain of reactions. In the other cases the hydroxyl radical
is eventually regenerated but only after several further steps which
interlink the chemistries of the various species (Table 12). Carbon
monoxide oxidation is a slow process and the lifetime of CO in the
atmosphere is several years. The oxidation rate of SO2 can be around
1 % h~ l resulting in an overall lifetime of a few days. Radicals other than
Table 12 Chemical reactions for the atmospheric oxidation of CO, SO2, CH4
Carbon monoxide
(M = an inert third body)
Sulfur dioxide

Methane

HO"2 to OH* conversion

OH* such as HO 2 , CH 3 O 2 , or other hydrocarbon peroxy radicals will
also attack SO2 but at a slower rate and their contribution to the overall
oxidation is thought to be relatively small.
4.1.2 Ozone. We must now address the question of where the hydroxyl radicals come from in the first place. One source, present even in
non-polluted atmospheres at a background level of 20-40 ppb, is ozone.
Ozone is a secondary pollutant formed through photochemical reactions
and can have a harmful effect on human health causing respiratory
problems, and on crop yields. It is the primary constituent of photochemical smog. Ultraviolet light of wavelength below 310 nm can
dissociate ozone producing electronically excited oxygen atoms O*
which rapidly split molecules of water vapour:
(18)
(19)
Aldehydes (R.CHO, including formaldehyde H.CHO) can also be
photolysed producing hydrogen atoms which eventually result in OH*
radicals via reactions already shown in Table 12.
Of basic importance to the understanding of polluted urban atmospheres is the photolysis of NO 2 and the subsequent formation of ozone
above background levels. It was noted earlier that only a small proportion of NO x emissions are in the form of NO 2 , the rest being NO. NO
emitted into the atmosphere can be slowly oxidized to NO 2 by the
reaction with molecular oxygen
(20)
Photolysis of NO 2 by UV light below 395 nm produces oxygen atoms
and subsequently ozone:
(21)
(22)
However the process is reversed by the reaction
(23)
so the net result is an ozone level in equilibrium with NO and NO 2 and
dependent on the intensity of the solar radiation. The observed levels of
ozone are higher than would be predicted on the basis of this limited

Table 13

Species lifetimes, kOH and POCP weighted emissions for selected
hydrocarbon species. POCP values are calculated relative to a POCP
of If or ethene based on 1990 data53

Species

koH x JO12
cm3 molecule" 1 S " 1

Lifetime
(h)

POCP weighted emission
(tonnes ethene yr~ l )

methyl benzene (toluene)
ethene
1,3-dimethyl benzene
butane
propene
benzene
ethane

5.96
8.52
23.6
2.54
26.3
1.23
0.268

58.3
40.7
14.7
136.7
13.2
282.3
1295

85423
78925
66086
58649
36946
9418
3017

scheme. High ozone levels imply a high NO2/NO ratio or significant
NO -> NO2 conversion which cannot be achieved by the molecular
reaction (20). The types of reactions responsible have already been shown
in Table 12; they are the transfer of an O-atom from HO 2 , CH3CT2 and
other peroxy radicals. Crudely we can say that the photochemical
reaction mixture catalyses the NO to NO2 conversion resulting in the
build-up of ozone. Hydrocarbon molecules differ in their photochemical
ozone creation potential (POCP),53"55 largely related to how quickly they
react with the OH* radical. Methane has a very low potential but other
species, including substituted benzenes such as toluene and the xylene
isomers, and light unsaturated hydrocarbons such as ethene, propene,
and but-2-ene have high POCPs. POCPs represent the increase in ozone
concentration caused by an increase in the concentration of the hydrocarbon species relative to ethene which has a value 1. The aromatics are
present in high concentration in gasoline and the unsaturated compounds are typical products of engine combustion. Table 13 gives some
examples of the rate of reaction with OH* for selected hydrocarbon
species, along with associated POCP weighted emissions. These are
calculated by multiplying the POCP by the annual emission and therefore indicate which VOCs are best targeted for control strategies.
As they are driven by the chemistry described above, we can see that
the concentrations OfO3 and NO2 depend on sunlight and emissions and
will therefore vary diurnally as well as with the time of year. The diurnal
variations of NO, NO2, and O3 typically detected are illustrated in
53

UK Photo-chemical Oxidants Review Group, 'Ozone in the United Kingdom', Department of the
Environment, London, 1993.
54
R. G. Derwent and M. E. Jenkin, 'Hydrocarbon involvement in photochemical ozone formation
in Europe', AERE-R13736, UK Atomic Energy Authority, Harwell.
55
UK Photo-chemical Oxidants Review Group, Fourth Report, 'Ozone in the United Kingdom',
Department of the Environment, Transport and the Regions, London, 1997.

Figure 9. The morning rush hour peak in the NO and hydrocarbon
emissions is followed by the gradual conversion to NO2 and subsequent
rise ofO 3 , which decays as the sun goes down in late afternoon.
The production of ozone is much greater in the trace representing
summer conditions than for winter due to higher photolysis rates. For
NO and NO2 the winter hourly averages are higher. In an air mass
moving downwind from a city, the ozone peak may be worse in the
surrounding countryside than in the city centre because of less destruction ofO 3 by NO. During the night most of the reactions that have been
described die down but there is an additional route for conversion of
NO2 to nitric acid via the nitrate radical NO 3 which is formed from
reaction OfNO2 with O3. NO 3 is photolytically unstable in daylight. Dry
deposition is also an important loss process for O3 and can account for
up to 30% of its loss.
Space does not permit a detailed discussion of the hydrocarbon
chemistry in the atmosphere, which is extremely complex. In addition to
reactions with OH* radicals and molecular oxygen similar to those for
methane shown in Table 12, hydrocarbon species are attacked by the
oxygen atoms released in reaction (21) and by ozone. One result of the
interaction of the hydrocarbon and NO x chemistries is the formation of a
group of lachrymatory substances including peroxyacetyl nitrate (PAN)
and peroxybenzoyl nitrate (PBzN). The thresholds for eye irritation for
these compounds are only 700 and 5 ppb respectively. They are formed
through the reaction of NO2 with oxidation products of aldehydes:

R = CH3 gives peroxy acetyl nitrate; R = C6H5 gives peroxy benzoyl
nitrate.
4.2

Trends in Ozone Levels

In the UK, elevated ozone levels generally occur in summer, anticyclonic
conditions such as those illustrated in Plate 3 when photolysis rates are
high and wind speeds low. Concentrations tend to be higher in southern
England where polluted air masses from continental Europe have a
significant effect. The highest recorded concentration in the UK was
258 ppb at Harwell in 1976 as compared with a high of 450 ppb in the
Los Angeles basin in 1979. Los Angeles provides an ideal environment
for ozone formation due to high traffic emissions and its meteorological
conditions. Ozone levels in LA are currently falling due to emissions
legislation, although the number of days a year on which the State limit
of 90 ppb is exceeded is still of the order of 100.

In Europe there is evidence that ozone is slightly increasing throughout the troposphere. Over 1000 monitoring stations are now reporting
ozone levels in Europe. In the summer of 1996 the EEC threshold for
warning the public of 1 h > 360 jig m~ 3 was exceeded at three stations
in Europe, two in Athens and one in Firenze in Italy. The threshold for
information of the public of 1 h > 180 /ig m~ 3 was exceeded in all EU
states except for Ireland. In terms of episodes of elevated ozone levels,
the UK and Scandinavia are the least affected out of all the European
countries with southern Europe being most affected. Urban areas show
less exceedence because of high NO conditions.
The formation of photochemical smog is governed by emissions of
NOx and volatile organic carbons (VOCs). Because the chemistry is
complex and non-linear, abatement strategies for ozone and smog are
difficult to devise. Within Europe some regions will respond better to
reductions in NO x emissions and some better to a reduction in VOCs. In
general whether a region is NO x or VOC limited depends on the
NO x : VOC ratio, with regions of high NO x often responding better to
reductions in VOCs. The UNECE Protocol now require a 30% reduction in VOCs from 1988 to 1999 and an EC directive on ozone requires
national monitoring networks to assess conditions according to the
reference levels given in Table 10.

5 PARTICLES AND ACID DEPOSITION
5.1 Particle Formation and Properties
5.1.1 Particle Formation. The nitric and sulfuric acids formed in the
gas reactions described earlier generally undergo further changes. They
are both water soluble and will be rapidly absorbed into water droplets if
these are present. They may react with solid particles forming sulfates
and nitrates. For example, limestone particles (calcium carbonate,
CaCO3) can be converted to calcium sulfate, and salt particles (NaCl),
of marine origin, can be converted to sodium sulfate, Na2SO4, or sodium
nitrate, NaNO3, with the displacement of hydrogen chloride gas, HCl.
However, the most common reactions are those involving ammonia:
(24)
(25)
(26)
(27)

Reactions (24) and (25) are actually reversible and the position of
equilibrium depends on the concentrations and the temperature. Significant dissociation of NH4Cl and NH4NO3 occurs during warm
summer weather. One end product of the oxidation of SO2 in the
atmosphere is ammonium sulfate, a substance better known as a
fertilizer. The average total mass of these secondary sulfates and nitrates
in the UK varies from about 10-12 jugm" 3 in the south east to 67jugm~~3 in Scotland56 and, being fine particles, they contribute to
PMlO. These figures should be compared to the typical PMlO levels in
the atmosphere discussed earlier.
Particles formed by gaseous reactions or condensation are initially
very small (< 0.1 /im) but grow rapidly either by surface accumulation of
material from the gas or by particle-particle coagulation. Once in the size
range 0.1 to 2.0 /mi, they become relatively stable towards further
growth and can remain airborne for periods of days. Most smoke
emissions are in this category along with the sulfates and nitrates. At
the other end of the size spectrum (2-50 /mi) is coarse dust, either emitted
from industrial processes or raised by the wind from the ground, which
has a shorter lifetime.
5.1.2 Particle Composition. Figure 13 presents a typical breakdown
of the components of the total suspended particulate matter for an urban
area based on results from Brisbane, Australia.57 The fine particles
(<2.5 jum) are dominated by the ammonium sulfate and nitrates plus
carbonaceous material. About one-third of this is elemental carbon and
the other two-thirds organic carbon {i.e. high molecular weight hydrocarbons). The coarser particles are dominated by wind-blown dusts
(crustal materials such as clays, silica, limestone, etc.) and include the
sea salt component but have smaller amounts of ammonium sulfate and
carbon. These results are fairly similar to those reported in the UK51
although the sea salt component is higher.
5.7.5 Deliquescent Behaviour. The particles of water soluble compounds such as sulfates, nitrates, and chlorides will exist in the atmosphere either as solid particles or as droplets depending on the relative
humidity. For pure compounds the transition from solid to liquid is a
sharp one. For example, it takes place with salt, NaCl, at 75% relative
humidity and with ammonium sulfate at 86%. Particles of mixed
composition may continuously grow in size by water absorption from
60-70% up to 100% relative humidity. Insoluble carbonaceous particles are, of course, not subject to this phenomenon. Particles are
56
57

A. G. Clarke, M. J. Willison, and E. M. Zeki, Atmos. Environ., 1985, 19, 1081.
T. C. Chan, R. W. Simpson, G. H. McTainsh and P. D. Vowles, Atmos. Environ., 1997, 31, 3773.

Percent

Fine
Coarse

Crustal
Figure 13

Sea salt

Elemental
Carbon

Organic
matter

Ammonium
sulphate

Nitrate

Composition of coarse ( > 2.5 fim) and fine atmospheric particles in Brisbane,
Australia51

important in cloud and fog formation since they act as condensation
nuclei on which the much larger droplets may begin to be formed from
water vapour.
5.1.4 Optical Properties.5* Fine particles in the 0.1-2/mi size range
are effective at scattering light, and some, especially soot, are also
effective at absorbing light. These effects contribute to a reduction in
visibility through the atmosphere. Quantitatively, the visibility is the
maximum distance at which a large dark object can be seen against the
horizon sky (sometimes called the visual range). In clean air the
distance can be over 50 km but in air polluted by particles this can be
severely reduced. A mass concentration (TSP) of 200-300 fig m ~ 3 will
reduce the visibility to below 5 km. The phenomenon of a mid-summer
haze on a hot day reflects high particulate pollutant levels. It is a
totally different phenomenon to an early morning mist which is caused
by water droplets at a relative humidity near 100% in Europe. Such
highly polluted summer days are usually associated anticyclonic conditions (cf. Plate 3). Wintertime visibility has been improved in many
cities and the frequency of fogs reduced in urban areas by the measures
taken to control smoke and SO 2 emissions. Any additional measures to
reduce pollutant emissions, especially of NO x and SO 2 , will bring a
corresponding benefit in terms of improved visibility throughout the
58

A. P. Waggoner, R. E. Weiss, et al., Atmos. Environ.,1981, 15, 1891.

year. This trend has been confirmed by measurements taken throughout the United States where visibility in the East worsened between
1970 and 1980 but slightly improved by 1990 in line with SO2
emissions.50

5.2 Droplets and Aqueous Phase Chemistry
Liquid water occurs in the atmosphere as clouds, mists, and fog within
which the concentration of water can be up to 1 g m~ 3 . Smaller amounts
of water are also present in association with deliquescent particles as
discussed above. At relative humidities below 95%, this secondary
amount will normally be less than 1 mg m~~3 and will not be considered
further.
Water droplets can accumulate pollutants by adsorption of either
gases or particles and within the droplets chemical reactions can proceed,
changing the nature of the adsorbed species. Solution of SO2 into water
results in a mixture of the species SO 3 " (sulfite), HSO^" (bisulfite) and
H2SO3 (undissociated sulfurous acid) depending on the pH which is
defined as — logio (H + ion concentration). At typical cloudwater pH
values the bisulfite ion is the dominant ion formed from:
(28)
There are several mechanisms for the oxidation of bisulfite to sulfuric
acid in the aqueous phase. In a cloud the most important oxidants
appear to be ozone and hydrogen peroxide, which is formed from the
reaction of two HO 2 radicals. The oxidants must first be adsorbed into
water from the gas phase and then take part in the reactions:
(29)
(30)
The sulfuric acid formed will be completely dissociated to ions H + and
SO4~. The ozone reaction is inhibited by low pH whereas the H 2 O 2
reaction is not. In acidic droplets the oxidation by H 2 O 2 is therefore
dominant whereas at high pH the O 3 becomes more significant. In a
cloud or fog where there has been a significant input of particulate
pollution, the oxidation of SO2 by atmospheric oxygen catalysed by
metal ions (iron, Fe 3 + , or manganese, Mn 2 + ) or by soot can also be
important. Although not indicated by the simple reactions 29 and 30,
free radical reactions are as important in solution chemistry as they are in

the gas phase, and many different species are involved. The results of a
recent European Project in this area have been published.59
It is difficult to distinguish experimentally between the photochemical
reaction mechanism leading to H2SO4 with subsequent absorption of the
acid into water and the aqueous phase oxidation of SO2. However, it
appears that both routes are important.
In the case of nitrogen oxides the routes for nitrate formation in clouds
are
• dissolution of gaseous nitric acid vapour formed by reactions
previously described in Section 4
• dissolution of nitrate-containing particles into droplets
• absorption of nitrogen oxides or nitrous acid, HONO, into droplets
followed by oxidation of nitrite ions NO2T by oxidants such as
H2O2.
Because of the low solubility of NO2 and especially NO in water it is
likely that the dominant processes are the first two.
Despite partial neutralization of dissolved acids by ammonia, the
water in polluted fogs and clouds can be much more acidic than in
collected rainfall. pH values down to 2 have been measured in urban
fogs—more acid than vinegar. Similarly low values have been measured
in the plumes of large power stations.

5.3

Deposition Mechanisms

5.3.1 Dry Deposition of Gases. As illustrated in Figure 12, the life
cycle of an air pollutant normally involves emission, dispersion and
transport, chemical transformation and finally deposition to the ground.
Understanding the rates and mechanisms of deposition is important to
the assessment of the environmental impact of many pollutants. Experimentally we can measure the concentration of the pollutant (jugm"3)
and the total rate of deposition (pig m~2 s~l). The higher the ground
level concentration, the more rapid the deposition but the ratio of these
two quantities gives a useful measure of the efficiency of the deposition
process. It is called the deposition velocity.
^

, .

deposition rate , .

Deposition velocity = —
: — (units: m s
air concentration
59

)

'Heterogeneous and Liquid-Phase Processes', ed. P. Warneck, Springer, Berlin, 1996.

Different surfaces (water, soil, ice, etc.) will have correspondingly
different deposition velocities. The description 'dry deposition' is used
in all cases even if the removal is to a wet surface.
Deposition to vegetation is rather more complex than deposition to a
plane surface. The pollutant's progress may be retarded either by the
slowness of diffusion through the air within the canopy of vegetation (i.e.
between the leaves, etc.) or by the rate of transfer from air to leaf surface
(the cuticle) or to the interior of the leaf via stomata. Moisture makes a
considerable difference, in that transfer to a wet surface is generally faster
than to a dry surface. Based on experimental data, dry deposition
velocities for SO2 used in modelling large scale deposition are typically
assumed to be in the range 2-5 mm s~l. Similar considerations apply to
other pollutants which are subject to a significant rate of adsorption at
the ground or onto vegetation (NO, NO2, HNO3, etc.). NO2 has a lower
value of 1 mm s" 1 while HNO3 deposits very rapidly and values up to
40 mm s~l are assumed.42
5.3.2 Wet Deposition. The term 'wet deposition' is used to describe
pollutants brought to ground either by rainfall or by snow. This
mechanism can be further subdivided depending on the point at which
the pollutant was absorbed into the water droplets. In-cloud absorption
followed by precipitation is termed 'rain-out'; below cloud absorption,
i.e. pollutants collected as the raindrops fall, is termed 'wash-out'. The
rate of removal of a pollutant by wash-out will increase in proportion to
the rainfall rate. Overall we can define a scavenging coefficient which is
the fractional loss of the pollutant from the gas phase per second. For
SO2 the scavenging ratio is of the order of 10" 6 S" 1 in drizzle
(<1 mm h" 1 ) and an order of magnitude greater in heavy rain. Half of
the gas below the clouds can be removed in several hours of heavy rain.
The rates for the nitrogen oxides are lower due to their reduced solubility
in water. In remote areas the majority of the wet deposition of sulfur
appears to be due to rain-out. Wash-out becomes relatively more
significant near the sources of pollution where the gas concentrations
are high.
Another mechanism of deposition is when fog or cloud droplets are
removed directly to the ground or to vegetation. This is termed 'occult
deposition'. It becomes significant at elevated locations such as mountains or hill tops. There is the potential for more severe damage to foliage
than with acid rain since, as was previously mentioned, polluted fog or
cloud droplets can contain much higher concentrations of acidic pollutants than raindrops.
Dry deposition and wet deposition are both important in the total
deposition of sulfur and oxidized nitrogen compounds. Dry deposition is

SULPHUR

OXIDISED NITROGEN (HTN)

(kT S)

EXPORT

INPUT

Emission
Deposition

1600
350

EXPORT

INPUT

Emission
Deposition

780
150

Figure 14 Average annual emission and deposition of sulfur and oxidized nitrogen for the
UK1992-1994. The figures for the inputs are estimates of the European
contribution to the total deposition figures*2

most significant where ground level concentrations are high, in other
words close to the sources. Figure 14 shows the annual budgets for the
UK between 1992 and 1994.42 Deposition of reduced nitrogen (NH 3 ) is
also important as discussed below. For sulfur, dry and wet deposition are
respectively 40 and 60% of the total, for nitrogen the split is 27% dry and
73% wet. Modelling work suggests that about 45% of the wet S
deposition and 25% of the dry S deposition in the UK comes from
other European countries. The UK is however, a net exporter of air
pollutants, while other countries of Europe, for example Norway and
Sweden, are net importers. The UK is the largest single contributor to
sulfur deposition in southern Norway although emissions from many
other countries are transported there.
5 J J Deposition of Particles. The mechanism of deposition of particles depends on the particle size. Large particles with diameter greater
than 10 jum fall slowly by gravitational settlement. The larger the
particles, the more rapidly they fall. The sedimentation velocities for
particles of density 2 g c m " 3 are as follows:
Diameter (//m)
5
10
20

Velocity (m s ~ l )
1.5 x 10~ 3
6.1 x 10" 3
2.4 x 10~ 2

Diameter (^m)
50
100
150

Velocity (m s ~ l )
1.4 x 10" 1
4.6 x 10" 1
8.0 x 10" 1

Particles larger than 150 jam diameter, falling at over 1 m s" 1 , remain
airborne for such a short time that they do not need to concern us as air
pollutants. Particles less than 5 ^m have sedimentation velocities which

are so low that their movement is determined by the natural turbulence
of the air, just as for gases.
Intermediate particles, between 1 and 10 /mi diameter, can be removed
by impaction onto leaves and other obstacles. Particles in the 0.1 to 1 jum
range, which include most of the nitrates and sulfates, are only removed
very slowly by dry deposition. The deposition velocities are of the order
of 1 mm s" 1 , much lower than for SO2. The most likely route for their
removal is rain-out following water vapour condensation and droplet
growth in clouds. Wash-out is not very efficient for these fine particles
although it becomes more significant for larger particles such as coarse
dust.

5.4

Acid Rain

5.4.1 Rainwater Composition. Even in the absence of air pollutants,
rain-water is slightly acidic (pH 5.6) due to atmospheric carbon dioxide.
'Acid rain' therefore refers to rain with a pH below about 5. Annual
average acidities over much of central and eastern Europe correspond to
pH values between 4.2 and 4.8 whilst rain in the Mediterranean region
and the western extremes of Europe has a pH of 5.0 or above. Similarly
the pH in the north-eastern USA is in the region 4.3-4.7 but increases to
5.0 or above in the mid-west and western states. Table 14 shows the
annually averaged composition at three UK sites.40 Barcombe Mills is in
Table 14

Precipitation-weighted mean concentrations (fieq l~*) and mean
annual rainfall for three UK sites, 1992-9442

Site

Barcombe Mills
(SE England)

Thorganby
(NE England)

Eskdalemuir
(S Scotland)

Rainfall mm
PH
H+
Cations
NH 4 +
Na +
Me 2 +
Ca^ +
Cations
Total

733
4.72
19
26
125
34
24
228

494
4.16
69
68
51
17
40
245

1523
4.68
21
19
65
18
10
133

Anions a

52(37)
25
146 ( < 0 )
223

86 (79)
42
108 (48)
236

Anions
a

SO41NO 3 Cl "
Total

36 (29)
18
74 ( < 0 )
128

Note: Figures in brackets are non-seasalt sulphate and chloride based on a standard ionic
composition of seawater and the assumption that all Na + is sea-salt derived. (<0) indicates that
subtraction of the seasalt component leaves a negative number.

the south-east of England, Thorganby in the north-east and Eskdalemuir
in southern Scotland. Thorganby is the UK monitoring site with the
highest annual acidity of precipitation (pH 4.16). It is in a region close to
several major coal-fired power stations and is unusual in having a very
high non-sea-salt chloride arising from near-source HCl wet deposition.
Excluding this site, sodium and chlorine in rain are predominantly
provided by sea-salt. The acid has been partially neutralized by
ammonia and other ions such as Ca 2+ which may have originated as
calcium carbonate, CaCO3. The sulfate contribution to the acidity is
larger than that of nitrate but the last decade has seen a progressive
increase in the relative importance of nitrate due to the greater reductions in SO2 emissions than NOx emissions. This trend is expected to
continue.
5.4.2 The Effects. The phrase 'acid rain' has come to be used very
loosely to mean almost anything to do with acidification, whether or
not rain is actually involved. Three particular effects have received most
attention. Historically the first of these was the increased acidification
of lakes and streams in Scandinavia leading to loss of fish and other
aquatic organisms. This was attributed to the acidity of rain polluted by
sulfur and nitrogen oxides and the UK was blamed as being the chief
culprit. Similarly the north-eastern USA was blamed for acidification in
Canada. Acidification of fresh waters in the UK itself was demonstrated later.60
The second type of environmental damage is damage to forests.61'62
The effects became very noticeable in Germany in the early 1980s with
the worst effects being noted in the south-west (the Black Forest) and on
the eastern border with the Czech Republic, a country which shares the
same problem. Since then other countries have reported similar phenomena. Although the reasons for the damage have been the subject of much
debate, it seems likely to be a combination of factors. Predisposing
factors include drought and high altitude, and in some cases disease plays
a role. Some forests in the former Eastern European countries suffer
from the effects of very high SO2 levels but this is not the case elsewhere.
Possible mechanisms include:
• the effect of ozone initiating an attack on cell walls, with subsequent
further deterioration being due to acid rain or acid mists and fogs
60

UK Acid Waters Review Group, 'Acidity in United Kingdom Fresh Waters', HMSO, London,
1986 and Second Report, 1988.
61
UK Terrestrial Effects Review Group, The Effects of Acid Deposition on the Terrestrial
Environment in the United Kingdom', HMSO, London, 1988.
62
UK Terrestrial Effects Review Group, Second Report, 'Air Pollution and Tree Health in the
United Kingdom', Department of the Environment, London, 1993.

leaching nutrients and resulting in the breakdown of chlorophyll.
Reduced root growth and nutrient uptake follow,
acidification of the ground with consequent effects on the soil
chemistry including elevation of mobile aluminium levels which
can damage the roots.
excess deposition of nitrogen (as nitrate and ammonium) which can
have a variety of effects. In the ground NH4" can release H + during
the process of being oxidized to NO3" by bacteria. The H + can then
be leached out. From the point of view of acidification phenomena,
ammonia should therefore not be regarded as an ally even though,
prior to its transformation to nitrate, it reduces the rain-water
acidity.
Given the complexity of the biological and chemical processes mentioned it is clear that control of the effects of 'acid rain' on biological
systems must focus on all the relevant pollutants—SO2, NOx, NH3, and
hydrocarbons (as precursors of ozone).
The third problem associated with acid rain is the attack on stonework
and the decay of famous cathedrals and other buildings constructed of
limestone.63 Both wet and dry deposition of sulfur dioxide are involved.
Under moist conditions SO2 or sulfuric acid will convert calcium
carbonate to gypsum, CaSO4.2H2O. Since the sulfate is more soluble
than the carbonate, the reacted stone can be removed by dissolution. The
solid gypsum also occupies a larger volume than the original carbonate
and this leads to spalling of material from the surface. In polluted urban
areas a black 'crust' of soot and gypsum builds up on stonework not
washed by rain. A combination of these factors leads to a rate of loss of
stone which depends on the deposition rate of SO2 to the surface.
Deposition to a moist surface is more rapid than to a dry surface so the
fraction of time the surface is wetted as well as the SO2 concentration is
important. It is generally assumed that in urban areas the dry deposition
OfSO2 gas is the major factor rather than the acidity of rain itself.
5.4.3 Patterns of Deposition and Critical Loads Assessment. In order
to highlight areas where acidity is causing damage, and therefore to
enable more effective control measures, a critical loads approach has
been introduced throughout Europe. A critical load for a particular
receptor-pollutant combination is defined as the highest deposition load
that the receptor can withstand without long term damage occurring. A
gridded critical loads map is prepared based on the geology and other
factors affecting the response of fresh waters or terrestrial ecosystems to
63

UK Building Effects Review Group, 'The Effects of Acid Deposition on Buildings and Building
Materials', HMSO, London, 1989.

pollutant input. Plates 4a and 4b show where in the UK the critical loads
for acidity have been exceeded for soils and fresh waters respectively.64
For soils, there is a regional pattern, with the highest exceedence
occurring in the most sensitive regions, notably Wales, the Pennines,
and the Scottish Highlands which are also regions of high rainfall. The
exceedence for waters is more localized in particular parts of the
Pennines, western Scotland, and northern and southern Wales. Future
patterns of exceedence of critical loads can be predicted on the basis of
expected reductions in pollutant emissions. The needs for more stringent
control measures can then be assessed.
Following emissions reductions of SO2 within the EU the total area of
Europe with exceedences of critical loads for sulfur has been reduced by
50% between 1980 and 1994. Exceedences for acidity do occur, however,
partly because similar reductions in NO x and ammonia have not
occurred.45 The situation is similar in the US where a decreasing trend
in sulfate ions has been seen in recent years but along with slight
increases in nitrate ions.
Questions
1. Discuss the potential uncertainties which affect the predictions of
global climate models.
2. Discuss the origins of turbulence in the atmospheric boundary
layer, and its potential impact on pollution dispersion processes.
3. Choose two of the world's megacities (i.e. those with a population
over 10 million) and compare their air quality problems. Include
in your discussion any relevant differences in meteorological
conditions and emissions profiles.
4. Contrast the factors influencing air pollution in major cities of
developing countries with the situation in major cities in Europe
and the USA.
5. Why have particulate emissions from vehicles become an increasing cause for concern in the last few years?
6. Describe, with reference to chemical emissions and meteorological
processes, why the exceedences for ozone are much higher in the
cities of Los Angeles and Athens than for other European and US
cities.
7. Explain the reasons why defining abatement strategies for ozone
in Europe is potentially difficult, and explain why the strategies
might be different for different countries.
64

http://www.environment-agency.gov.uk/s-enviro.html 'UK State of the Environment Report',
January 1998.

8. Recent years in the UK have seen a slight increase in rural ozone
but a slight decrease in average urban values. Explain these trends
in terms of emissions and chemical processes.
9. Trace the path of NO x emissions from their emission from a
power station to the point of impact in a mountain lake. In your
answer refer to all possible chemical and physical processes which
may be undergone.
10. What factors determine the relative importance of wet and dry
deposition for different pollutants?
11. Describe how critical loads are used to design abatement strategies
for acidity. Highlight the problems associated with the critical
loads approach.
12. VOCs arise from a very wide range of sources. Discuss practical
steps that may be taken to reduce these emissions.

Previous Page
Ca)

CWF

10~5 M, activity coefficients should be calculated using equations (3)(5). Where the ionic strength of the solution is <10~ 2 3 M,10 activity
coefficients can be calculated using the Debye-Hiickel (DH) expression.
In more concentrated solutions, values calculated using equation (3)
differ from experimental data and this stems from the assumption that
the solute ions are point charges, Le. of infinitely small radius. The
extended DH expression, equation (4), includes a parameter, a, which is
related to ion size. For solutions of ionic strength <10~ 1 M either
equation (4) or the Giintleberg approximation, equation (5), which
incorporates an average value of 3 for a, should therefore be used.10
The Giintleberg approximation is particularly useful in calculations
where a number of ions are present in solution. It should be noted that
single ion activity coefficients are not measurable and are only used to
simplify calculations. Only products or ratios of ion activity coefficients
are directly measurable. Using the above theory, but replacing zf with
z+z_, a mean ion activity coefficient can be calculated for an electrolyte
(Example 1).
EXAMPLE 1 Calculate the mean ion activity coefficient of (i) a 0.001 M and
(H) a 0.01 M aqueous solution of magnesium sulfate.
Ionic strength,
= 0.004 M
= 0.04 M
Log mean activity coefficients
= -0.13
= -0.33
Mean activity coefficients

W. Stumm and J. J. Morgan, 'Aquatic Chemistry', 3rd Edn., John Wiley & Sons, New York, 1996.

These equations tend to underestimate activity coefficients for solutes
in higher ionic strength solutions ( > 0 . 1 M ) and various empirical
correction factors have subsequently been added to the Giintleberg
approximation. The Davies equation incorporates an empirical factor
— 0.2/or —0.3/within equation (5) to calculate activity coefficients for
intermediate ionic strength aqueous solutions (0.1 to 0.7 M). This was
developed further in the Specific Ion Interaction Theory (SIT),11'12
which adds an ion- and electrolyte-specific correction factor to extend
the range of application to 3.5 M. At low ionic strength these models are
less accurate than the extended DH model because the DH expression
used in Davies and SIT equations has a fixed average ion size
parameter, a = 3 and a = 4.6, respectively. For high ionic strength
solutions (e.g. brines), the Pitzer equations13 add terms for binary and
ternary interactions between solute species to a DH expression in
calculations of activity coefficients. Although naturally occurring
brines and some high ionic strength contaminated waters may require
the more complicated expressions developed in the Davies, SIT, or
Pitzer models, the use of equations (3)-(5) is justified for the ionic
strengths of many freshwaters.
2.1.3 Equilibria and Equilibrium Constants. Many of the important
reactions involving solutes in freshwaters can be described by equilibria.
This approach means that an equilibrium constant, K, relating the
activities of the solutes, can be defined for each stoichiometric equilibrium expression.
(6)
(7)
The equilibrium constant can also be defined in terms of the concentrations of the solutes.
(8)
11

I. Grenthe and H. Wanner, 'Guidelines for the Extrapolation to Zero Ionic Strength', Report
NEA-TDB-2.1, F-91191, OECD, Nuclear Energy Agency Data Bank, Giv-sur Yvette, France,
1989.
12
D. K. Nordstrom and J. L. Munoz, 'Geochemical Thermodynamics', Benjamin/Cummings,
Menlo Park, CA, 1985.
13
K. S. Pitzer, Thermodynamic Modelling of Geological Materials: Minerals, Fluids and Melts',
ed. I. S. E. Carmichael and H. P. Eugster, Reviews in Mineralogy, Mineralogical Society of
America, 1987, Vol. 17, p. 97.

For infinitely dilute solutions, { } = [ ] and the equilibrium constant can
be written as:
(9)
For solutions of fixed ionic strength, or, for example, where major ions
in solution are present at concentrations several orders of magnitude
greater than the solute, it can be assumed that the solute activity
coefficients are also constants and can be incorporated into the equilibrium constant. The equilibrium constant for a fixed ionic strength
aqueous solution is termed a constant ionic strength equilibrium constant, CK.
where

(10)

This again enables the use of concentrations rather than activities in
equilibrium calculations.
It is sometimes advantageous to use a mixture of activities and
concentrations and a mixed equilibrium constant, K\ is defined as:
where

(H)

For example, a mixed acidity constant is frequently used where pH has
been measured according to the IUPAC convention as the activity of
hydrogen ions but the concentrations of the conjugate acid-base pair are
used.
The relationship between K and CK or K and K' as defined above can
be used to calculate the effect of ionic strength of solution on the true
equilibrium constant, K. 0Kand K' can be calculated using equations (3)(5) together with experimentally determined values of ^(Example 2).
EXAMPLE 2 Use the Guntleberg approximation to calculate the mixed acidity
constant, YJ ,for a 1 litre solution containing 1 x 10 ~4 moles of an organic acid
(K = 1.77 x 10~4) and (i) 0.01 M, (H) 0.1 M with respect to sodium chloride.
(Hint: take logs of both sides of Y^ —

aci

K and substitute Guntleberg approxi-

ianion

motion expressions for yacid and yanion;

use zacid = 0 and zanion =

-1).

Clearly, the effect of increasing the ionic strength of solution is to decrease the
value of pK\ i.e. increasing K'. For / = 0.1 M9K' is 2.34 x 10~4 compared with
1.77 x 10~4for an infinitely dilute solution.

2.2

Dissolution/Precipitation Reactions

Section 2.1 has provided relationships between solute activities and
concentrations in solution in order that solute behaviour can be
quantified. This section discusses dissolution and precipitation reactions
that impart or remove solutes in natural waters and therefore modify the
chemical composition of natural waters.
2.2.1 Physical and Chemical Weathering Processes. Initial steps involving physical weathering by thermal expansion and contraction or
abrasion lead to the disintegration of rock. Disintegration increases
the surface area of the rock which can, in the presence of water,
undergo chemical weathering. Water acts not only as a reactant but
also as a transporting agent of dissolved and particulate components
and so weathering processes are extremely important in the hydrogeochemical cycling of elements. Mineral dissolution reactions often
involve hydrogen ions from mineral or organic acids, e.g. acid hydrolysis of Na-feldspar in equation (14). Alternatively, the transfer of
electrons (sometimes simultaneously with proton transfer) may
promote the dissolution of minerals, e.g. reductive dissolution of
biotite.
Chemical weathering of minerals results not only in the introduction
of solutes to the aqueous phase but often in the formation of new
solid phases. Dissolution is described as congruent, where aqueous
phase solutes are the only products, or incongruent, where new solid
phase(s) in addition to aqueous phase solutes are the products. These
reactions can be represented by equilibria, where the equilibrium
constant is related only to the activities of the aqueous solutes, i.e. an
assumption is made that the activities of solid phases and that of water
have the value of unity. Examples of weathering of primary minerals,
those formed at the same time as the parent rock, are shown in
equations (12)—(15):
Congruent dissolution of quartz:
(12)
quartz

silicic acid

(13)

At pH < 9, the equilibrium between quartz and silicic acid can be
represented as in equation (12) and the value of K at 298K is
1.05 x 10~ 4 mol P 1 , 1 4 indicating that the solubility of quartz is low. At
higher pH values, the dissociation of silicic acid results in the increased
solubility of quartz (see Section 2.4.1).
Incongruent dissolution of Na-feldspar:
Na-feldspar

silicic acid

kaolinite

(14)
(15)

Alteration of primary minerals such as the feldspars gives rise to
secondary minerals such as kaolinite (Al2Si2O5(OH)4), smectites {e.g.
Na 0 ^ 3 Al 233 Si 367 O 10 (OH) 2 ), and gibbsite (Al(OH)3). As the composition
of the new mineral phases and the potential for further alteration of
secondary minerals are a function of the prevailing geochemical conditions, many equilibrium expressions must be used to fully describe
chemical weathering processes in natural systems. For example, reactions leading to the formation of kaolinite from the primary mineral, Nafeldspar, as well as alteration of the secondary mineral, e.g. to gibbsite,
must be considered. The composition of the aqueous phase is of major
importance in determining the nature of the solid products formed
during chemical weathering. In particular, the solution activities of
silicic acid, metal ions, and hydrogen ions are key parameters influencing
the formation and alteration of new solid phases (Example 3).8
2.2.2 Solubility. Dissolution of minerals during chemical weathering
releases species into solution but aqueous phase concentrations are
limited by the solubility of solid phases such as amorphous silica,
gibbsite, and metal salts.
For example, with respect to the dissolution of magnesium sulfate and
magnesium chloride:
(16)
(17)
The equilibrium solubilities are defined by the respective solubility
products, KST>.
14

J. W. Ball and D. K. Nordstrom, 'Users Manual for WATEQ4F, with Revised Thermodynamic
Data Base and Test Cases for Calculating Speciation of Major, Trace and Redox Elements in
Natural Waters', US Geological Survey Open File Report, Menlo Park, CA, 1991, p. 91.

(18)
(19)
A solution is considered to be undersaturated, saturated or oversaturated with respect to a solid phase, for example magnesium sulfate,
if K3* > {Mg2 + }{SOi-} o b s e r v e d , ^sp = {Mg2 + }{SOi-} o b s e r v e d or
A:SP < {Mg2 + }{SC>4 }Observed, respectively.
EXAMPLE 3 Investigate the stability relationship between gibbsite and kaolinite in natural waters atpH

< 7. ^kaoiimte-gibbsite

10~4'4 at 25 0C.

Considering the formation of gibbsite from Na-feldspars (pH < 7):

This equilibrium expression represents congruent dissolution of Na-feldspar but
the solution quickly becomes saturated with respect to the solid phase Al(OH)3
(gibbsite, A:SP = 10 ~ 3 3 9 ).
e.g.
The concentrations of these species may also be influential in biological systems
where, for example, the presence of silicic acid may be responsible for the
decreased toxicity of aluminium (see Section 3.1.2). The concentration of silicic
acid in solution is a key parameter influencing the stability of gibbsite. In the
presence of even low concentrations of silicic acid, e.g. 10~4M, gibbsite is
converted to kaolinite.

In natural systems, the role of water as a transporting agent is important.
Geochemical conditions promoting removal of silicic acid from the solid-water
interface {e.g. water flow) favour the formation of gibbsite over kaolinite.
(Adapted from Drever.8)
2.2.3 Influence of Organic Matter. Natural organic matter is present
in most natural systems. With respect to weathering processes, its
importance in freshwaters, and indeed in soils and sediments in contact
with freshwaters, can be attributed to the presence of organic acids,
which include low molecular mass compounds, e.g. oxalic acid, and
extremely complex, high molecular mass coloured compounds described
as humic substances—a heterogeneous mixture of poly functional macromolecules ranging in size from a few thousand to a few hundred
thousand Da. In addition to inorganic acids, e.g. H 2 CO 3 , these provide

hydrogen ions for the acid hydrolysis of minerals. As well as promoting
dissolution, natural organic matter can influence the formation of new
mineral phases.15

2.3

Complexation Reactions in Freshwaters

In this section, as a starting point, it is assumed that all species in solution
are in hydrated form, e.g. Al(H2O)^+ where the six water molecules form
the first co-ordination sphere of Al. The hydrated form is often
represented as, for example, Al(aq).
2.3.1 Outer and Inner Sphere Complexes. Outer sphere complexation
involves interactions between metal ions and other solute species in
which the co-ordinated water of the metal ion and/or the other solute
species are retained. For example, the initial step in the formation of ion
pairs, where ions of opposite charge approach within a critical distance
and are then held together by coulombic attractive forces, is described as
outer sphere complex formation.
Mg& + SO^aq) ^ (Mg2+ (H2O)(H2O)SOi-)(aq)

(20)

The formation of ion pairs is influenced by the nature of the oppositely
charged ions, the ionic strength of solution, and ion charge. Ion pairs are
generally formed between hard (low polarizability) metal cations, e.g.
Mg2 + , Fe3 + , and hard anions, e.g. CO 2 ~, SC>4~, and ion pair formation
is generally most significant in high ionic strength aqueous phases. In
freshwaters, which are frequently of low ionic strength, ion charge is a
key parameter. Greater coulombic interactions occur between oppositely
charged ions with higher charge.
Inner sphere complexation involves interactions between metal ions
and other species in solution which possess lone pairs of electrons. Inner
sphere complexation involves the displacement of one or more coordinating water molecules and the transfer of at least one lone pair of
electrons. Those species which possess electron lone pairs are termed
ligands and complexation reactions may involve inorganic and/or
organic ligands.
2.3.2 Hydrolysis. In aqueous systems, hydrolysis reactions are an
important example of inner sphere complexation for many metal ions.
The interaction between hydrated metal ions and water can be written as
15

F. J. Stevenson, 'Humus Chemistry: Genesis, Composition, Reactions', John Wiley & Sons, New
York, 1994.

a series of equilibria for which the stepwise equilibrium constants are
denoted *ATnf.
(21)
(22)
(23)
(24)
Hydrolysis and, more generally, complex formation equilibria may be
described by cumulative stability constants, jS. Using equations (21)(24):
(25)
(26)
(27)
(28)
Clearly, these hydrolysis reactions are dependent on hydrogen ion
activity. The relationship between speciation and pH, which is influenced
by the characteristics of the metal ion, will be discussed further in Section
2.4.1.
2.3.3 Inorganic Complexes. The main inorganic ligands in oxygenated freshwaters, in addition to O H " , are HCO 3 ", C O 3 " , Cl", SO4-,
and F " , and, under anoxic conditions, HS~ and S 2 " (see Section 2.4.4).
The stability of complexes formed between metal ions and inorganic
ligands depends on the nature of the metal ion as well as the properties of
fThe hydrolysis of hydrated metal ions can also be written as the interaction between the hydrated
metal ion and the hydroxyl ligand. For example:

K\ and *K\ are related as follows:

so
The values for *Kn are those obtained by Wesolowski and Palmer;16 all values are given in unitless
form but fundamentally are defined in terms of the products of the activities of ions on the molar or
molal scales.
16
D. J. Wesolowski and D. A. Palmer, Geochim. Cosmochim. Ada, 1994, 58, 2947.

Table 2

Major

species

freshwater'S

Metal ion

Major species

[M}aqJ]/[MrTOTAL (aq)]

Mg"
Ca"
Al 111
MnIV
Fe 111
Ni 11
Cu 11
Zn"
Pb"

Mg2+
Ca1+
Al(OH)?, Al(OH) 2 + , Al(OH) 4 MnO2*
Fe(OH)?, Fe(OH) 2 + , Fe(OH) 4 Ni 2 + , NiCO?
CuCO?, Cu(OH) 2 3
Z n 2 + , ZnCO?
PbCO?

0.94
0.94
1 x 10"9
2 x 10"u
0.4
0.01
0.4
0.05

Adapted from Sigg and Xue, 17 and reproduced with permission from W. Stumm and J. J. Morgan,
'Aquatic Chemistry', 3rd Edn., © John Wiley & Sons, Inc., 1996.

the ligand. Table 2 shows some of the major complexed metal species
involving inorganic ligands in oxygenated freshwaters at pH 8.
2.3.4 Surface Complex Formation. Metal ions form both outer and
inner sphere complexes with solid surfaces, e.g. hydrous oxides of iron,
manganese, silicon and aluminium. In addition, metal ions, attracted to
charged surfaces, may be held in a diffuse layer which, depending upon
ionic strength, extends several nanometres from the surface into solution. Diffuse layer metal retention and outer sphere complex formation
involve electrostatic attractive forces which are characteristically weaker
than co-ordinative interactions leading to inner sphere surface complex
formation. A number of factors influence metal interactions with
surfaces, including the chemical composition of the surface, the surface
charge, and the nature and speciation of the metal ion. The importance
of the pH of the aqueous phase in these interactions will be discussed
further in Section 2.4.1.
2.3.5 Organic Complexes. Dissolved organic matter consists of a
highly heterogeneous mixture of compounds, including low molecular
mass acids and sugars as well as the high molecular mass coloured
compounds termed humic substances. Humic substances are secondary
synthetic products derived from decaying organic debris. Although they
are structurally poorly defined, it is accepted that large numbers of
mainly oxygen-containing functional groups are attached to a flexible,
predominantly carbon backbone. Individual organic molecules, in particular those of humic substances, can provide more than one functional
group for complex formation with a hydrated metal ion. Ligands which
17

L. Sigg and H. B. Xue, 'Chemistry of Aquatic Systems: Local and Global Perspectives', ed. G.
Bidoglio and W. Stumm, Kluwer Academic, Dordrecht, 1994.

provide two and three functional groups are termed bidentate and
tridentate respectively. The formation of complexes where more than
one functional group from the same organic molecule is involved are
more stable than those where functional groups are from discrete organic
molecules. The concentration of dissolved organic matter in freshwaters
is generally low ( 2 - 6 m g C P 1 ) and humic substances, comprising
molecules which possess large numbers of functional groups in numerous different chemical environments, are implicated as the component of
natural organic matter most important in metal binding.15
2.4 Species Distribution in Freshwaters
2.4.1 pH as a Master Variable. pH is one of the key parameters which
influence species distribution in aqueous systems. Many equilibrium
expressions contain a hydrogen ion activity term, e.g. in acid-base,
complexation and surface charge formation equilibria. It is therefore
useful to consider the relationship between the activity of the species of
interest {e.g. contaminant metal ions, organic pollutants, naturally
occurring inorganic and organic solutes, and weakly acidic functional
groups on mineral surfaces) and the activity of hydrogen ions.
For example, an acid-conjugate base pair can be represented as HA and
A " , respectively:
(29)
The activity of acid initially in solution, C, termed the analytical activity,
is equal to the sum of the equilibrium activities of the acid and its
conjugate base.
(30)
Equations (29) and (30) can be combined to give an expression for the
activity of the conjugate base in terms of the equilibrium constant, the
analytical activity, which is also a constant, and the hydrogen ion
activity.
(31)

(32)

Similarly an expression for the activity of the acid can be written:
(33)

Alternatively, the activities of acid and conjugate base can be represented
as a fraction of the analytical activity:
(34)
(35)
(36)
a0 and (X1 are termed dissociation fractions for the acid and conjugate
base, respectively. The expressions for the dissociation fractions are
independent of the analytical activity.
A plot of dissociation fraction against pH for an acid-conjugate base
pair is shown in Figure 2.
pH = pK

2,4,6-TCP"

2,4,6-TCP

PH
Figure 2

The influence ofpH on dissociation fractions OLQ and (Xj for 2,4,6-trichlorophenol

More frequently log { } against pH is plotted using equations (37) and
(38):
(37)
(38)
The equilibrium pH can be obtained directly from the log { } versus pH
plot by adding lines showing log {H+ } and log {OH~} versus pH and by
using an expression for charge balance:
(39)
For a solution containing a weak acid, the charge balance relationship,
equation (39), is satisfied at the equilibrium pH (Example 4). The same
log { } versus pH plot can also be used to determine the equilibrium pH
of a solution containing the salt of a weak acid, e.g. NaA, by using the
appropriate charge balance expression. (Hint: {Na + } = C, so {HA} +
{H + } = {OH"}).
EXAMPLE 4 An organic pollutant, 2,4,6-trichlorophenol (2,4,6-TCP), has
been released into a natural water (assume I = 0 M). Illustrate graphically the
relationship between (i) log {2,4,6-TC1P) and pH; (H) log {2,4,6-TCP~} and
pH. Use {2,4,6-TCPTOTAL} = 4X 10~4 MandK = 10~613.
Graphical representation is best approached by dividing the pH range into two
regions, pH < pK and pH > pK. The slope of the lines in each of these two
regions can be determined by differentiating equations (32) and (33) with respect
topH.
« fco**stant) = 0 ; K i s
dpH
large relative to {H+ } at pH >pK.
Note that

d lo

At pH < pK:

s m a l l re l a tive

to {H+ } at pH < pK; K is

-log

{a c tiv ity }

{H+} = {2,4,6-TCP'} + {OH'}

PH
Graph to Example 4

At pH > pK:

log{2,4,6-TCP} = log(4 x 1(T4) for pH < pK and log {2,4,6-TCP~} =
log (4 x 10~4) for pH > pK. Using the full expressions for log {2,4,6-TCP}
and log {2,4,6-TCP-}, calculate several points on lines of slope H-1 and — 1 to
complete the graph.
Carbonate Equilibria. Another example of the importance of acidbase behaviour, not only in aerated freshwaters but also in seawater
(see Chapter 4), is the dissociation of diprotic acids such as carbonic
acid, H 2 CO 3 . The hydration of dissolved CO2(aq) in natural waters
gives rise to carbonic acid. The dissociation of carbonic acid not only
influences the pH of the water but provides ligands which can complex

trace metals. Example 5 illustrates the relationship between log { }
and pH for a closed aqueous system which contains dissolved CO2
(the presence of mineral phases has been ignored). By assuming that a
system is closed to the atmosphere, it is possible to treat carbonic acid
as a non-volatile acid and to consider only the hydration reaction
which converts CO2(aq) to H2CO3. This means that the analytical
activity of carbonic acid, C, is a constant for a given system. The
dissociation of carbonic acid is described by equations (40) and (41)
and the equilibrium constants, K\ and K2, are defined by equations
(42) and (43).
(40)
H2CO3* is used to represent the sum of CO2(aq) and H2CO3, i.e. to take
account of the presence of CO2(aq) which is in equilibrium with H 2 CO 3 .
(41)
(42)
(43)

As before, the analytical activity of a weak acid can be written as the sum
of undissociated and dissociated acid species:
(44)
Combining equations (42)-(44), an expression for each of these
(H2CO3*, HCO3", and CO 3 ") is obtained:
(45)
(46)
(47)
The dissociation fractions, a0, aJ? and a2, are obtained by dividing
each of equations (45)-(47) by the analytical activity, C (cf.

equation 44):
(48)
(49)
(50)

where
(51)

Graphical representation of log { } versus pH can again be used to
obtain the equilibrium pH (Example 5).
EXAMPLE 5 Illustrate
graphically
the
relationship
between
(i)
log [H2CO3*), (H) log [H2CO3], (Ui) log [HCOJ), and (iv) log [COJ') and
pH for a closed aqueous system. Use [carbonic acidT), C = 2 x 10~5 M,
K 7 = 5.7 x 10-7andK2 = 5.1 x 10~u.
The presence of dissolved CO 2 is taken into account by the following
equilibrium:
CO2(aq) + H 2 O 0 ) ^ H2CO3(aq)

^hydration = ^

= 1.54 X 10"3 at 298 K

By defining {HCO3*} as (H 2 CO 3 *) = {CO 2(aq) } + (H 2 CO 3 ) and using the
approximation (HCO 3 *) ~ (CO2(aq)} because the hydration equilibrium
lies far to the left, £hydration = {H2CO3}/{H2CO3*} and (H 2 CO 3 ) =
(H 2 CO 3 *) x 1.54 x 10~ 3 . In this way the true activity of carbonic acid can be
determined.
By assuming a closed system, the total analytical activity of carbonic acid, C,
is constant [this is not true for an open system (Example 6)]. Thus equations
(42)-(44) can be combined to give equations (45)-(47). The graphical representation can be constructed by dividing the pH range into three regions,
pH < pATi, pATi < pH < pK2, and pH > pK2, and the procedure outlined in
Example 4, i.e. obtaining logarithmic expressions for (H 2 CO 3 *), ( H C O f ) and
(CO 3 - } and differentiating each with respect to pH, can then be utilized to
construct the plot of log { } against pH.

AtpH pK2:

The charge balance expression (52) can be used to determine the equilibrium
pH.
Example 6 extends the consideration of the carbonate system to include
equilibration of an aqueous system with an atmosphere (e.g. air)
containing CO2(g), i.e. an open system.

EXAMPLE 6 Illustrate graphically the relationship between (i) log
[H2CO3*), (H) 1Og[H2CO3], (Ui) log[HCOj), and (iv) log[CO23~) andpH
for an open aqueous system. Use pCo2 = 3.5 x 10~4 atm, KH = 4.8 x 10~2 (at
288K), K1 = 5.1 x 10-7andK2 = 5.1 x 10~n.

where KH is the Henry's Law constant for atmosphere-aqueous phase equilibria
involving gases.
Clearly {H2CO3*} is now a constant for a fixed partial pressure of CO2 and is
independent of pH, i.e.

This means that {H2CO3} is also constant over the entire pH range.
Combining this additional equilibrium expression with equations (42) and (43)
gives expressions for (HCO3"} and (CO 3 "}:

- l o g { a c tiv ity }

and

PH
Graph to Example 6

Additionally, the expression for C becomes:
C = Kupcojtto, which increases with increasing pH according to the function

Plotting log { } versus pH for (HCO3"} and (CO 3 ") gives lines of slope + 1 and
+ 2 respectively. The charge balance expression, which is the same as that for
the closed system, can again be used to determine the equilibrium pH.
Alkalinity. An important definition arising from considerations of
carbonate equilibria is that of alkalinity which is derived with reference
to the charge balance expression for carbonic acid:
(52)
Alkalinity is described as the acid neutralizing capacity of an aqueous
system or equivalently as the amount of base possessed by the system:
Alkalinity =

(53)

It is important to understand that, by definition, alkalinity is independent of addition or removal of CO 2 (or H 2 CO 3 ) from the system (cf.
equation (52)—H 2 CO 3 does not appear in the charge balance expression). Alkalinity is an important parameter in assessing the effects of
environmental change on aqueous systems (see Section 3.4.1). Species
other than carbonate can contribute to alkalinity and an alternative
definition is:
Alkalinity
(54)
pH Dependence of Complex Formation. Other equilibria such as metal
complexation reactions can be considered as acid-base reactions and
plots of log { } against pH also provide information about the dominant
species present in solution under different geochemical conditions.
Hydrolysis of metal cations occurs progressively with increasing pH,
e.g. M&), M(OH)&), M(OH) 2 + ^, M(OH)§ (aq) , and M(OH)« aq) are the
hydrolysis products for many M cations (see Section 2.3.2). M(aq) is
generally found at low pH and only at extremely high pH is M(OH)4"
formed. The dependence of hydrolysis reactions on pH is illustrated
further in Example 7 and in Section 3.1.2.

EXAMPLE 7 What are the dominant species of Fe111 present in an aqueous
solution in the absence of dissolved CO2 at pH 1, 6, and 11? Use
[Fe1I1) = 10~9 M and log *pj = -3.05, log *p2 = -6.31, log *j?5 = -13.8 and
log *p4 = —22.7. Assume that no Fe(OH)3 is precipitated, i.e. homogeneous
solution.

This equation can be solved for {Fe3 + } at various pH values by using the
equilibrium constants for each of the four equilibria above.

-log

{activity}

The concentrations of all other species can be calculated using {Fe3 + } at each
selected pH value. Plotting —log { } against pH illustrates that the dominant
species of Fe111 are Fe 3 + , Fe(OH)2+ and Fe(OH)4T at pH 1, 6, and 11,
respectively.

pH
Graph to Example 7

It should be noted that, for higher values of (Fe"1), equilibria for polymeric
iron species should be included and that Fe(OH)3 may precipitate. (The
formation of polymeric iron species may be an intermediate step leading to the
formation of the solid phase.10)

Example 8 illustrates the influence of pH on species distribution for a
Cu n -CO 2 -H 2 O system.

EXAMPLE 8 What are the dominant species of Cu11 at pH 4, 9, and 12 in an
aqueous solution where Ccarbonate = 2 x 10~3 M and [Cu") =5 x 10~8 M?
Use log*p}=-8.0,
log *p2 =-16.2,
log *fi3 =-26.8,
log *jS4 = -39.9,
log*Picarbonate = 6-77, log *P2carbonate = ^'0.01. Also required are the acidity
constants for carbonic acid:

It is first necessary to calculate {CO3""} for different pH values. These values can
then be used together with the stability constants to solve for {Cu2 + } at the
different pH values. Then the activity of each species can be calculated and plots
of — log { } against pH, { } against pH or % distribution

- l o g { a c tiv lty}

illustrate that Cu 2+ , CuCO 3 , and Cu(OH)3" are the dominant species at pH 4,9,
and 12, respectively.

PH
Graph to Example 8

At low pH values, ligands such as F and SO4 may compete
successfully with the hydroxyl ion for a metal cation (see Section 3.1.2).
Most significantly, the formation of such complexes increases the
solubility of the metal ion by several orders of magnitude over that
predicted by the solubility product of the solubility-limiting solid phase,
e.g. hydroxides. The presence of even trace concentrations of organic
ligands has an even greater effect over a wide pH range on metal cation
speciation (Example 9) and solubility, increasing the aqueous phase
concentration by up to seven orders of magnitude.18 Complexation
clearly has an important role to play, not only in reducing metal toxicity
(see Section 3.1.2), but in markedly increasing metal ion mobility.
EXAMPLE 9 What are the dominant species of Cu11 at pH 8,
dissolved carbonate species, where {CUT} = 5 x 10~8 M and
(Yi=IO-413,
Porganic= JO78 at pH5) is present at 2
log *PJ = -8.0, log *p2 = -16.2, log *fc = -26.8, log *p4 =

in the absence of
an organic acid
x 10~7 M? Use
-39.9.

/?organic{OrganicT

Hence, at pH 5 more than 90% Cu11 is bound by the organic ligand.
18

D. Langmuir, 'Aqueous Environmental Geochemistry', Prentice-Hall, Englewood Cliffs, NJ,
1997.

Influence of Ionic Strength. The ionic strength of solution will influence
the activities of species and so equilibrium constants must be adjusted to
take account of ionic strength as discussed in Section 2.1.2. This applies
not only to the acid-base equilibrium constants but also to the dissociation constant for water and the stability constants for complex formation. The values of equilibrium constants also apply only at a specified
temperature, e.g. 298 K.
2.4.2 ps as a Master Variable. Chemical speciation is also influenced
by the redox conditions prevailing in natural waters. Although redox
reactions are often slow, and therefore species are present at activities far
from equilibrium, they are commonly represented by thermodynamic
equilibrium expressions which can provide the boundary conditions
towards which a system is proceeding.
pe, a parameter describing redox intensity, gives the hypothetical
electron activity at equilibrium. It measures the relative tendency of a
solution to accept or donate electrons with a high pe being indicative of a
tendency for oxidation, while a low pe is indicative of a tendency for
reduction, pe = — log {e~} is analogous to the pH scale
(pH = — log{H + }) since a low value for ps is obtained where the
hypothetical {e~} is large (pH is low where (H + } is large) and conversely
a high value of ps is obtained where {e ~ } is small (pH is high where {H+ }
is small). pe is related to the electrode potential, EH, by the expression
pe = 16.9 Eu (V). A full treatment of this relationship can be found in
Reference 10. Although the electron activity is a hypothetical phenomenon, pe is a useful parameter to describe the redox intensity of natural
systems and hence the distribution of species under prevailing redox
conditions (Example 10).
EXAMPLE 10 Calculate the ps values for the following solutions (298 K,
I=O):
(i) A solution at pH 2 containing [Fe3 + ) = 10~45 M and [Fe2+) = 10~2J M
where log K = 13.
(U) A neutral solution containing [Mn2+) = 10~6 M in equilibrium with the solid
phase, Mn(IV)O2 and log K = 40.84.
(i) At pH 2, the hydrated metal ion has not undergone hydrolysis to any
significant extent

-log

{a c tiv ity }

For a fixed pH value (e.g. pH = 2), the species distribution (e.g. where
{FeT} = 1 x 10~3 M) at different ps values can be illustrated by plotting
log { } against pe. This is analogous to the treatment of pH as a master variable.

pe
Graph to Example 10

Redox intensity or electron activity in natural waters is usually
determined by the balance between processes which introduce oxygen
(e.g. dissolution of atmospheric oxygen, photosynthesis) and those
which remove oxygen (e.g. microbial decomposition of organic matter).
Often these processes are controlled by the availability of inorganic
nutrients such as phosphate and nitrate, e.g. as utilized in the formation
of organic matter during photosynthesis (see Section 3.4):

(55)

Table 3

Redox processes—terminal

electron

Terminal electron acceptor

O2
NO^" (reduction to N 2 )
MnIV
NO 3 " (reduction to NH 4 + )
Fe 111
reducible organic matter
SOI"
CO 2

13.75
12.65
8.9
6.15
-0.8
—3.01
-3.75
-4.13

acceptors

Adapted with permission from W. Stumm and J. J. Morgan, 'Aquatic
Chemistry', 3rd Edn., © John Wiley & Sons, Inc., 1996.

The decay of organic matter produced in this manner leads to the
subsequent consumption of oxygen, e.g. by respiration:
(56)
The decay of organic matter requires the presence of a terminal electron
acceptor and in equation (56) molecular oxygen is reduced to water.
Other terminal electron acceptors present in natural waters include
NO3", MnIV, Fe111, SO4~, and CO2. Once all molecular oxygen has
been consumed, organic matter is decomposed via reactions involving
other terminal electron acceptors in a series determined by the ps
intensity as shown in Table 3.
The sequence of redox reactions involving organic matter, all of which
are microbially mediated, can be thought of as progressing through
decreasing levels of pe and so, for example, NO^" in denitrification
reactions would be utilized as the electron acceptor before MnIV and
SO^" would be utilized before CO2 (see Section 3.3.2).
Redox processes are important for elements which can exist in more
than one oxidation state in natural waters, e.g. Fe11 and Fe111, Mn11 and
Mn Iv . These are termed redox-sensitive elements. The redox conditions
in natural waters often affect the mobility of these elements since the
inherent solubility of different oxidation states of an element may vary
considerably. For example, Mn11 is soluble whereas MnIV is highly
insoluble. In oxic systems, MnIV is precipitated in the form of oxides
and oxyhydroxides. In anoxic systems, Mn11 predominates and is able to
diffuse along concentration gradients both upwards and downwards in a
water column. This behaviour gives rise to the classic concentration
profiles observed for Mn (and Fe) at oxic-anoxic interfaces as illustrated
in Figure 3.

OxIc
flnoxic

Relative depth

Insoluble
Fe(HI) [hydrjoxides
Production of Fe(HI) by oxidation of Fe(H) and
Oxidation
dissolution of Fe(HI) [hydrjoxide by catalysis
of Fe(HHigand complexes
Setting of Fe(HI)
Diffusion Fe(H)
Redox boundary
Fe(IH) [hydrloxide 5. In particular, Al-F complexes were the dominant form at pH
5-6 whereas at pH > 6, Al-OH species became the major form. In the
presence of organic ligands, however, they found that, over a broad pH
range of 4.3-7.0, alumino-organic complexes were a major component of
23

C. T. Driscoll and W. D. Schecher, Environ. Geochem. Health, 1990,12, 28.

the monomeric aluminium. It is the chemically labile inorganic monomeric aluminium that has been identified as the toxic agent for fish
which, in acidified lakes and streams at aluminium concentrations of
100-20O)UgP1, are unable to maintain their osmoregulatory balance
and are susceptible to respiratory problems from coagulation of mucus
and Al(OH)3 on their gills at their physiological pH of 7.2.24
With respect to human exposure to aluminium in drinking water, it
must be remembered that aluminium is often deliberately added, in the
form of Ak(SC^)3, to water supplies at treatment works to remove
coloration by organic compounds in reservoirs in upland catchments.
This it achieves by hydrolysing to a gelatinous, high surface area,
precipitate of Al(OH)3, which helps to remove the coloured organic
colloids. It has been suggested, partially because of observed water
aluminium-related dialysis dementia in some chronic renal patients on
artificial kidney dialysis machines in the 1960s and 1970s, that
exposure to low concentrations of aluminium in drinking water, for
which there is a current EC Maximum Admissible Concentration of
200/^gI"1, might be implicated in Alzheimer's Disease.25 There is no
proof of this at present, although it is known that the incidence of
senile dementia on the remote Pacific island of Guam, where bauxite is
mined and the local water contains elevated levels of aluminium, is a
hundred-fold greater than in the USA. It has been mooted that, for
humans, a high intake of silicic acid (H4SiO4) could reduce the
bioavailability of aluminium by two orders of magnitude for, at
slightly alkaline intestinal pH, hydroxyaluminosilicates are stable and
unavailable for absorption.26 This would mimic the situation in nature,
where, in the absence of a strong acid input, the bioavailability of
aluminium is kept low by the formation of hydroxyaluminosilicate
species.
3.1.3 Acid Mine Drainage and Ochreous Deposits. One of the major
pollution problems affecting freshwaters is that of acid waters discharged
from coal and metal mines.27 In an active underground coal mine,
pumping lowers the water table. The deeper strata thus become exposed
to air with the result that pyrite (FeS2) present is subject to oxidation,
generating acid conditions.
(61)
24

N. J. Bunce, 'Environmental Chemistry', Wuerz, Winnipeg, 2nd Edn., 1994.
'Aluminium in Food and the Environment', ed. R. C. Massey and D. Taylor, Royal Society of
Chemistry, Cambridge, 1989.
26
J. D. Birchall, Chem. Br., 1990, 26, 141.
27
R. J. Pentreath, 'Issues in Environmental Science and Technology', ed. R. E. Hester and R. M.
Harrison, Royal Society of Chemistry, Cambridge, 1994, Vol. 1, p. 121.
25

When mining and pumping cease, the water table returns to its natural
level. While the flooding of the mine stops the direct oxidation of pyrite,
it does bring the sulfuric acid and iron sulfates into solution. Some of the
ferrous ion (Fe2 + ) may be oxidized to the ferric ion (Fe 3+ ), a slow
process at low pH, but one which can be catalysed by bacteria, and the
ferric ion may react further with pyrite.28
(62)
(63)
On reaching the surface, the acidic mine drainage water is mixed with air
and oxygenated water, leading to rapid oxidation from the ferrous to the
ferric form and the precipitation of the characteristic unsightly orange
ochreous deposits of iron (oxy)hydroxides, e.g. ferrihydrite (Fe(OH)3),
goethite (FeOOH), observed along many stream and river beds.
(64)
(65)
ferrihydrite

(66)
goethite

The overall process may be represented as:
(67)
Micro-organisms (e.g. Thiobacillus thiooxidans, Thiobacillus ferrooxidans, Metallogenium) are involved as catalysts in many of the oxidizing
reactions, which would be extremely slow under the prevailing conditions
of low pH (e.g. <4.5). 29 The combination of acid waters and coatings can
have devastating effects upon aquatic biota, with depletion of freeswimming and bottom-dwelling organisms, the loss of spawning gravel
for fish, and direct fish mortalities. Typical approaches to treatment of
acid mine drainage have included methods to remove the iron floe by
oxidation and to adjust the pH through the use of limestone filter beds.
3.1.4 Acid Mine Drainage and Release of Heavy Metals. The production of sulfuric acid from sulfide oxidation in mines can also lead to the
leaching of metals other than iron, with the result that the emerging
acidic waters may be laden with heavy metals. Thus in the former
metalliferous mining area of south-west England, where there are now
many abandoned mine workings, the closure and flooding of the famous
Wheal Jane tin mine in 1992 led to a highly acidic cocktail of dissolved
28

J. E. Andrews, P. Brimblecombe, T. D. Jickells, and P. S. Liss, 'An Introduction to Environmental
Chemistry', Blackwell Science, Oxford, 1996.
29
C. F. Mason, 'Biology of Freshwater Pollution', Longman, Harlow, 3rd Edn., 1996.

metals (Cu, Pb, Cd, Sn, As) entering the Carnon and FaI rivers at 7-15
million litres per day and spreading throughout the surrounding estuaries and coastal waters.30
In the USA, the EPA has identified over 31 000 hazardous waste sites,
with the largest complex of 'Superfund' sites to be remediated in western
Montana, in the Clark Fork River Basin where there have been more
than 125 years of copper and silver mining and smelting activities. Moore
and Luoma31 have characterized three types of contamination resulting
from large-scale metal extraction: primary, consisting of wastes produced during mining, milling, and smelting and deposited near their
source of origin; secondary, resulting from transport of contaminants
away from these sites by rivers or through the atmosphere to soils,
groundwaters, rivers, etc.; and tertiary, where contaminants may be
remobilized many kilometres away from their point of origin.
Some of the largest waste deposits occur in tailing ponds containing
acid mine water. In the Clark Fork River Complex, it is estimated that
ponds contain approximately 9000 t arsenic, 200 t cadmium, 90 000 t
copper, 20000 t lead, 200 t silver, and 50000 t zinc. These metals enter
streams and rivers as solutes and particulates and contaminate sediments
in the river and reservoirs far downstream from the primary sources
(Figure 4).31 Downstream concentrations follow an exponential decline
when viewed over several hundred kilometres. The sediment of Milltown
Reservoir, more than 200 km from the mines and smelters at Butte and
Anaconda, is highly contaminated with various metals and arsenic. It is
believed that oxidation-reduction processes release arsenic from the
reservoir sediments and cause contamination of an aquifer from which
water drawn through wells, now closed, contains arsenic at concentrations higher than the EPA drinking water standards.31
When open-pit mining ended in the Clark Fork River Complex in
1982, pumping was discontinued, with the result that water began filling
underground shafts and tunnels and the 390 m deep Berkeley pit.
Contaminated acid water containing individual metal and sulfate concentrations thousands of times those in uncontaminated water could
flow into an adjacent alluvial aquifer and eventually over the rim of the
pit. Large-scale hard-rock mining in the western United States, especially
Nevada, has greatly increased in recent years, with the result that deep
'pit lakes' are likely to form as open-pit metal mines intersecting
groundwater are depleted and shut down.32 Pit lakes in high sulfide
rock will tend to have poor quality acidic water, although oxidized rock
that contains significant carbonate will produce better quality, near30

L. E. Hunt and A. G. Howard, Mar. Pollut. Bull, 1994, 28, 33.
J. N. Moore and S. N. Luoma, Environ. ScL Technol., 1990, 24, 1279.
32
G. C. Miller, W. B. Lyons, and A. Davis, Environ. ScL Technol., 1996, 30, 118A.
31

Primary
1a. Waste
rock
1b. Tailings
1c. Slag

Figure 4

Secondary

2a. Ground water at
open pits
2b. Ground water
beneath ponds
2c. Sediment in
river channels

26. Floodplain
sediment/soil
2e. Reservoir
sediment
2f. Soils from
air pollution

Tertiary
3a. River sediment
reworked from
floodplain
3b. Groundwater
from contaminated
reservoir sediment

Types of contamination resulting from large scale metal extraction31
(Reprinted with permission from Environ. Sci. TechnoL, 24, 1279. © 1990
American Chemical Society)

neutral pit lake water. Most of the larger existing pit lakes currently
contain water that does not meet standards for drinking water, agricultural water quality, or aquatic life.32 Factors such as the oxygen status of
the lake, pH, the hydrogeologic flow system, composition of the wallrock, evapoconcentration, biological activity, and hydrothermal inputs
are all important to the modelling of future water quality and impact.
Remediation approaches in circumstances like those described here
have been based on a variety of physical, chemical, and biological
systems. These include the construction of ponds where sufficient
organic matter is available to establish anaerobic conditions and immobilize at least some of the metals {e.g. Cu, Pb, Cd, Zn) as sulfides. Similar
passive treatment of tailings-impacted groundwater has also employed
precipitation of metal sulfides as the key clean-up step.33 Active chemical
treatment with lime, producing sludges, and biosorption by reeds, wetlands, etc., are other methods which have been tried. Although there
have been considerable technical improvements around the world in
recent years in metal extraction and smelting, waste treatment, and
containment, the regulatory requirements are becoming much more
stringent. For example, the new National Pollutant Discharge Elimina33

R. J. Allan, 'Proceedings of the 10th International Conference on Heavy Metals in the
Environment', CEP, Edinburgh, 1995, Vol. 1, p. 293.

tion System permits of the US EPA stipulate > 9 0 % survival of selected
test organisms for 48 hours in water discharges and effluents. While
fathead minnows and water fleas, sensitive at the m g l " 1 level, were
formerly used, the new standard will be the much more sensitive
arthropod, Ceriodaphnia dubia, which is affected at the jxg \~l level. It is
understood that most mines and mine-mill discharges in Missouri's New
Lead Belt may well fail this test.34
3.2

Metals in Water

3.2.1 Arsenic in Groundwater. In what has been described as the
largest arsenic poisoning epidemic in the world, hundreds of thousands
of people in West Bengal, India, have been seriously affected by arsenic
poisoning resulting from the consumption of water drawn from tube
wells sunk some 20 to 150 metres below the ground into aquifers.35 In
West Bengal and other parts of the world {e.g. Bangladesh, Chile,
Mexico, Argentina, Ghana, Mongolia, and Taiwan) where inorganic
arsenic concentrations in drinking water are elevated ( m g l " 1 against
WHO recommended limits of 1 0 / i g P 1 ) , exposure has resulted in
hyperpigmentation, hyperkeratosis, and cancer of the skin.36'37
Sulfide minerals are one of the most important natural sources of
arsenic in groundwater. Oxidation of arsenopyrite (FeAsS), in analogous
fashion to pyrite (FeS2), may release high concentrations of arsenic into
solution. 38
(68)
In Bengal, it has been suggested that large-scale withdrawal of groundwater for irrigation produces seasonal fluctuation of the water table,
which in turn results in intake of oxygen into the pore waters of
sediments that are arsenic-rich in the form of arsenopyrite.39 The exact
speciation of soluble inorganic arsenic, e.g. as the undissociated forms or
different oxyanions of the acids H 3 As m O3 and H3AsvO4, will be
dependent upon the prevailing ps and pH. Sorption of arsenic, especially
34

N. L. Gale, B. G. Wixson, D. Forciniti, and D. Murphy, 'Proceedings of the 15th Annual
European Meeting of the Society for Environmental Geochemistry and Health', British Geological Survey, Keyworth, 1997.
35
T. R. Chowdhury, B. Kr. Mandal, G. Samanta, G. Kr. Basu, P.P. Chowdhury, C. R. Chanda, N.
Kr. Karan, D. Lodh, R. Kr. Dhar, D. Das, K. C. Saha, and D. Chakraborti, 'Arsenic: Exposure
and Health Effects', ed. C. O. Abernathy, R. L. Calderon, and W. R. Chappell, Chapman & Hall,
London, 1997, p. 93.
36
'Arsenic: Exposure and Health', ed. W. R. Chappell, C. O. Abernathy, and C. R. Cothern, Science
and Technology Letters, Northwood, 1994.
37
'Arsenic: Exposure and Health Effects', ed. C. O. Abernathy, R. L. Calderon, and W. R. Chappell,
Chapman & Hall, London, 1997.
38
P. L. Smedley, W. M. Edmunds, and K. B. Pelig-Ba, 'Environmental Geochemistry and Health',
ed. J. D. Appleton, R. Fuge, and G. J. H. Ball, Geological Society, London, 1997, p. 163.
39
P. Bagla and J. Kaiser, Science, 1996, 274, 174.

pentavalent arsenate (e.g. AsOl , HASO4 , H2AsO4 ), onto ferric
hydroxide (Fe(OH)3) produced under oxidizing conditions may,
however, restrict its mobility and availability, although fertilizer phosphate present in groundwater could perhaps displace the arsenate.
Arsenite, especially as H3As111O3 (pK\ = 9.2), the predominant form
under reducing conditions at pH < 9.2, is much less strongly sorbed. As
an alternative to arsenopyrite oxidation, it is possible that elevated
arsenic concentrations in Bengal groundwater could result from the
release of adsorbed arsenate during the dissolution of ferric hydroxide
under reducing conditions (see Section 3.3.2).
Although trivalent inorganic arsenic, with its propensity for binding to
the SH group of enzymes, is acknowledged to be more toxic to humans
than pentavalent inorganic arsenic, it must be recognised that As v can be
converted to As111 in the human body as part of the reduction/biomethylation pathway of excretion:
(69)
where MMAA (CH3AsO(OH)2, monomethylarsonic acid) and DMAA
((CH3)2AsO(OH), dimethylarsinic acid) are less toxic metabolites, the
latter predominating as the major form of arsenic excreted in urine.36'37
It is possible that such methylation could also occur naturally by
microbial action in freshwaters, although reported occurrences suggest
that the effect is small.40
3.2.2 Lead in Drinking Water. The naturally soft, slightly acidic,
plumbosolvent water of the Loch Katrine water supply for the Glasgow
area was recognized many years ago to release lead from the lead pipes
and tanks in the domestic plumbing of the Victorian and subsequent
(even post-World War II) eras.41
(70)
(71)
(72)
Essentially, elemental lead becomes soluble in acidic conditions due to its
oxidation by dioxygen. Furthermore, compounds such as carbonate and
hydroxycarbonate compounds of lead, i.e. PbCO 3 and Pb3(CO3)2(OH)2,
that may coat the pipes, will dissolve under acid conditions.
In view of the concern over detrimental effects of exposure to lead
upon human health, in particular the possible impact upon intelligence
and behaviour of young children, steps were taken in the mid-1970s to
40
41

W. R. Cullen and K. J. Reimer, Chem. Rev., 1989, 89, 713.
M. R. Moore, W. N. Richards, and J. G. Sherlock, Environ. Res., 1985, 38, 67.

reduce the lead content of tap water in Glasgow and other susceptible
areas, which often exceeded the WHO maximum guideline at the time of
100 fig I""1. The method chosen was adjustment of pH to 8-9 by lime
dosing. The effects of liming the Glasgow water supply were quite
dramatic. Whereas pre-1978, when the pH was 6.3, only 50% of
random daytime samples were < 100 fig l~l in lead, the figure increased
to 80% during 1978-1980, when the adjusted pH was 7.8. After 1980,
when the pH was increased further to 9.0, 95% of samples were
< 100 fig I" 1 . 4 1 It appears that carbonate and hydroxycarbonate lead
compounds present in the coatings on the pipes were stabilized. Significant reductions in blood lead levels of key exposed groups (e.g.
pregnant women) were also observed.
Since 1989, as regulatory upper limits for lead in drinking water have
fallen, e.g. to 50 fig 1"* (EC) and now to 10 fig I" 1 (WHO), orthophosphate has been added to the water supply in Glasgow to precipitate
insoluble lead compounds such as Pb 3 (PO 4 ) 2 and Pb 5 (PO 4 ) 3 OH. This
has resulted in a fall in the proportion of households with water lead
> 10 fig \~l from 49% in 1981 to 17% in 1993.42 Despite this improvement, an estimated 13% of infants were still exposed via bottle feeds to
tap water lead concentrations in excess of 10 fig I" 1 and it seems very
unlikely that further treatment of the water supply will be able to
guarantee water lead concentrations < 10 fig l~l.
3.2.3 Cadmium in Irrigation Water. In the 1950s in Japan, many
people, especially menopausal women suffering from malnutrition, low
vitamin D intake and calcium deficiency, suffered a condition known as
Itai-Itai (Ouch-Ouch) disease, with symptoms ranging from lumbagotype pains to multiple fractures of softened bones. The cause was
irrigation water from a river (Jintsu) chronically contaminated with
dissolved cadmium from a zinc mining and smelting operation. Contaminated rice from the irrigated paddy fields was eaten and the Ca 2 + in
bones replaced by Cd 2 + , an ion of the same charge and size.
Although this is a particularly extreme example of acute effects
resulting from high exposure to this non-essential element, concern has
been expressed about chronic effects {e.g. kidney damage, hypertension)
from possible enhanced exposure of humans through increased application of sewage sludge to agricultural land, in view of EC-enforced
cessation of dumping at sea by 1999. Compared with other heavy
metals, cadmium exhibits an especially high mean sludge concentration
(20mgkg~ J ) relative to mean soil concentration (0.4 mg kg" 1 ). 4 3 In
42

G. C. M. Watt, A. Britton, W. H. Gilmour, M. R. Moore, G. D. Murray, S. J. Robertson, and J.
Womersley, Brit. Med.J., 1996, 313, 979.
43
P. O'Neill, 'Environmental Chemistry', Chapman & Hall, 2nd Edn., London, 1993.

acidic soils the concentration OfCd2+ available for uptake by plants can
be substantial,44 as it adsorbs only weakly onto clays, whereas at pH > 7
it readily precipitates, e.g. as CdCO3 for which the solubility product,
KSp = 1.8 x 10~14, is indicative of the advantages of liming. Similarly,
in drinking water, the presence of dissolved carbonate at a concentration
of 5 x 10~4 M can reduce the solubility of cadmium from 637 mg 1~l to
0.11 mg I" 1 , in line with evidence that hard water, with a high calcium
content, can protect against cadmium.45
3.2.4 Selenium in Irrigation Water. In 1983 high rates of embryonic
deformity and death, attributed to selenium toxicosis, were found in wild
aquatic birds at the Kesterson National Wildlife Refuge in the San
Joaquin Valley of California.46 Kesterson Reservoir was a regional
evaporation pond facility where drainage waters, often containing high
levels of salts and trace elements (including selenium), from irrigated
farmland had been collected since 1978.
With evapotranspiration, greatly in excess of precipitation, bringing
soluble salts to the surface of farmland in the arid climate of the westcentral San Joaquin Valley, and crop productivity, after irrigation,
threatened by shallow saline groundwater near the root zone, grids of
sub-surface tile drains were constructed to divert the saline waters to a
collective drain (San Luis), which flowed into Kesterson Reservoir. It
was the geologic setting of the San Joaquin Valley as well as the climate,
however, which led not only to the soil salinization but also to the
presence of selenium (in the form of SeO^ ~~) in Kesterson inflow waters
at concentrations in excess of the US EPA designation of 1000 /ig I" 1 for
selenium as a toxic waste, never mind its 10/igP 1 limit for drinking
water and the <2.3 /ig I" 1 limit subsequently suggested for the protection of aquatic life.
To the west of the San Joaquin River (Figure 5), selenium in the soils is
believed to be of natural origin. During the Jurassic and Cretaceous
periods, there was deposition of marine sediments comprising sandstones, shales, and conglomerates, including seleno-sulfides of iron
(FeS2 + FeSe2). Subsequent uplifting of these sediments produced the
Coast Ranges and subjected the sediments to weathering under oxidizing
conditions. With resultant acid neutralized by the carbonate component
of the sediments, the predominant form of selenium produced from
oxidation of selenide, Se~ n , under alkaline conditions, is selenate,
SeVIO4~, in preference to selenite, SeIVO3~, or biselenite, HSeIVO^".47
44

J. E. Fergusson, 'The Heavy Elements: Chemistry, Environmental Impact and Health Effects',
Pergamon Press, Oxford, 1990.
J. E. Fergusson, 'Inorganic Chemistry and the Earth', Pergamon Press, Oxford, 1982.
46
T. S. Presser, Environ. Management, 1994,18, 437.
47
R. H. Neal, 'Heavy Metals in Soils', ed. B. J. Alloway, Blackie, Glasgow, 2nd Edn., 1995, p. 260.

Sacramento

U033Q 3\fP°40 mg kg" 1 ) and fish, much of the mercury in the latter
being in the form of the highly toxic CH3Hg + . Despite the closure of the

chlor-alkali plant in the late 1980s, mercury is still flowing into the lake
from tributaries, although concentrations of soluble mercury compounds in the lake are declining.69

3.4

Nutrients in Water and Sediments

3.4.1 Phosphorus and Eutrophication. Eutrophication can be considered as the excessive primary production of algae and higher plants
through enrichment of waters by inorganic plant nutrients, usually
nitrogen and phosphorus. The latter, in the form of phosphate, is
normally the limiting nutrient because the amount of biologically
available phosphorus is small in relation to the quantity required for
algal growth.70 Sources of nutrients can be discrete {e.g. specific sewage
outfall) or diffuse {e.g. farmland fertilizers). Eutrophic lakes, highly
productive and often turbid owing to the presence of algae, can be
contrasted with oligotrophic lakes, which exhibit low productivity and
are clear in summer. There have been many examples of unsightly algal
blooms affecting freshwater bodies throughout the world, from Lake
Erie in North America to the Norfolk Broads in East Anglia, England.29
Public concern has increased along with the reported incidences of
toxicity of the bloom-forming organisms, in particular the cyanobacteria
(blue-green algae), which have been implicated in fish fatalities, for
example in Loch Leven, Scotland.71
The chemical form of phosphorus in the water column available for
uptake by biota is important. The biologically available phosphorus is
usually taken to be 'soluble reactive phosphorus (orthophosphate)', i.e.
that which, upon acidification of a water sample, reacts with added
molybdate to yield molybdophosphoric acid which is then reduced with
SnCl2 to the intensely coloured molybdenum blue complex and is
determined spectrophotometrically (Amax = 882 nm).72 Reduction in
inputs of phosphate, for example from point sources or by creating
water meadows and buffer strips to contain diffuse runoff, has obviously
been one of the major approaches to stemming eutrophication trends
and encouraging the restoration of affected lakes. That this has not
always been successful, however, can be attributed in many cases to the
release/recycling of phosphorus previously deposited to and incorporated within the bottom sediments of the lake systems in question.73
70

D . M. Harper, 'Eutrophication of Freshwaters: Principles, Problems and Restoration', Chapman
& Hall, London, 1992.
S. G. Bell and G. A. Codd, 'Issues in Environmental Science and Technology', ed. R. E. Hester
and R. M. Harrison, Royal Society of Chemistry, Cambridge, 1996, Vol. 5, p. 109.
72
J. Murphy and J. R. Riley, Analyt. Chim. Ada, 1962, 27, 31.
73
M. W. Marsden, Freshwater BioL, 1989, 21, 139.
71

The potential mobility and bioavailability of sedimentary phosphorus are to a large extent governed by the chemical associations and
interactions of phosphorus with different sedimentary components.74
Bearing in mind the removal or transport of phosphorus to the
sediments, especially important phases are likely to be 'organic phosphorus' from deposited, dead, decaying biota, and 'sorbed orthophosphate' on inorganic particulates (e.g. iron (hydr)oxides). Many
sequential extraction schemes have been developed to investigate
phosphorus fractionation in lake sediments. Perhaps the most sophisticated is that of Psenner et al.75 who identify, operationally define, and
separate the following fractions: 'labile, loosely bound or adsorbed'
(NH4Cl-extractable); 'reductant-soluble, mainly from iron (hydr)oxide
surfaces' (buffered dithionite-extractable); 'adsorbed to metal oxides
(e.g. Al2O3)' (NaOH-extractable), subsequently distinguishable from
'organic' (also NaOH-extractable); 'apatite-bound' (HCl-extractable);
and 'residual' (persulfate-digestible).
Many such studies of sedimentary phosphorus profiles, also incorporating pore water measurement of soluble reactive phosphate, have
demonstrated that redox-controlled dissolution of iron (hydr)oxides
under reducing conditions at depth releases orthophosphate to solution.
This then diffuses upwards (and downwards) from the pore water
maximum to be re-adsorbed or co-precipitated with oxidized Fe111 in
near-surface oxic sections. The downwards decrease in solid phase
'organic' phosphorus indicates increasing release of phosphorus from
deposited organic matter with depth, some of which will become
associated with hydrous iron and other metal oxides, added to the pool
of mobile phosphorus in pore water or contribute to 'soluble unreactive
phosphorus'. The characteristic reactions involving inorganic phosphorus in the sediments of Toolik Lake, Alaska, are shown in Figure
7.76 If, at depth, the concentrations of Fe 2+ and phosphate are high
enough, authigenic vivianite (Fe3(PO4)2.8H2O) may precipitate out.
With redox control largely responsible for phosphorus mobility in
sediments, what might the consequences of oxygen depletion in the
hypolimnion be? If conditions in the surface sediments are not sufficiently oxidizing to precipitate iron (hydr)oxides and thereby adsorb the
phosphate (i.e. the redox boundary for iron may be in the overlying
water column), the phosphate from previously deposited sediments
would stream off into the water column and promote eutrophication.
This process is called internal loading of phosphorus.
74

J. G. Farmer, A. E. Bailey-Watts, A. Kirika, and C. Scott, Aquat. Conserv., 1994, 4, 45.
R. Psenner, B. Bostrom, M. Dinka, K. Pettersson, R. Pucsko, and M. Sager, Arch. Hydrobiol.
Ergebn. LimnoL, 1988, 30, 98.
76
J. C. Cornwell, Arch. Hydrobiol, 1987, 109, 161.

P concentration

Reaction zone

Solid P

Sediment depth

III
Porewater DRP

III

Figure 7

Characteristic reactions involving phosphorus in the sediments ofToolik Lake,
Alaska. The primary processes controlling porewater phosphorus concentrations
are adsorption to and desorption from iron oxyhydroxides and the precipitation
of authigenic vivianite16
(Reproduced with permission from Arch. Hydrobiol., 109, 161. © 1987
E. Schweizerbart'sche Verlagsbuchhandlung)

Redox control may not be the only process affecting release of
phosphorus from sediments. During the enhanced photosynthesis of
algal blooms, the pH of lake water increases as CO2 is used up and
HCO^" increases. Thus, in summer, both Lough Neagh77 in Northern
Ireland and Lake Glanningen78 in Sweden have shown an increase in
water phosphate concentration. It seems likely that OH ~ is exchanging
with sorbed phosphate in alkaline lakes, thus releasing phosphorus
from association with iron and aluminium (hydr)oxides (see Section
77

B. Rippey, 'Interactions between Sediments and Freshwater', ed. H. L. Golterman, Dr. W. Junk
B. V., The Hague, 1977, p. 349.
S-O. Ryding and C. Forsberg, 'Interactions between Sediments and Freshwater', ed. H. L.
Golterman, Dr. W. Junk B. V., The Hague, 1977, p. 227.

2.4.3). If the waters are calcium-rich, however, this could have the
effect of precipitating phosphate as hydroxyapatite (Cai0(PO4)6(OH)2)
under the high pH conditions prevailing. There can be other factors,
such as temperature, which promote phosphorus release. An increase
in temperature can lead to increased bacterial activity, which increases
oxygen consumption and decreases the redox potential. Turbulence
may also be a factor. For example, wind-induced bottom currents in
shallow lakes could destroy any pH gradients across a buffered
sediment-water interface and re-suspend P-rich sediment in water of
high pH.79
There have been many proposals for restoration of eutrophic lakes.80
For example, 34 options have been put forward for Loch Leven,81
Scotland, with the aim of reducing algal biomass and the incidence of
the bloom-forming cyanobacteria in particular. These strategies fall into
two categories: (i) those aimed at stemming the production of algae in the
first place, including reduction in the supplies of light and nutrients, and
(ii) those aimed at reducing existing algal biomass, including physical
methods such as increased flushing of the loch and harvesting of blooms,
chemical treatment with algicides, and biological methods involving
viruses, parasitic fungi and grazing protozoans, rotifers, and microcrustaceans. Thus far, progress has largely been restricted to reducing
external inputs of phosphorus from point sources to the loch, an
approach which could perhaps be supplemented by the creation of
water meadows and/or buffer strips to reduce inputs from diffuse
runoff. Elsewhere, there was a surge in wetlands construction in the
1980s, using a mixture of plants to clean water contaminated with
nitrates and phosphates.82 One of the largest wetlands restoration
projects is planned for the Florida Everglades, to the south and east of
the rich farmlands around Lake Okeechobee, where runoff waters
enriched in phosphorus from fertilizers have disrupted the flora and
fauna. As water flows through cattails and sawgrass, phosphorus
concentrations are expected to decline from 170 /^g P 1 to 50 /ig I" 1 . It
should be noted that the addition of ferric sulfate or chloride to mop up
phosphate in eutrophic lakes,29 may not work in the long term as there
may be subsequent release of deposited phosphate from sediments under
reducing conditions.
79

L. Hakanson and M. Jansson, 'Principles of Lake Sedimentology', Springer, Berlin, 1983.
A. J. D. Ferguson, M. J. Pearson, and C. S. Reynolds, 'Issues in Environmental Science and
Technology', ed. R. E. Hester and R. M. Harrison, Royal Society of Chemistry, Cambridge, 1996,
Vol. 5, p. 27.
81
A. E. Bailey-Watts, I. D. M. Gunn, and A. Kirika, 'Loch Leven: Past and Current Water Quality
and Options for Change', Report to the Forth River Purification Board, Institute of Freshwater
Ecology, Edinburgh, 1993.
82
P. Young, Environ. Sci. TechnoL, 1996, 30, 292A.
80

A novel use of phosphate to counter acidification has emerged in
recent years.83 Traditional remedial methods which have been adopted
include the direct liming of lakes, e.g. in Sweden,84 or of catchments, e.g.
around Loch Fleet in south-west Scotland.85 Although helpful, such
approaches based on neutralization are costly and usually need to be
repeated at regular intervals. Furthermore, the resulting Ca-rich waters
may turn out to support biota quite different from those found in natural
softwater lakes. An alternative approach has recently been tried by
Davison and co-workers on Seathwaite Tarn, an upland reservoir in the
English Lake District.83 Phosphate fertilizer was added to stimulate
primary productivity and thereby increase the assimilation of nitrate.
This generates base according to the equation:

(86)
As concentrations of nitrate are increasingly high in acid waters, the
addition of modest amounts of phosphate may generate sufficient base to
combat acidity without inducing excessive productivity. An increase in
pH of 0.5 and a marked increase in biological productivity at all levels
were observed over the three-year period of the experiment. In the longer
term, additional quantities of base should be generated by the anoxic
decomposition of organic material accumulating on the lake bed
(through the dissimilative reduction of inorganic oxidants such as
nitrate, sulfate or iron (hydr)oxides present in the sediments—equations
(76) to (79)). If oxygen is the electron acceptor there is no net gain of base
(equation (75)), there is no advantage in adding nitrate because 1 mole of
nitrate is required to generate 1 mole of base, and the generated base
(contributing to alkalinity) should not be confused with the temporary
rise in pH associated with CO2 consumption which affects neither
alkalinity nor acidity.
3,4.2 Nitrate in Groundwater. Principal sources of nitrate in water are
runoff and drainage from land treated with agricultural fertilizers and
also deposition from the atmosphere as a consequence of NO x released
from fossil fuel combustion. The nitrogen present in soil organic matter
may also be released as nitrate through microbial action. Nitrate's role as
a nutrient contributes significantly to blooms of algae, which upon death
83
84
85

W. Davison, D. G. George, and N. J. A. Edwards, Nature, 1995, 377, 504.
P. Nyberg and E. Thornelof, Water, Air, Soil Pollute 1988, 41, 3.
'Restoring Acid Waters: Loch Fleet, 1984-1990', ed. G. Howells and T. R. K. Dalziel, Elsevier,
London, 1992.

are decomposed first by aerobic bacteria, thereby depriving fish and
other organisms of oxygen.29'81
There has long been concern expressed over the presence of nitrate in
drinking water at concentrations exceeding the EC guideline of
50 mg I""1 because of the risk of methaemoglobinaemia (blue baby
syndrome). Here, NO;T is reduced in the baby's stomach to NO2T
which, on absorption, reacts with oxyhaemoglobin to form methaemoglobin. There have also been fears over the reaction of NO^" with
secondary amines from the breakdown of meat or protein to produce
carcinogenic N-nitroso compounds but there is as yet no clear evidence
of a link between stomach cancer and nitrate in water.86 Nevertheless,
there is growing concern over the contamination of groundwater by
nitrate, for example in regions such as the Sierra Pelona Basin, California.87 There the local groundwaters, the major source of drinking water
from private water wells located near each private residence in the rural
communities, often exceed the US EPA maximum contaminant level for
drinking water of lOmgl" 1 (NO3-N). Isotopic investigations, based
upon 15N/14N, have confirmed the predominance of anthropogenic,
organic human and/or animal waste and decay of irrigation-enhanced
vegetation rather than natural nitrate sources.
3.5

Organic Matter and Organic Chemicals in Water

5.5.1 BOD and COD. The solubility of oxygen in water in equilibrium
with the atmosphere at 25 0C is 8.7 mg I" 1 . Causes of oxygen depletion
include decomposition of biomass {e.g. algal blooms) and the presence of
oxidizable substances (e.g. sewage, agricultural runoff, factory effluents)
in the water. The addition of oxidizable pollutants to streams produces a
typical sag in the dissolved oxygen concentration. The degree of oxygen
consumption by microbially mediated oxidation of organic matter in
water is called the Biochemical (or Biological) Oxygen Demand (BOD)
(equation (75)). Another index is the Chemical Oxygen Demand (COD),
which is determined by using the powerful oxidizing agent, dichromate
(Cr2Oy"), to oxidize organic matter,
(87)
followed by back-titration of excess added dichromate with Fe2 + ,
(88)
86

T. M. Addiscott, 'Issues in Environmental Science and Technology', ed. R. E. Hester and R. M.
Harrison, Royal Society of Chemistry, Cambridge, 1996, Vol. 5, p. 1.
87
A. E. Williams, L. J. Lund, J. A. Johnson, and Z. J. Kabala, Environ. Sci. Techno!., 1998, 32, 32.

As Cr2Oy oxidizes substances not oxidized by O2, the COD is usually
greater than the BOD and to some extent overestimates the threat posed
to oxygen content.
3.5.2 Synthetic Organic Chemicals. A large number of organic compounds are synthesized for agricultural use, mainly as pesticides, and for
industrial use in solvents, cleaners, degreasers, petroleum products,
plastics manufacture, etc.ss Many organic micropollutants percolate
into the soil and accumulate in aquifers or surface waters. They can
contaminate drinking water sources via agricultural runoff to surface
waters or percolation into groundwaters, industrial spillages to surface
waters and groundwaters, runoff from roads and paved areas, industrial
waste water effluents leaching from chemically treated surfaces, domestic
sewage effluents, atmospheric fallout, and as leachate from industrial
and domestic landfill sites.88
Pesticides. Many pesticides in aquifers have resisted degradation and
are more likely to persist there because of reduced microbial activity,
absence of light, and lower temperatures.89 Numerous aquifers, for
example in eastern England, have been found to exceed the Maximum
Admissible Concentration (MAC) guidelines for total pesticides in
drinking water (0.5 fig I" 1 ), due primarily to the presence of herbicides
of the carboxy acid and basic triazine groups. Gray88 has listed the 12
pesticides most often found in UK drinking waters in two groups,
frequently occurring (atrazine, chlortoluon, isoproturon, MCPA, mecoprop, simazine) and commonly occurring (2,4-D, dicamba, dichlorprop,
dimethoate, linuron, 2,4,5-T). A major seasonal source of insecticide in
freshwaters is sheep dipping. With organophosphorus compounds,
which themselves replaced the more persistent organochlorine insecticides such as lindane, now falling out of favour because of health risks to
users, the use of synthetic pyrethroids as alternatives has resulted in the
death of aquatic organisms in UK rivers. Pyrethroids, which do not
persist in the environment and are largely non-toxic to mammals, are
toxic to invertebrates and are estimated to be at least 100 times more
toxic in the aquatic environment than organophosphorus pesticides. It
appears to be the pouring of waste dip into holes in the ground that has
caused the problem.90 To protect the aquatic environment, the Environment Agency in England assesses water quality against Environmental
Quality Standards (EQSs). Defined as the concentration of a substance
which must not be exceeded within the aquatic environment in order to
88

N. F. Gray, 'Drinking Water Quality', John Wiley & Sons, Chichester, 1994.
K. R. Eke, A. D. Barnden, and D. J. Tester, 'Issues in Environmental Science and Technology',
ed. R. E. Hester and R. M. Harrison, Royal Society of Chemistry, Cambridge, 1996, Vol. 5, p. 43.
90
F . Pearce, New Scientist, 11 January, 1997, 4.
89

protect it for its recognized uses, EQSs are specific to individual
substances, including pesticides.
Polychlorinated BiphenyIs (PCBs). The Great Lakes, the largest body
of freshwater in the world, with long hydraulic residence times, long
food chains, and multiple sources of PCBs, have been the focal point
of PCB research in aquatic systems. The distribution of PCBs between
the dissolved and particulate phases is dependent on the concentration
of suspended particulate matter, dissolved and particulate organic
carbon concentrations, and the extent to which the system is at
equilibrium. In the 1990s, water concentrations of up to 0.6 ng I" 1
have been observed in the most contaminated lakes (Michigan, Erie,
and Ontario), which may be compared with the US EPA Great Lakes
Water Quality Guidance criteria of 0.017 ng I" 1 . Sediment concentrations peaked in the 1970s, the maximum period of PCB production in
the USA. Following a ban upon their North American production
during the 1970s, there was a significant decline in the PCB concentrations of Great Lakes' fish from the mid-1970s to the mid-1980s but the
rate of decrease has since slowed or stopped for all lakes.66 In common
with other organochlorine compounds, it is the stability, persistence,
volatility, and lipophilicity of PCBs which lead to considerable
biomagnification along the food chain, often far from the place of
release.
Endocrine Disruptors. Concern has recently been expressed over the
possible role of various synthetic organic chemicals {e.g. organochlorine pesticides, PCBs, phthalates, alkylphenolethoxylates, alkylphenols,
etc.) as disruptors of endocrine systems of wildlife and perhaps even of
humans.91 Reproductive changes in male alligators from Lake
Apopka, Florida, embryonic death, deformities, and abnormal
nesting behaviour in fish-eating birds in the Great Lakes region and
the occurrence of hermaphroditic fish near sewage outfalls on some
British rivers have been attributed to postulated oestrogenic or antiandrogenic effects of some of these chemicals.90"92 Most alkylphenolethoxylates (APEs), which are used as detergents, emulsifiers, wetting
agents, and dispersing agents, enter the aquatic environment after
disposal in wastewater. During biodegradation treatment of the latter,
the APEs are transformed into more toxic, short chain ethoxylates,
alkylphenol carboxylic acids, and alkylphenols.93 The threshold concentration of nonylphenol in water for production of the female egg
91

T. Colborn, J. P. Meyers, and D. Dumanoski, 'Our Stolen Future', Little, Brown, Boston, 1996.
M . Lee, Chem. Br., 1996, 32, 5.
93
R. Renner, Environ. Sci. TechnoL, 1997, 31, 316A.

yolk protein, vitellogenin, in male rainbow trout is ~10//gl l.94 It
also appears that nonylphenol concentrations in many European rivers
are up to 10 times higher than those found in US rivers, which are
typically 20 ^gI" 1 , the estimated lower limit of the US EPA
draft drinking water health advisory. As MTBE generally was not
found in shallow urban groundwater, along with benzene, toluene,
ethylbenzene, or xylene, which are commonly associated with petrol
spills, it was concluded that possible sources of MTBE in groundwater
include leaking storage tanks and non-point sources such as recharge
of precipitation and stormwater runoff. As MTBE is thought to be
potentially carcinogenic to humans, its use poses an interesting
dilemma for regulators, given that it helps to reduce carbon monoxide
emissions from cars.

4
4.1

TREATMENT
Purification of Water Supplies

Other than those already discussed in Section 3.1.2, i.e. coagulation of
colloids using A12(SO4)3, and in Section 3.2.2, i.e. pH adjustment using
lime, measures for the purification of drinking water supplies are largely
outwith the scope of this Chapter. The main unit processes are summarised in Table 6.10°
98

M. Cooney, Environ. Sci. TechnoL, 1997, 31, 269A.
P. J. Squillace, J. S. Zogorski, W. G. Wilber, and C. V. Price, Environ. Sci. TechnoL, 1996, 30,
1721.
100
H. Fish, 'Understanding Our Environment: An Introduction to Environmental Chemistry
and Pollution', ed. R. M. Harrison, Royal Society of Chemistry, Cambridge, 2nd Edn., 1992,
p. 53.

Table 6

Summary of treatment processes used in purification of public water
supply

Process

Purpose

Raw water storage (short term)

Sedimentation. Die-off of faecal organisms.
Balancing of intake water quality. Raw
water reserve

Raw water storage (long term)

Oxidation of organic matter. Partial
removal OfNO3", HCO 3 ", PO4", SiO2, by
algal uptake

Chemical precipitation using A12(SO4)3
or activated SiO2, or Fe salts plus
polyelectrolytes

Coagulation, flocculation, and settlement of
turbidity and colour

Microstraining

Straining through very fine mesh rotary
screens

Rapid

filtration

Slow sand

filtration

Rapid up- or down-flow filtration through
sand
Filtration plus bio-oxidation by slow gravity
downward flow

Chlorination and/or ozonation or UV

Disinfection

Softening by lime, lime-soda, or ion
exchange
Activated carbon treatment by powder
addition before filtration or passage
through active carbon filters

Removal of Ca and Mg hardness (no longer
fashionable)
Reduction in residual organic matter

Desalination by flash distillation or
reverse osmosis

Production of freshwater from saltwater or
superpurification of wastewater

Adapted with permission from 'Understanding Our Environment', 2nd Edn., p. 85. © 1992 The
Royal Society of Chemistry.

4.2

Waste Treatment

Methods for the conventional primary, secondary, and tertiary treatment of wastewaters and sewage are listed in Table 7. Less conventional
treatments, such as constructed wetlands where the localized diffusion of
oxygen and release of organic nutrients by the roots of mixed plants
enables fast degradative activity by both aerobic and anaerobic organisms as well as plant assimilation of pollutants, have been introduced in
recent years.101 In addition, reedbed systems employing a single plant
species, most commonly Phragmites australis in Europe, are becoming
increasingly popular, especially for industrial applications, as they are
simple to operate and cheap to maintain.101
101

R. Cobban, D. Gregson, and P. Phillips, Chem. Br., 1998, 34, 40.

Table 7

Conventional methods of waste treatment

Treatment stage

Method

Preliminary treatment

Coarse solid and grit screening

Primary treatment

Suspended solids sedimentation

Secondary treatment—relies on microorganisms extracting and chemically
transforming nutrients from primary
effluents

Anaerobic and aerobic digestion
Activated sludge processes

Tertiary treatment—chemically,
physically, or biologically removing
nutrients, including phosphorus and
nitrogen. Because these methods
necessitate removing minutely sized
trace quantities of compounds from
large volumes of waste water they are
usually more expensive

Membrane filtration—removing particles
that are too small for ordinary filtration,
Generally this involves the pressurized flow of
liquid across a membrane. There are three
types of membrane filtration systems, which
differ mainly in the size of particles they
remove.
(i) Microfiltration retains and concentrates
particles the size of paint pigments and
bacteria
(ii) Ultrafiltration is used to reduce the
biochemical oxygen demand of
wastewater by removing substances such
as sugars, fats, oils, and greases
(iii) Reverse osmosis removes the smallest
particles and can be used to retain
substances such as dissolved salts or
metal ions
Ion exchange resins—used to exchange ions
from solution
Extended aeration with microbial uptake—
aerobic micro-organisms are used to digest
the last remaining nutrients found in
wastewater
Precipitation with iron or aluminium—
mixing wastewater with iron or aluminium
sulfates causes a chemical reaction forcing
pollutants to come out of solution so that they
can be collected by sedimentation or filtration

Questions

1. Calculate the ionic strength of a solution which is 0.01 M with
respect to potassium sulfate and 0.002 M with respect to magnesium
chloride. Use this value of ionic strength and the Guntleberg
approximation to determine the mixed acidity constant, K\ for
methanoic acid (K = 10~ 375 ).

2. Construct a plot to show how — log {activity} varies with pH for all
species in one litre of groundwater containing 3 x 10~~4 moles of 4chlorophenol {K = 10~ 918 ). Annotate the plot to show the equilibrium pH of the groundwater.
3. Calculate the equilibrium pH of rainwater in equilibrium with
CO2(g) at a partial pressure of 0.00035 atm. (Hint: the charge
balance equation can be approximated by (H + } = (HCO^"})

4. Construct a pe-pH diagram for arsenic species present in natural
waters. Assume that sulfur species are absent and that the total
dissolved concentration of arsenic is 10~6 M. From the diagram,
predict the major arsenic species in:
groundwaters
alkaline surface waters

5. Draw, describe, and compare the concentration profiles of the
chemical species of manganese, iron, arsenic, and phosphorus
likely to be found in the solid and solution phases of unmixed
sediments at the bottom of a seasonally anoxic, eutrophic freshwater lake during (i) summer and (ii) winter.
6. With respect to human health and ecological impact, discuss the
short and long term consequences of the products of the 20th
Century synthetic organic chemical industry upon the quality of
surface waters and groundwaters.

7. Explain the chemistry underlying (i) the major detrimental effects
of metal mining upon rivers and groundwaters and (ii) associated
preventive or remedial measures.
8. Critically discuss the effectiveness of measures to counter the effects
of (i) acidification and (ii) eutrophication in freshwater lakes.
FURTHER READING
E. K. Berner and R. A. Berner, 'Global Environment: Water, Air and
Geochemical Cycles', Prentice-Hall, Englewood Cliffs, NJ, 1996.
'Freshwater Quality: Defining the Indefinable?', ed. P. J. Boon and D. L.
Ho well, The Stationery Office, Edinburgh, 1997.
J. I. Drever, 'The Geochemistry of Natural Waters', Prentice-Hall,
Englewood Cliffs, NJ, 1997.
U. Forstner, 'Contaminated Sediments', Springer, Berlin, 1989.
A. J. Home, 'Limnology', McGraw-Hill, New York, 2nd Edn., 1994.
A. G. Howard, 'Aquatic Environmental Chemistry', Oxford University
Press, Oxford, 1998.
D. Langmuir, 'Aqueous Environmental Geochemistry', Prentice-Hall,
Englewood Cliffs, NJ, 1997.
'The Fresh Waters of Scotland', ed. P. S. Maitland, P. J. Boon, and D. S.
McLusky, John Wiley & Sons, Chichester, 1994.
C. F. Mason, 'Biology of Freshwater Pollution', Longman, 3rd Edn.,
1996.
'Groundwater Contaminants and their Migration', ed. J. Mather, D.
Banks, S. Dumpleton and M. Fermor, Geological Society, London,
Special Publication No. 128, 1998.
F. M. Morel and J. G. Hering, 'Principles and Applications of Aquatic
Chemistry', John Wiley & Sons, New York, 1993.
J. Pankow, 'Aquatic Chemistry Concepts', Lewis, Michigan, 1991.
W. Salomons and U. Forstner, 'Metals in the Hydrocycle', Springer,
Berlin, 1984.
W. Stumm, 'Chemistry of the Solid-Water Interface', John Wiley &
Sons, New York, 1992.
W. Stumm and J. J. Morgan, 'Aquatic Chemistry', John Wiley & Sons,
3rd Edn., New York, 1996.

CHAPTER 4

The Oceanic Environment
STEPHEN J. DE MORA

1 INTRODUCTION
The World Ocean is a complex mixture containing all the elements, albeit
in minute amounts in some cases. Seawater contains dissolved gases and,
apart from some exceptional environments, is consequently both well
oxygenated and buffered at a pH of about 8. There are electrolytic salts,
the ionic strength of seawater being approximately 0.7, and multitudinous organic compounds in solution. At the same time, there is a
wide range of inorganic and organic particles in suspension. These
comfortable distinctions become quite confused in seawater because
some molecules present in true solution are sufficiently large to be
retained by a filter. Moreover, surface adsorption allows particles to
scavenge dissolved elements and accumulate coatings of organic material
from solution. Some elements, particularly those with biochemical
functions, may be rapidly removed from solution. Concurrently, reactions involving geological time scales are proceeding slowly. Yet despite
this apparent complexity, many aspects of the composition of seawater
and chemical oceanography can now be explained with recourse to the
fundamental principles of chemistry. This chapter serves to bridge the
gap between those with environmental expertise and those with a
traditional chemical background.
1.1

The Ocean as a Biogeochemical Environment

A traditional approach utilized in geochemistry, and now also in
environmental chemistry, is to consider the system under investigation a
reservoir. For a given component, the reservoir has sources (inputs) and
sinks (outputs). The system is said to be at equilibrium, or operating
under steady-state conditions, when a mass balance between inputs and

outputs is achieved. An imbalance could signify that an important source
or sink has been ignored. Alternatively, the system may be perturbed,
possibly anthropogenically mediated, and therefore be changing toward
a new equilibrium state.
Processes within the reservoir that affect the temporal and spatial
distribution of a given component are transportation and transformations. Both physics and biology within the system play a role. Clearly,
transport effects are dominated by the hydrodynamic regime. Although
transformations could involve chemical (dissolution, redox reactions,
speciation changes) or geological (sedimentation) processes, biological
activity can control nutrient and trace metal distributions. Furthermore,
the biota influences concentrations of O2 and CO2 which in turn
determine the pH and ps (i.e. the redox potential see p. 157), respectively.
For these reasons, some fundamental aspects of descriptive physical and
biological oceanography are included in this chapter.
In terms of biogeochemical cycling, the ocean constitutes a large
reservoir. The surface area is 361.11 x 106 km2, encompassing nearly
71 % of the earth's surface. The average depth is 3.7 km, but depths in the
submarine trenches can exceed 10 km. The ocean contains about 97% of
the water in the global hydrological cycle. A schematic representation of
the ocean reservoir is presented in Figure 1. The material within it can be
operationally defined, usually based on filtration, as dissolved or
particulate. The ocean is divided into two layers, with distinct surface
and deep waters. The boundary regions are also distinguished, as the
composition in these regions can be quite different from that of bulk
seawater. Furthermore, interactions within these environments can alter
the mass transfers across the boundary. The rationale for such features
will be presented in subsequent sections.
Material supplied to the ocean originates from the atmosphere, rivers,
glaciers, and hydrothermal waters. The relative importance of these
pathways depends upon the component considered and geographic
location. River runoff commonly constitutes the most important source.
Transported material may be either dissolved or particulate, but discharges are into surface waters and confined to coastal regions. Hydrothermal waters are released from vents on the sea floor. Such
hydrothermal waters are formed when seawater circulates into the
fissured rock matrix and, under conditions of elevated temperature and
pressure, compositional changes in the aqueous phase occur due to
seawater-rock interactions. This is an important source of some elements, such as Li, Rb, and Mn. The atmosphere supplies particulate
material globally to the surface of the ocean. In recent years, this has
been the most prominent pathway to the world ocean for Pb, identified
by its isotopic signature as originating from petrol additives. Wind-borne

ATMOSPHERIC
INPUT

air / sea
interface
surface
ocean

RIVER
INPUT

GLACIAL
INPUT

river / sea
interface

SEA WATER
deep
ocean

HYDROTHERMAL
INPUTS AND
OUTPUTS

Sediment
water
interface

SEDIMENT
BASEMENT ROCK

Figure 1 A schematic representation of the ocean reservoir. The source and sink fluxes are
designated as g and n, referring to gross and net fluxes, thereby indicating that
interactions within the boundary regions can modify the mass transfer. Within
seawater, the p ^ d term signifies that substances can undergo paniculatedissolved interactions. However, it must be appreciated that several
transportation and transformation processes might be operative

(From Chester1)

transport is greatest in low latitudes and the Sahara Desert is known to
act as an important source of dust. In contrast, glacial activity makes
little impact on the world ocean. Glacier-derived material tends to
comprise physically weathered rock residue, which is relatively insoluble.
In addition, the input is largely confined to polar regions, with Antarctica responsible for approximately 90% of the material.
Sedimentation acts as the major removal mechanism. This is essentially a geological process in coastal environments but is biologically
mediated in the open sea through the sinking of shells of microorganisms and faecal pellets. However, volatilization and subsequent
1

R. Chester, 'Marine Geochemistry', Unwin Hyman, London, 1990, p. 698.

evasion to the atmosphere can be important for elements such as Se and
Hg that undergo bioalkylation.
Some definitions facilitate the interpretation of chemical phenomena
in the ocean. Conservative behaviour signifies that the concentration of a
constituent (or absolute magnitude of a property) varies only due to
mixing processes. Components or parameters that behave in this manner
can be used as conservative indices of mixing. Examples are salinity and
potential temperature, the definitions for which are presented in subsequent sections. In contrast, non-conservative behaviour indicates that
the concentration of a constituent may vary as a result of biological or
chemical processes. Examples of parameters that behave non-conservatively are dissolved oxygen and pH. Residence time, T, is defined as:

where A is the total amount of constituent A in the reservoir and dA/dT
can be either in rate of supply or the rate of removal of A. This represents
the average lifetime of the component in the system and is, in effect, a
reciprocal rate constant (see Chapter 6). Finally, the photic zone refers to
the upper surface of the ocean in which photosynthesis can occur. This is
typically taken to be the layer down to the depth at which sunlight
radiation has declined to 1% of the magnitude at the surface. This might
typically be > 100 m for visible light or photosynthetically active radiation (PAR) but generally < 20 m for UV wavelengths, of recent interest
due to the enhanced input invoked by stratospheric ozone depletion (see
Chapter 2).
1.2

Properties of Water and Seawater

Water is a unique substance, with unusual attributes because of its
structure. The molecule consists of a central oxygen atom with two
attached hydrogen atoms forming a bond angle of about 105°. As
oxygen is more electronegative than hydrogen, it attracts the shared
electrons to a greater extent. In addition, the oxygen atom has a pair of
lone orbitals. The overall effect produces a molecule with a strong dipole
moment, having distinct negative (O) and positive (H) ends. While there
are several important consequences, two will be considered here. Firstly,
the positive H atoms of one molecule are attracted towards the negative
O atom in adjacent molecules giving rise to hydrogen bonding. This has
important implications with respect to a number of physical properties,
especially those relating to thermal characteristics. Secondly, the large
dipole moment ensures that water is a very polar solvent.

Considering firstly the physical properties, water has much higher
freezing and melting points than would be expected for a molecule of
molecular weight 18. Water has high latent heats of evaporation and
fusion. This means that considerable energy is required to stimulate
phase changes, the energy being utilized in hydrogen bond rupturing.
Moreover, it has a high specific heat and is a good conductor of heat.
Consequently, heat transfer in water by advection and conduction gives
rise to uniform temperatures. The density of pure water exhibits
anomalous behaviour. In ice, O atoms have 4 H atoms orientated
about them tetrahedrally. These units are packed together with a
hexagonal symmetry. At the freezing point, 0 0C, ice is less dense than
water. Heating breaks hydrogen bonds, and molecules can achieve
slightly closer packing, which causes the density to increase. The
maximum density occurs at 4 0C because at higher temperatures
thermal expansion compensates for this compression effect. As will be
discussed later, seawater differs in this respect. Thus, fresh ice floats on
water, which in part explains how rivers and lakes can freeze over but
remain liquid at depth. With respect to other properties, water has a
high surface tension that is manifest in stable droplet formation and
has a relatively low molecular viscosity and therefore is quite a mobile
fluid.
Water is an excellent solvent. It is extremely polar and can dissolve a
wider range of solutes and in greater amounts than any other substance.
Water has a very high dielectric constant, a measure of the solvent's
ability to keep apart oppositely charged ions. The solvating characteristics of individual ions influence their behaviour in solution, i.e. in terms
of hydration, hydrolysis, and precipitation. Although water exhibits
amphoteric behaviour, electrolytic dissociation is quite small. Furthermore, dissociation gives equal ion concentrations to both H 3 O + and
OH ~ and so pure water is neutral. The amphoteric behaviour enhances
dissolution of introduced particulate matter through surface hydrolysis
reactions.
While the concept will be considered in detail below, the term salinity
(5%o) is introduced here as a measure of the salt content of seawater, a
typical value for oceanic waters being 35 g kg" 1 . In an oceanographic
context, the most important consequence of the addition of salt to water
is the effect on density. However, many of the characteristics outlined
above are also altered. The addition of electrolytes can cause a small
increase in the surface tension. This effect is not commonly observed in
seawater due to the presence of surfactants, which decrease the surface
tension and so facilitate foam formation. As illustrated in Figure 2, the
presence of salt does depress the temperature of maximum density and
the freezing point of the solution relative to pure water. Thus, seawater

Temperature, 0C
Figure 2

Salinity, g%<>
The temperature of maximum density (—) and freezing point (
as a function of dissolved salt content
(From Tchernia2)

) ofseawater

with a typical salt content of 35 g kg" l freezes at approximately — 1.9 0C
and the resulting ice is more dense than the solution. However, more
often than the formation of sea ice itself, the freezing process tends to
produce fresh ice overlying a more concentrated brine solution. Salts
can be precipitated at much lower temperatures, i.e. mirabilite
(Na2SO4.2H2O) at -8.2 0C and halite (NaCl) at - 2 3 0C. Some brine
inclusions and salt crystals can become incorporated into the ice.
From an oceanographic perspective, the fundamental properties of
seawater are temperature, salinity, and pressure (i.e. depth dependence).
Together, these parameters control the density of the water, which in
turn determines the buoyancy of the water and pressure gradients. Small
density differences integrated over oceanic scales cause considerable
pressure gradients and result in currents.
Surface water temperatures are extremely variable, obviously influenced by location and season. The minimum temperature found in polar
latitudes approaches —20C. Equatorial oceanic waters can reach 300C.
Temperature variations with depth are far from consistent. A region in
which mixing is prevalent, as observed especially in the surface waters,
2

P . Tchernia, 'Descriptive Regional Oceanography', Pergamon Press, Oxford, 1980, p. 253.

Figure 3

The distribution of mean annual salinity in the surface waters of the ocean
(From The Open University3)

produces a layer in which the temperature is relatively constant. The
zone immediately beneath normally exhibits a sharp change in temperature, known as the thermocline. The thermocline in the ocean extends
down to about 1000 m within equatorial and temperate latitudes. It acts
as an important boundary in the ocean, separating the surface and deep
layers and limiting mixing between these two reservoirs.
Below the thermocline, the temperature changes only little with depth.
The temperature in seawater is non-conservative because adiabatic
compression causes a slight increase in the in situ temperature measured
at depth. For instance in the Mindanao Trench in the Pacific Ocean, the
temperatures at 8500 m and 10 000 m are 2.23 0C and 2.48 0C, respectively. The term potential temperature is defined to be the temperature
that the water parcel would have if raised adiabatically to the ocean
surface. For the examples above, the potential temperatures are 1.22 0C
and 1.160C, respectively. Potential temperature is a conservative index.
Salinity in the surface waters in the open ocean ranges between 33 and
37 (Figure 3), the main control being the balance between evaporation
and precipitation. The highest salinities occur in regional seas where the
evaporation rate is extremely high, namely the Mediterranean Sea (3839) and the Red Sea (40-41). Within the World Ocean, the salinity is
3

The Open University, 'Seawater: its Composition, Properties and Behaviour', Open University &
Pergamon Press, Oxford, 1989, p. 165.

greatest in latitudes of about 20° where the evaporation exceeds
precipitation. Lower salinities occur poleward as evaporation diminishes
and near the equator where precipitation is very high. Local effects can
be important, as evident in the vicinity of large riverine discharges that
dilute the salinity. Salinity variations with depth are related to the origin
of the deep waters and so will be considered in the section on oceanic
circulation. A zone in which the salinity exhibits a marked gradient is
known as a halocline.
Whereas the density of pure water is 1.000 g ml" 1 , the density of
seawater (S%o = 35) is about 1.03 g ml" 1 . The term 'sigma-tee', ah is
used to denote the density (actually the specific gravity and hence a
dimensionless number) of water at atmospheric pressure based on
temperature and salinity in situ. Density increases, and so the buoyancy
decreases, with an increase in ot. It is defined as:
ot = (specific gravityS%O,T - 1) x 1000
In a plot of temperature against salinity (a T-S diagram), constant at
appear as curved lines which denote waters of constant pressure and are
known as isopycnals. A zone in which the pressure changes greatly is
known as a pycnocline. Within the water column, a pycnocline therefore
separates waters with distinctive temperature and salinity characteristics,
usually indicative of different origins. A T-S diagram can also be used to
estimate the properties resulting from the mixing of two water masses. As
noted above, the temperature is not a conservative property, and therefore at is also non-conservative. To circumvent the associated difficulties
of interpretation, an analogous term known as the potential density, a Q9
is defined on the basis of potential temperature instead of in situ
temperature. The oe is therefore a conservative index.
1.3

Salinity Concepts

Salinity is a measure of the salt content of seawater. Developments in
analytical chemistry have led to an historical evolution of the salinity
concept. Intrinsically it would seem to be a relatively straightforward
task to measure. This is true for imprecise determinations that can be
quickly performed using hand-held refractometers. The salinity affects
seawater density and thus the impetus for high precision in salinity
measurements came from physical oceanographers.
The first techniques utilized for the determination of salinity, involving
the gravimetric analysis of salt left after evaporating seawater to dryness,
were fraught with difficulties. Variable amounts of water of crystallization might be retained. Some salts, such as MgCl2, can decompose

leaving residues of uncertain composition. Other constituents, especially
organic material, might be volatilized or oxidized. Overall, such methods
led to considerable inconsistencies and inaccuracies. The second set of
procedures for salinity measurement made use of the observation from
the Challenger expedition of 1872-76 that sea salt composition was
apparently invariant. Hence, the total salt content could be calculated
from any individual constituent, such as Cl"" that could be readily
determined by titration with Ag+ . At the turn of the century, Knudsen
defined salinity to be the weight in grams of dissolved inorganic matter
contained in 1 kg of seawater, after bromides and iodides were replaced
by an equivalent amount of chloride and carbonate was converted to
oxide. Clearly from the adopted definition, the analytical technique was
not specific to Cl" and so the term chlorinity was introduced. Chlorinity
(Cl%o) is the chloride concentration in seawater, expressed as g kg" 1 , as
measured by Ag + titration {i.e. ignoring other halide contributions by
assuming C P to be the only reactant). The relationship of interest was
that between £%> and Cl%o, as given as:

As a calibrant solution for the AgNO3 titrant, Standard Seawater was
prepared that had certified values for both chlorinity and salinity.
Unfortunately, the above salinity-chlorinity relationship was derived
based on only nine seawater samples that were somewhat atypical. It has
since been redefined using a much larger set of samples representative of
oceanic waters to become:

The third category of salinity methodologies was based on conductometry, as the conductivity of a solution is proportional to the total salt
content. Standard Seawater, now also certified with respect to conductivity, provides the appropriate calibrant solution. The conductivity of a
sample is measured relative to the standard and converted to salinity in
practical salinity units (psu). Note that although psu has replaced the
outmoded %o, usually units are ignored altogether in modern usage.
These techniques continue to be the most widely used methods because
conductivity measurements can provide salinity values with a precision
of ±0.001 psu. Highly precise determinations require temperature
control of samples and standards to within ±0.0010C. Application of a
non-specific technique like conductometry relies upon the assumption
that the sea salt matrix is invariant, both spatially and temporally. Thus,
the technique cannot be reliably employed in marine boundary environ-

ments where the seawater composition differs from the bulk characteristics.
There are two types of conductometric procedures commonly used.
Firstly, a Wheatstone Bridge circuit can be set up whereby the ratio of
the resistance of unknown seawater to standard seawater balances the
ratio of a fixed resistor to a variable resistor. The system uses alternating
current to minimise electrode fouling. Alternatively, the conductivity can
be measured by magnetic induction, in which case the sensor consists of a
plastic tube containing sample seawater that links two transformers. An
oscillator establishes a current in one transformer that induces current
flow within the tube, the magnitude of which depends upon the salinity
of the sample. This in turn induces a current in the second transformer,
which can then be measured. This design has been exploited for in situ
conductivity measurements.
1.4

Oceanic Circulation

The distribution of components within the ocean is determined by both
transportation and transformation processes. A brief outline of oceanic
circulation is necessary to ascertain the relative influences. Two main
flow systems must be considered. Surface circulation is established by
tides and the prevailing wind patterns, and deep circulation is determined
by gravitational forces. Both are modified by Coriolis force, the acceleration due to the earth's rotation. It acts to deflect moving fluids {i.e.
both air and water) to the right in the northern hemisphere and to the left
in the southern hemisphere. The magnitude of the effect is a function of
latitude, being nil at the equator and increasing poleward.
Surface oceanic circulation is depicted in Figure 4. For the most part,
the circulation patterns describe gyres constrained by the continental
boundaries. The prevailing winds acting under the influence of Coriolis
force result in clockwise and counter-clockwise flow in the northern and
southern hemispheres, respectively. The flow fields are non-uniform,
exhibiting faster currents along the western margins. These are manifest,
for example, as the Gulf Stream, Kuroshio Current, and Brazil Current.
Circulation within the Indian Ocean is exceptional in that there are
distinct seasonal variations in accord with the monsoons. The absence of
other continents within the immediate boundary region of Antarctica
gives rise to a circumpolar current within the Southern Ocean.
The surface circulation is restricted to the upper layer influenced by the
wind, typically about 100 m. However, underlying water can be transported up into this zone when horizontal advection is insufficient to
maintain the superimposed flow fields. This process is called upwelling
and is of considerable importance in that biochemical respiration of

North Equatorial
North Equator.il CounterJ
South Equatorial Counter
South Equatorial

. Equatorial Counter

Nq.ua
Eto
qu
torial
E
ian
late
Coru
r

^Antarctica CircumpolJ' or "West Wind DnIt

Figure 4

The surface circulation in the ocean
(From Stowe4)

organic material at depth ensures that the ascending water is nutrientrich. Upwelling occurs in the eastern oceanic boundaries where longshore winds result in the offshore transport of the surface water.
Examples are found off the coasts of Peru and West Africa. Similar
processes cause upwelling off Arabia, but this is seasonal due to the
monsoon effect. A divergence is a zone in which the flow fields separate.
In such a case, upwelling may result as observed in the equatorial Pacific.
It should be noted that a region in which the streamlines come together is
known as a convergence, and water sinks in this zone.
The density of the water controls the deep circulation. If the density of
a water body increases, it has a tendency to sink. Subsequently it will
spread out over a horizon of uniform OQ. AS the density can be raised by
either an increase in the salinity or a decrease in the temperature, the
deep water circulatory system is also known as thermohaline circulation.
The densest waters are formed in polar regions due to the relatively low
temperatures and the salinity increase that results from ice formation.
Antarctic Bottom Water (ABW) is generated in the Weddell Sea and
flows northward into the South Atlantic. North Atlantic Deep Water
(NADW) is formed in the Norwegian Sea and off the southern coast of
Greenland. The flow of the NADW can be traced southwards through
the Atlantic Ocean to Antarctica where it is diverted eastward into the
Southern Indian Ocean and South Pacific. There it heads northwards
4

K . S. Stowe, 'Ocean Science', John Wiley & Sons, New York, 1979.

and either enters the North Pacific or becomes mixed upward into the
surface layer in the equatorial region. The transit time is of the order of
1000 years. As noted previously, the thermocline acts as an effective
barrier against mixing of dissolved components in the ocean. Consequently, this deep water formation process in high latitudes is important
because it facilitates the relatively rapid transport of material from the
surface of the ocean down to great depths. The deep advection of
atmospherically derived CO2 is a pertinent example.
Diverse processes can form intermediate waters within the water
column. In the southern South Atlantic, the NADW overrides the
denser ABW. Antarctic Intermediate Water results from water sinking
along the Antarctic Convergence (~ 50 0S). Relatively warm, saline
water exits the Mediterranean Sea at depth and can be identified as a
distinctive layer within the North and South Atlantic.
2
2.1

SEAWATER COMPOSITION AND CHEMISTRY
Major Constituents

The major constituents in seawater are conventionally taken to be those
elements present in typical oceanic water of salinity 35 that have a
concentration greater than 1 mg kg" 1 , excluding Si which is an important nutrient in the marine environment. The concentrations and main
species of these elements are presented in Table 1. One of the most
significant observations from the Challenger expedition of 1872-76 was
that these major components existed in constant relative amounts. As
already explained, this feature was exploited for salinity determinations.
Inter-element ratios are generally constant, and often expressed as a ratio
to Cl%o as shown in Table 1. This implies conservative behaviour, with
concentrations depending solely upon mixing processes, and indeed,
salinity itself is a conservative index.
Because of this behaviour, individual seawater constituents can be
utilized for source apportionment studies in non-marine environments.
For instance, an enrichment factor (EF) for a substance X is defined as:
sample
'seawater

An enrichment factor of 1 indicates that the substance exists in comparable relative amounts in the sample and in seawater, thereby giving a
good indication of a marine origin. If EFx > 1, then it is enriched with
respect to seawater. Conversely, depletion is signified when values
EFx < 1. Another example of the application of inter-element ratios

Table 1

Chemical species and concentrations of the major elements in seawater
Concentration for S = 35

Element

Chemical species

(mol 1"x)

Na
Mg
Ca
K
Sr
Cl
S
C (inorganic)
Br
B
F

Na +
Mg 2+
Ca^ +
K+
Sr 2+
CP
SO2T, NaSO4HCO3", CO2T
Br"
B(OH)3, B(OH)4"
F " , MgF +

4.79
5.44
1.05
1.05
9.51
5.59
2.89
2.35
8.62
4.21
7.51

x
x
x
x
x
x
x
x
x
x
x

10" 1
10" 2
10~2
10" 2
10~ 5
10" 1
10" 2
10" 3
10" 4
10" 4
10" 5

(g kg" ! )

Ratio to chlorinity
(Cl - 19.374%o)

10.77
1.29
0.4123
0.3991
0.00814
19.353
0.905
0.276
0.673
0.0445
0.00139

5.56
6.66
2.13
2.06
4.20
9.99
4.67
1.42
3.47
2.30
7.17

x
x
x
x
x
x
x
x
x
x
x

10" 1
10~2
10~2
10~2
10~4
10" 1
10~2
10~2
10~ 3
10~ 3
10~5

Based on Dyrssen and Wedborg 5

can be found in examining the geochemical cycle of sulfur. Concentrations of SC>4~ and Na + in ice cores and marine aerosols exhibit a
SC>4~ :Na + greater than that observed in seawater. This excess can be
readily calculated and is known as non-sea salt sulfate (NSSS). Contributions to NSSS include SO2 derived from both volcanic and anthropogenic sources, together with dimethyl sulfide (DMS) of marine
biogenic origin.
Not all the major constituents consistently exhibit conservative behaviour in the ocean. The most notable departures occur in deep waters
where Ca2 + and HCO^ exhibit anomalously high concentrations due to
the dissolution of calcite. The concept of relative constant composition
does not apply in a number of atypical environments associated with
boundary regions. Inter-element ratios for major constituents can be
quite different in estuaries and near hydrothermal vents. Obviously,
these are not solutions of sea salt (with the implication that accuracy of
salinity measurements by chemical and conductometric means is
limited).
The residence times for some elements are presented in Table 2. The
major constituents normally have long residence times. The residence
time is a crude measure of a constituent's reactivity in the reservoir. The
aqueous behaviour and rank ordering can be appreciated simply in terms
of the ionic potential given by the ratio of electronic charge to ionic
radius (Z/r). Elements with Zjr < 3 are strongly cationic. The positive
charge density is relatively diffuse, but sufficient to attract an envelope of
water molecules forming a hydrated cation. As the ionic potential
5

D . Dyrssen and M. Wedborg, T h e Sea', ed. E. Goldberg, John Wiley & Sons, New York, 1974, p.
181.

Table 2

The residence time and speciation of some elements in the ocean

Element

Principal species

Concentration
(moll" 1 )

Li
B
F
Na
Mg
Al
Si
P
Cl
K
Ca
Sc
Ti
V
Cr
Mn
Fe
Co
Ni
Cu
Zn
Br
Sr
Ba
La
Hg
Pb
Th
U

Li +
B(OH)35B(OK)4F ~ , MgF +
Na+
Mg 2 +
Al(OH)4-, Al(OH)3
Si(OH)4
HPO 4 ", P O 4 " , MgHPO 4
Cl"
K+
Ca 2 +
Sc(OH)3
Ti(OH)4
H 2 VO 4 -, HVO 2 ,-,NaVO 4 CrO^", NaCrO 4 Mn 2 + , MnCl +
Fe(OH) 3
Co 2 + , CoCO 3 , CoCl +
Ni 2 + , NiCO 3 , NiCl +
CuCO 3 , CuOH + , Cu 2 +
ZnOH + , Zn2 + , ZnCO 3
Br"
Sr 2+
Ba 2+
La3 + , LaCO 3 + , LaCl 2+
HgClJPbCO 3 , Pb(CO 3 )I-,PbCl +
Th(OH) 4
UO2(CO3)4T

2.6 x
4.1 x
6.8 x
4.68 x
5.32 x
7.4 x
7.1 x
2 x
5.46 x
1.02 x
1.02 x
1.3 x
2 x
5 x
5.7 x
3.6 x
3.5 x
8 x
2.8 x
8 x
7.6 x
8.4 x
9.1 x
1.5 x
2 x
1.5 x
2 x
4 x
1.4 x

10~ 5
1(T 4
10" 5
10" l
10" 2
10" 8
10~ 5
10" 6
10" 1
10- 2
10" 2
10"n
10" 8
10~ 8
10" 9
10~ 9
10~ 8
10" 1 0
10~ 8
10" 9
10~ 8
10" 4
10~ 5
10~ 7
10" 11
10" 1 0
10" 1 0
10"u
10~ 8

Residence time
(yr)
2.3 x
1.3 x
5.2 x
6.8 x
1.2 x
1.0 x
1.8 x
1.8 x
Ix
7 x
1x
4 x
1.3 x
8 x
6 x
1x
2 x
3 x
9 x
2 x
2 x
1x
4 x
4 x
6 x
8 x
4 x
2 x
3 x

106
107
105
107
107
102
104
105
108
106
106
104
104
104
103
104
102
104
104
104
104
108
106
104
102
104
102
102
106

Based on Brewer6 and Bruland.7

increases, the force of attraction towards the water similarly rises to the
extent that one oxygen-hydrogen bond in the molecule breaks. This
causes the solution pH to fall and metal hydroxides to form. Neutral
hydroxides tend to be relatively insoluble and so precipitate. However, in
the more extreme case for which Z/r > 12, the attraction toward the
oxygen is so great that both bonds in the associated water molecules are
broken. The reaction product is an oxyanion, usually quite soluble
because of the associated anionic charge. Thus in seawater, those
elements (Al, Fe) having a tendency to form insoluble hydroxides have
short residence times. This is also true for elements that exist preferen6

P . Brewer, 'Chemical Oceanography', ed. J. P. Riley and G. Skirrow, Academic Press, London, 2nd
Edn., 1975,VoI. l,p.415.
7
K. W. Bruland, 'Chemical Oceanography', ed. J. P. Riley and R. Chester, Academic Press,
London, 1983, Vol. 8, p. 157.

tially as neutral oxides (Mn, Ti). Hydrated cations (Na + , Ca2 + ) and
strongly anionic species (CP, Br", UO2(CO3)2~) have long residence
times. This treatment is, of course, somewhat of an oversimplification,
ignoring the rather significant role that biological organisms play in
nutrient and trace element chemistry.
2.2

Dissolved Gases

2.2.7 Gas Solubility and Air-Sea Exchange Processes. The ocean
contains a vast array of dissolved gases. Some of the gases such as Ar
and chlorofluorocarbons behave conservatively and can be utilized as
tracers for water mass movements and ventilation rates. Equilibrium
processes at the air-sea interface generally lead to saturation, and then
the concentration remains unchanged once the water sinks. Thus, the gas
concentration is characteristic of the lost contact with the atmosphere.
Deep waters usually contain no CFCs as such anthropogenic compounds
have only a recent history of use. There are several important nonconservative gases, which exhibit wide variations in concentration due to
biological activity. O2 determines the redox potential in seawater and
CO2 buffers the ocean at pH 8. Ocean-atmosphere exchange processes
for gases such as CO2 and dimethyl sulfide may play an important role in
climate change.
Both temperature and salinity affect the solubility of gases in water.
Empirical relationships can be found elsewhere.8'9 The trends are such
that gas solubility increases with a decrease in temperature or an increase
in salinity. The changes in solubility are non-linear and differ dramatically for various gases. Figure 5 depicts the solubility of several gases as a
function of temperature.
At the ocean-atmosphere interface, exchange of gases occurs to
achieve equilibrium between the two systems and, consequently, gases
become saturated. However, supersaturation can be achieved by several
mechanisms. Firstly, bubbles that form from white cap activity can be
entrained and dissolved at depth. The slight but significantly different
pressure relative to the surface favours gas dissolution and results in a
higher equilibrium concentration. Secondly as evident from Figure 5, if
two water masses that have been equilibrated at different temperatures
are mixed, then the resulting water body would be supersaturated.
Thirdly, gases that are produced in situ by biological activity may
become supersaturated, particularly when evasion to the atmosphere is
hindered.
8
9

R. F. Weiss, Deep-Sea Res., 1970,17, 721.
D . Kester, 'Chemical Oceanography', ed. J. P. Riley and G. Skirrow, Academic Press, London, 2nd
Edn., 1975,VoI. l,p.497.

GAS SOLUBILITY (ml/1 atm)

Ar
O2
N2
Ne
He

TEMPERATURE(0C)

Figure 5

The solubility of various gases in seawater as a function of temperature
(From Broecker and Peng10)

The gas solubility for a water body in equilibrium with the overlying
air mass can be expressed in several ways. It is convenient to consider
Henry's Law that states:

where H is the Henry's law constant and ca and cw refer to the
concentration of a gas in air and water, respectively. As discussed by
Liss (1983)11, air-sea exchange occurs when a concentration gradient
exists {i.e. AC = cji~l — cw) and the magnitude of the consequential
flux, F, is given as:

10
n

W. S. Broecker and T. H. Peng, 'Tracers in the Sea', Lamont-Doherty Geological Observatory,
Palisades, 1982, p. 690.
P . S . Liss, 'Air-Sea Exchange of Gases and Particles', ed. P. S. Liss and W. G. N. Slinn, Reidel,
Dordrecht, 1983, p. 241.

Table 3
Gas

The net global fluxes of some trace gases across the airj sea interface
Global air-sea direction*

Flux magnitude^

+
+

+
+
+
+
From Chester.J
+ sea to air, — air to sea, = no net flux.

where the proportionality constant, K, has dimensions of velocity and so
is frequently referred to as the transfer velocity (see also Chapter 6).
Air-sea exchange processes are consequently dependent upon the
concentration gradient and the transfer velocity. The transfer velocity is
not a constant, but rather depends upon several physical parameters
such as temperature, wind speed, and wave state. The exchange can also
be attenuated by the presence of a surface film or slick. Alternatively, the
exchange can be facilitated by bubble formation. The concentration
gradient determines the direction of the flux, into or out of the ocean. Net
global fluxes for some gases are presented in Table 3. The atmosphere
serves as the source of material for conservative gases, especially those of
anthropogenic origin, but several gases produced in situ by biological
activity evade from the ocean.
2.2.2 Oxygen. Oxygen is a non-conservative gas and a typical oceanic
profile is shown in Figure 6. The concentration varies throughout the
water column, its distribution being greatly influenced by biological
activity. The generalized chemical equation for carbon fixation is often
given as:

During photosynthesis this reaction proceeds to the right, thereby
producing organic material, designated by (CH2O)m and O2. The
surface waters become equilibrated with respect to atmospheric O2, but
they can get supersaturated during periods of intense photosynthetic
activity. Respiration occurs as the above reaction proceeds to the left and

Depth (m)

Molecular Oxygen (jig kg ! )

Figure 6 A profile of molecular oxygen in the North Pacific Ocean
(Data from Bruland 12 )

O2 is consumed. Photosynthesis is obviously restricted to the upper
ocean (in the photic zone) and ordinarily exceeds respiration. However,
the relative importance of the two processes changes with depth. The
oxygen compensation depth is the horizon in the water column at which
the rate of O2 production by photosynthesis equals the rate of respiratory O2 oxidation.
Below the photic zone, O2 is utilized in chemical and biochemical
oxidation reactions. As evident in Figure 6, the concentration diminishes
with depth to develop an oxygen minimum zone. Thereafter, the O2
concentration in deeper waters begins to increase because these waters
originated from polar regions. They were cold and in equilibrium with
atmospheric gases at the time of sinking, but subsequently lost little of
the dissolved O2 because the flux of organic material to deep waters is
relatively small.
The dissolved O2 content of seawater has a significant control on the
redox potential, often designated in environmental chemistry by pe. This

K. W. Bruland, Earth Planet. ScL Lett., 1980, 47, 176.

is defined with reference to electron activity in an analogous fashion to
pH and thus:

The relationship between pe and the more familiar electrode potential E
or E\i is:

and for the standard state:

where F is Faraday's constant, R is the universal gas constant, and T is
the absolute temperature in Kelvin. Whereas a high value of ps indicates
oxidizing conditions, a low value signifies reducing conditions. Oxygen
plays a role via the reaction:
O2 + 4H+ + 4e" ^2H 2 O
At 200C, K= 1083-1 and so water of pH - 8.1 in equilibrium with
atmospheric O2 (po2 = 0.21 atm) has ps = 12.5. This conforms to
surface conditions but the pa decreases as the O2 content diminishes
with depth. The oxygen minimum is particularly well developed beneath
the highly productive surface waters of the eastern tropical Pacific Ocean
where there is a large flux of organic material to depth and subsequently
considerable oxidation. The O2 becomes sufficiently depleted that the
resulting low redox conditions causes NO3" to be reduced to NO2".
When circulation is restricted vertically due to thermal or saline
stratification and horizontally by topographic boundaries, the water
becomes stagnant and the oxygen may be completely utilized producing
anoxic conditions. Such regions represent atypical marine environments
where reducing conditions prevail. Well known examples include the
Black Sea, which is permanently anoxic below 200 m, and the Cariaco
Trench, a depression in the Venezuelan continental shelf. Some fjords,
such as Saanich Inlet in western Canada and Dramsfjord of Norway,
may be intermittently anoxic. Periodic flushing of these inlets by dense,
oxygenated waters displaces deep anoxic water to the surface causing
massive fish mortality.
O2 can be used as a tracer to help identify the origin of water masses.
The warm, saline intrusion into the Atlantic Ocean from the Mediterra-

nean Sea is relatively O2 deficient. Alternatively, the waters downwelling
from polar regions have elevated O2 concentrations.
2.2.3 Carbon Dioxide and Alkalinity. Marine chemists sometimes
adopt activity conventions quite different from those traditionally used
in chemistry. It is useful to preface a discussion about the carbon
dioxide-calcium carbonate system in the oceans with a brief outline of
pH scales. Although originally introduced in terms of ion concentration,
today the definition of pH is based on hydrogen ion activity and is:

where a^ refers to the relative hydrogen ion activity (i.e. dimensionless,
as is pH). Defined using concentration scales, the pH can be:

where cH and mH represent molar and molal concentrations, c° and m°
are the respective standard state conditions (1 mol I" 1 and 1 mol kg" 1 ),
and yH is the appropriate activity coefficient. Obviously yH differs in these
two expressions as c° ^ m°. However, different activity scales may also
be used. In the infinite dilution activity scale, yn -» 1 as the concentration
of hydrogen ions and all other ions approach O. For analyses, pH meters
are calibrated using dilute buffers prepared in pure water. Alternatively,
in the constant ionic medium activity scale, yH -* 1 as the concentration of
hydrogen ions approaches O while all other components are maintained
at some constant level. Calibrant buffers are prepared in solutions of
constant ionic composition, and in marine chemistry this is often a
solution of synthetic seawater. While these two methodologies are
equally justifiable from a thermodynamic point of view, it is important
to appreciate that pH scales so defined are quite different. As a further
consequence, the absolute values for dissociation constants also differ.
The biogeochemical cycle of inorganic carbon in the ocean is extremely complicated. It involves the transfer of gaseous carbon dioxide
from the atmosphere into solution. Not only is this a reactive gas that
readily undergoes hydration in the ocean, but also it is fixed as organic
material by marine phytoplankton. Inorganic carbon can be regenerated
either by photochemical oxidation in the photic zone or via respiratory
oxidation of organic material at depth. Surface waters are supersatu-

rated with respect to aragonite and calcite, forms of CaCO3, but
precipitation is limited to coastal lagoons such as found in the
Bahamas. However, several marine organisms utilize calcium carbonate
to form shells. Sinking shells can remove inorganic carbon from surface
waters which is then regenerated following dissolution in the undersaturated waters found at depth. Nonetheless, calcitic oozes of biogenic
origin constitute a major component in marine sediments. Finally, the
inorganic carbon equilibrium is responsible for buffering seawater at a
pH near 8 on time scales of centuries of millennia.
There are several equilibria to be considered. Firstly, CO2 is exchanged
across the air-sea interface:

The equilibrium process obeys Henry's Law, but the dissolved CO2
reacts rapidly with water to become hydrated as:

Relative to the exchange process, the hydration reaction forming
carbonic acid occurs quite quickly. This means that the concentration
of dissolved CO2 is extremely low. The two processes can be considered
together as:

The equilibrium constant is then:

where pco2 is the partial pressure of CO2 in the marine troposphere.
Carbonic acid undergoes dissociation:

for which the first and second dissociation constants (using (H + ) rather
than (H3O + )) are:

The hydrogen ion activity can be established with a pH meter.
However, as discussed above, this measurement must be operationally
defined. On the other hand, the individual ion activities of bicarbonate
and carbonate ions cannot be measured. Instead, ion concentrations are
determined, as outlined below, by titration. Accordingly, the equilibrium
constants are redefined in terms of concentrations. These are then known
as apparent rather than true equilibrium constants and distinguished
using a prime notation. It must be appreciated that apparent equilibrium
constants are not invariant, but rather are affected by temperature,
pressure, salinity, and, as outlined previously, the pH scale adopted. The
apparent dissociation constants are:

It should be noted that whereas ion activities are denoted by curly
brackets { }, concentrations are designated by square brackets [].
Analogous to the pH, pK conventionally refers to — logK. Numerical
values for the constants pA^Co2> P^i > and pA^ based on a constant ionic
medium scale (i.e. seawater with chlorinity = 19%o) are given in Table 4.
This provides sufficient information to calculate the speciation of
carbonic acid in seawater at a given temperature as a function of pH.
Table 4

Equilibrium constants for the carbonate system

T(0C)

pKco*

pK\

pK'2

0
5
10
15
20
25
30

1.19
1.27
1.34
1.41
1.47
1.53
1.58

6.15
6.11
6.08
6.05
6.02
6.00
5.98

9.40
9.34
9.28
9.23
9.17
9.10
9.02

Adapted from Stumm and Morgan. 13

W. Stumm and J. J. Morgan, 'Aquatic Chemistry', John Wiley & Sons, New York, 3rd Edn., 1996,
p. 1022.

Species Distribution

pH
Figure 7

The distribution of carbonic acid species in seawater of 35 psu at 15 0 C as a
function ofpH

This is shown for carbonic acid at 150C in seawater equilibrated with
atmospheric carbon dioxide (~ 3.5 x 10~4atm) in Figure 7. While there
are several confounding features, the pH of seawater can be considered
to be buffered by the bicarbonatexarbonate pair. The pH is generally
about 8, but is sensitive to the concentration ratio [HCO3Ji[COf"] as
evident from rearranging the expression for K2 to become:

To understand the response of the oceanic CO2 system to in situ
biological activity or enhanced CO2 concentrations in the atmosphere, it
is necessary to consider in more detail the factors influencing the
inorganic carbon cycle. Two useful parameters can be introduced.
Firstly, the total concentration of inorganic carbon, SCO2, in seawater
is:

The first term is negligible and as evident in Figure 7, the major species at
pH 8 are HCO3" and CO 3 ".
Alkalinity is defined as a measure of the proton deficit in solution and
should not be confused with basicity. Alkalinity is operationally defined

by titration with a strong acid to the carbonic acid end point. This is
known as the titration alkalinity (TA). Seawater contains weak acids
other than bicarbonate and carbonate and so TA is given as:

The influence of [OH ~] and [H + ] on the TA are small and can often be
ignored. The borate contributes about 3% of the TA and, if not
determined independently, can be estimated from the apparent boric
acid dissociation constants and the salinity, relying upon the relative
constancy of composition of sea salt. This would give the carbonate
alkalinity (CA):

Considering the dissociation constants above, this can be alternatively
expressed as:

This equation can be rearranged to give the following quadratic expression that can be solved for the pH:

Worked Example 1

For seawater (35 psu, 150C) with an alkalinity of 2.30 meq 1 - 1 and in
equilibrium with atmospheric CO2 = 3.65 x 10~4 atm, calculate (i) the
pH and (ii) the speciation of carbonic acid.
(i) pH calculation
Data, including constants from Table 4, are: KCo2 = 10~L41,
K[ = 10" 605 , K'2 = 1(T9-23, CA = 2.30 x \0~\pco, = 3.65 x 10~4
The pH is obtained from the calculation of {H+ } using the above

quadratic equation. Thus,

where

Giving {H+ } = 6.22 x 10~ 9 andpH = 8.21
(H) Carbonic acid speciation calculations

Knowing the pH, each of the three major species (H2CO3, HCO^,
CO 3 ") can be calculated as a fraction (or percentage) of the ECO2. Note
that the negligible contribution due to dissolved CO2 is ignored. The
necessary expressions are derived from the definitions of the K[ and K'2
given previously.

Substituting into the expression for the summation of all carbonic
species gives:

Thereafter, the fractional contribution of each species to the total can be
calculated using:

Substituting the values for K\ and K'2 and using the previously calculated

pH of 8.21, the fractional contribution of each species is 0.006,0.907, and
0.087 for H2CO3, HCO3", and CO 3 ", respectively.
Consider now the effect of altering the/?Co2 i n the water. The alkalinity
should not change in response to variations in CO2 alone because the
hydration and dissociation reactions give rise to equivalent amounts of
H + and anions. CO2 can be lost by evasion to the atmosphere (a process
usually confined to equatorial regions) or by photosynthesis. This causes
the ZCO2 to diminish and the pH to rise, an effect that can be quite
dramatic in tidal rock pools in which pH may then rise to 9. Conversely,
an increase in /?co2? cither by invasion from the atmosphere or release
following respiration, prompts an increase in ECO2 and a fall in pH.
Thus, the depth profiles of pH would mimic that of O2 but the ZCO2
would exhibit a maximum at the oxygen minimum.
There are further confounding influences, in particular concerning
CaCO3. CaCO3 in the form of aragonite or calcite is used by many
organisms to form calcareous shells (tests). The shells sink and dissolve
when the organism dies. The solubility is governed by:

Surface waters are supersaturated with respect to CaCO3, but precipitation rarely occurs, possibly due to an inhibitory effect by Mg 2+ forming
ion pairs with CO 3 ". The solubility OfCaCO3 increases with depth, due
to both a pressure effect and the decrease in pH following respiratory
release of CO2, with the result that the shells dissolve. This behaviour not
only increases the alkalinity but also accounts for the non-conservative
nature OfCa2+ and inorganic carbon in deep waters. The depth at which
appreciable dissolution begins is known as the lysocline. At a greater
depth, designated as the carbonate compensation depth (CCD), no
calcareous material is preserved in the sediments. The depths of the
lysocline and CCD are influenced by the flux of organic material and
shells, and tend to be deeper under high productivity zones.
The CO2 and CaCO3 systems are coupled in that the pH buffering in
the ocean is due to the reaction:

In addition to the effects noted previously, an input of CO2 promotes the
dissolution OfCaCO3. The reaction does not proceed to the right without
constraint, but rather meets a resistance given by the Revelle factor, R:

This value is approximately 10, indicating that the ocean is relatively well
buffered against changes in XCO2 in response to variations in atmospheric ^CO2- Although the ocean does respond to an increase in the
atmospheric burden of CO2, the time scales involved are quite considerable. The surface layer can become equilibrated on the order of decades,
but as the thermocline inhibits exchange into deep waters, the equilibration of the ocean as a whole with the atmosphere proceeds on the order
of centuries. The ventilation of deep water by downwelling water masses
in polar latitudes only partly accelerates the overall process.
2.2.4 Dimethyl Sulfide and Climatic Implications. The Gaia hypothesis of Lovelock14 states that the biosphere regulates the global environment for self-interest. This presupposes that controls, perhaps poorly
understood or unknown, serve to maintain the present status quo.
Charlson and co-workers have made use of this hypothesis to suggest
that biogenic production of dimethyl sulfide (DMS) and the consequent
formation of atmospheric cloud condensation nuclei (CCN, i.e. small
particles onto which water can condense) acts as a feedback mechanism
to counteract the global warming resulting from elevated greenhouse gas
concentrations in the atmosphere.15 The cycle is illustrated in Figure 8.
Global warming, with concurrent warming of the ocean surface, leads to
enhanced phytoplankton productivity. This promotes the production
and evasion to the atmosphere of DMS. The DMS undergoes oxidation
to form CCN which promote cloud formation and increase the planetary
albedo {i.e. reflectivity with respect to sunlight) thereby causing a cooling
effect. From a biogeochemical perspective, the two key features are the
controls on the biogenic production of DMS and the formation of CCN
following aerial oxidation of DMS. These will be considered below in
more detail. With respect to the physics, the most important aspects of
the proposed climate control mechanism are that the enhancement of the
albedo is due to an increase in the number and type of CCN, and that this
CCN production occurs in the marine boundary layer. The albedo of
calm seawater is very low (~2%) in comparison to vegetated regions
(10-25%), deserts (-35%), and snow covered surfaces (~90%).
That biological processes within the oceans act as a major source of
reduced sulfur gases is well established.16 Of particular importance is the
generation of DMS. Surface concentrations, approximately in the range
14

J. Lovelock, 'Gaia. A New Look at Life on Earth', Oxford University Press, Oxford, 1979.
R. J. Charlson, J. E. Lovelock, M. O. Andreae, and S. G. Warren, Nature, 1987, 326, 655.
16
M. O. Andreae, 'The Role of Air-Sea Exchange in Geochemical Cycling', ed. P. Buat-Menard,
Reidel, Dordrecht, 1986, p. 331.
15

Number concentration
of cloud droplets
(Fixed LWP)

Scattering of solar
radiation by droplets

Cloud
albedo

Cloud nucleation

Loss of solar
radiation to space

CCN
Formation of
water* soluble particles

NSS - SO*
Oxidation

Surface
temperature
of Earth

Solar
irradiance
below clouds

DMS (gas)
Sea-to-air
transport
Atmosphere
Ocean

DMS (aq)

Figure 8

Production of DMS
by marine phytoplankton

The possible climatic influence of dimethyl sulfide of marine biogenic origin
(From Charlson et al.15)

0.7-17.8 nmol I" 1 , exhibit large temporal and geographic variations.
Oceanic distributions indicate that DMS is produced within the photic
zone, which is consistent with a phytoplankton source, but DMS
concentrations are poorly correlated with normal indicators of primary
productivity. While Phaeocystis and Coccolithoporidae have been identified as important DMS producers, there is still uncertainty as to the full
potential for biological DMS formation. With respect to climate modification, questions remain as to the biological response to global
warming. For the model of Charlson and co-workers15 to hold, either
organisms might increase DMS formation or biological succession could
change in such a way as to favour DMS producers. Thus, marine
biogenic source strengths and the controlling factors remain important
unresolved issues in sulfur biogeochemistry.
DMS concentrations in the remote marine troposphere vary in the
range of 0.03-32 nmol m~ 3 . Not surprisingly, and as with the seawater
concentrations, considerable temporal and geographic disparities occur.
Furthermore, atmospheric DMS concentrations exhibit diurnal variations, with a night-time maximum and an afternoon minimum consistent

with a photochemical sink. Whereas oxidation involves HO* free radicals
during the day, a reaction with NO3* may be important at night.
Relatively low levels are associated with air masses derived from continental areas, owing to the enhanced concentrations of oxidants. While
oceanic venting rates are dependent upon a number of meteorological
and oceanographic conditions, there is no question that the marine photic
zone acts as the major source of DMS to the overlying troposphere.
The oxidation of DMS in the atmosphere could yield several products,
namely dimethyl sulfoxide (DMSO), methanesulfonate (MSA) or
sulfate. Insofar as aerosol formation is concerned, the two key products
are MSA and SO4-. Atmospheric particles in the sub-micron size range
exert a significant influence on the earth's climate. The effect can be
manifested via three mechanisms. Firstly, the particles themselves may
enhance backscatter of solar radiation. Secondly, they act as cloud
condensation nuclei promoting cloud formation and so increasing the
earth's albedo. Thirdly, such clouds affect the hydrological cycle. The
evidence for such a biofeedback mechanism limiting global warming
remains circumstantial.
2.3

Nutrients

Although several elements are necessary to sustain life, traditionally in
oceanography the term 'nutrients' has referred to nitrogen (usually
nitrate but also ammonia), phosphate, and silicate. The rationale for
this classification was that analytical techniques had long been available
that allowed the precise determination of these constituents despite their
relatively low concentrations. They were observed to behave in a
consistent and explicable manner, but quite differently from the major
constituents in seawater.
The distributions of these three nutrients are determined by biological
activity. Nitrate and phosphate become incorporated into the soft parts
of organisms. As evident in the modified carbon fixation equation of
Redfield given below:17

the uptake of these nutrients into tissues occurs in constant relative
amounts. The ratio (i.e. Redfield ratio) for C:N:P is 106:16:1. Silicate is
utilized by some organisms, particularly diatoms (phytoplankton) and
radiolaria (zooplankton), to form siliceous skeletons. Such skeletons
17

A. C. Redfield, Am. J. ScL, 1958, 46, 205.

Depth (m)

Phosphate (jig kg O

Nitrate & Silicate (jig kg')
Figure 9

The depth distribution of nitrate (%), phosphate (M), and silicate (A.) in the
North Pacific Ocean
(Date from Bruland 12 )

consist of an amorphous, hydrated silicate, SiO2^H2O, often called
opaline silica.
A depth profile of nitrate, phosphate, and silicate in the North Pacific
Ocean is presented in Figure 9. Nutrients behave much like SCO2 and
are removed in the surface layer, especially in the photic zone. Thus,
concentrations can become quite low, and indeed sufficiently low to limit
further photosynthetic carbon fixation. The organisms sink following
death. The highest concentrations occur where respiration and bacterial
decomposition of the falling organic material are greatest, that is at the
oxygen minimum. The nutrients, including silica, are consequently
regenerated and their concentrations in deep waters are much greater
than those observed in the surface waters, thereby accounting for the
fertilizing effect of upwelling. It should be noted that the siliceous
remains behave differently from the calcareous shells discussed previously. The oceans everywhere are undersaturated with respect to
silica. Its solubility exhibits no pronounced variation with depth and
there is no horizon analogous to the CCD (see Section 2.2.3). Silica is
preserved to any great extent only in deep sea sediments associated with
the highly productive upwelling zones in the ocean.

2.4

Trace Elements

Trace elements in seawater are taken to be those that are present in
quantities less than 1 mg I" 1 , excluding the nutrient constituents. The
distribution and behaviour of minor elements have been reviewed in the
light of data that conform to an oceanographically consistent manner.1'7
Analytical difficulties are readily comprehensible when it is appreciated
that the concentration for some of these elements can be extremely low,
i.e. a few pg 1" l for platinum group metals.18 Some trace elements, such
as Cs + , behave conservatively and therefore absolute concentrations
depend upon salinity. More often, the elements are non-conservative and
their distributions in both surface waters and the water column vary
greatly, reflecting the differing source strengths and removal processes in
operation. Generalizations regarding residence times cannot be made, as
biologically active elements are removed from seawater relatively rapidly
but conservative constituents and platinum group metals have rather
long residence times of the order of 105 years.
Considering firstly the distribution in surface waters, several elements
exhibit high concentrations in coastal waters in comparison to levels in
the centres of oceanic gyres. Typically, this distribution arises because
the elements originate predominantly from riverine inputs or through
diffusion from coastal sediments. However, as they are effectively
removed from the surface waters in the coastal regions, little material is
advected horizontally to the open sea. Examples of elements that behave
in this way are Cd, Cu, and Ni. In contrast, the concentration of Pb,
including 210Pb, is greater in the gyres. This results from a strong
widespread aeolian (wind-borne) input coupled with less effective
removal from surface waters in the gyres.
Clearly the removal mechanisms have an appreciable effect on
dissolved elemental abundances. The two major processes in operation
are uptake by biota and scavenging by suspended particulate material. In
the first instance, the constituent mimics the behaviour of nutrients. This
is evident in the metal:nutrient correlation for Cd:P and Zn:Si (Figure
10).
No consistent pattern for depth profiles of trace elements exists.
Conservative elements trend with salinity variations, provided they
have no significant submarine sources. Non-conservative elements may
exhibit peak concentrations at different depths in oxygenated waters as:
1. surface enrichment;
2. maximum at the O2 minimum;
18

E. Goldberg, M. Koide, J. S. Yang, and K. K. Bertine, 'Metal Speciation: Theory, Analysis and
Applications', ed. J. R. Kramer and H. E. Allen, Lewis Publishers, Chelsea, 1988, p. 201.

cadmium (nmol kg" 1 )

phosphate (^mol kg"')

zinc (nmol kg l )

(a)

silicate (/nmol k g ' l )
(b)

Figure 10 Metal: nutrient correlations in the North Pacific for Cd: P and Zn: Si
(FromBruland12)

3. mid-depth maximum not associated with the O2 minimum;
4. bottom enrichment.
The criteria for an element, such as Pb, to exhibit a maximum concentration in surface waters are that the only significant input must be at the
surface (aeolian supply) and it must be effectively removed from the
water column. Constituents such as As, Ba, Cd, Ni, and Zn exhibit
nutrient type behaviour. Those elements (Cd) associated with the soft
parts of the organism are strongly correlated with phosphate and are
regenerated at the O2 minimum. Elements (e.g. Zn) associated with the
skeletal material may exhibit a smooth increasing trend with depth. The
third case pertains to elements, notably Mn, that have a substantial input
from hydro thermal waters. These are released into oceanic waters from
spreading ridges. Ocean topography is such once these waters are
advected away from such regions towards the abyssal plains, they are
then found at some intermediate depth. Bottom enrichment is observed
for elements (Mn) that are remobilized from marine sediments. The
behaviour of Al combines features outlined above, resulting in a middepth minimum concentration. Surface enrichment evident in mid
(410N) but not high (-60 0 N) latitudes in the North Atlantic results
from the solubilization of aeolian material. Removal occurs via scavenging and incorporation into siliceous skeletal material. Subsequent
regeneration by shell dissolution increases deep water Al levels.
2.5

Physico-chemical Speciation

Physico-chemical speciation refers to the various physical and chemical
forms in which an element may exist in the system. In oceanic waters, it is
difficult to determine speciation directly. Whereas some individual
species can be analysed, others can only be inferred from thermodynamic
equilibrium models as exemplified by the speciation of carbonic acid in
Figure 7. Often an element is fractionated into various forms that behave
similarly under a given physical (e.g. filtration) or chemical (e.g. ion
exchange) operation. The resulting partition of the element is highly
dependent upon the procedure utilized, and so known as operationally
defined. In the following discussion, speciation will be exemplified with
respect to size distribution, complexation characteristics, redox behaviour, and methylation reactions.
Physico-chemical speciation determines the environmental mobility of
an element, especially with respect to partitioning between the water and
sediment reservoirs. The influence can be manifested through various
mechanisms as summarized in Figure 11. Settling velocities, and by
implication the residence time, are controlled by the size of the particle.

fibrils

Living
biota

exudates

Inorganic ltgands
Low-moe
l cua
lr
weg
i ht organic
ligands

sorption, bio uptake
Metal
colloid
compe
lx
Hm
yde
rated
ionstal

inoM
rgeata
nlic
&
[compe
l xes

Ternary
compe
l xes

Hydrol
compleyx1iecs',j

Coatn
i gs Precipitation

Sorption

Sorption / Precipitation

Suspended mater

Sorption

Flux of low molecular weight organics from pore water

Flux of iron, manganese and high molecular weight organics
from reduced pore waters

Iron, manganese
High-molecular
weg
i ht organics

Sedm
i entato
i n/Resuspenso
i n Dissolution
Depostied sediments, g
i neous minerals

Figure 11

Speciation of metal ions in seawater and the main controlling mechanisms
(From Ohman and Sjoberg19)

Thus, dissolved to participate interactions involving adsorption, precipitation, or biological uptake can effectively remove a constituent from the
water column. The redox state can have a comparable influence. Mn and
Fe are reductively remobilized in sediments. Following their release from
either hydrothermal sources or interstitial waters, they rapidly undergo
oxidation to form colloids, which are then quickly removed from the
water column. Speciation also determines the bioavailability of an
element for marine organisms. It is generally accepted that the uptake
of trace elements is limited to free ions and some types of lipid-soluble
organic complexes. This is important in that some elements may be
essential, but toxic at elevated concentrations. Via complexation reactions, organisms may be able to modify the free ion concentration.
An element can exist in natural waters in a range of forms that exhibit
a size distribution as indicated in Table 5. Entities in true solution include
ions, ion pairs, complexes, and a wide range of organic molecules that
can span several size categories. At the smallest extreme are ions, which
exist in solution with a co-ordinated sphere of water molecules as
discussed previously. Na + , K + , and C P exist predominantly as free
hydrated ions. Collisions of oppositely charged ions occur due to
electrostastic attraction, and can produce an ion pair. An ion pair is the
19

L. Ohman and S. Sjoberg, 'Metal Speciation: Theory, Analysis and Applications', ed. J. R.
Kramer and H. E. Allen, Lewis Publishers, Chelsea, 1988, p. 1.

Table 5

The size distribution of trace metal species in natural waters

Size range

Metal species

Examples

< 1 nm

Free metal ions

Mn 2 + , Cd 2 +
+

Phase state
Soluble
2

1-10 nm

Inorganic ion pairs, inorganic
complexes, low molecular mass
organic complexes

NiCl , HgCl T,
Zn-fulvates

Soluble

10-100 nm

High molecular mass organic
complexes

Pb-humates

Colloidal

100-1000 nm

Metal species adsorbed onto
inorganic colloids, metals
associated with detritus

Co-MnO 2 , Pb-Fe(OH) 3 Paniculate

> 1000 nm

Metals adsorbed into living
cells, metals adsorbed onto,
or incorporated into mineral
solids and precipitates

Cu-clays, PbCO 3(s)

Particulate

From de Mora and Harrison. 20

transient coupling of a cation and anion during which each retains its coordinated water envelope. While impossible to measure directly, concentrations can be calculated with knowledge of the ion activities and
stability constant. For the formation of the ion pair NaSO4T via:

the stability constant is defined as:

Ion pair formation is important for Ca2 + , Mg2 + , SO4 , and HCO3 .
If the attraction is sufficiently great, a dehydration reaction can occur
leading to covalent bonding. A complex consists of a central metal ion
sharing a pair of electrons donated by another constituent, termed a
ligand, acting as a Lewis base. The metal ion and ligand share a single
water envelope. Ligands can be neutral (e.g. H2O) or anionic (e.g. C P ,
HCO3T) species. A metal ion can co-ordinate with one or more ligands,
which need not be the same chemical entity. Alternatively, the cation can
share more than one electron pair with a given ligand thereby forming a
ring structure. This type of complex, known as a chelate, exhibits
20

S. J. de Mora and R. M. Harrison, 'Hazard Assessment of Chemicals, Current Developments', ed.
J. Saxena, Academic Press, London, Vol. 3, 1984, p. 1.

enhanced stability largely due to the entropy effect of releasing large
numbers of molecules from the water envelopes.
Complex formation is an equilibrium process. Ignoring charges, for the
general case of a metal, M, and ligand, L, complex formation occurs as:

for which the formation constant, Kh is given by:

A second ligand may then be co-ordinated as:

The equilibrium constant for ML 2 can be expressed solely in terms of the
activities of the M and L:

where /J2 is the product KxK2 and is known as the stability constant. The
case can be extended to include n ligands as:

Worked Example 2
Assuming that y = 1 for all species, calculate the speciation of mercury in
typical seawater (35 psu at 25 0C) given the following values for stepwise
stability constants for successive chlorocomplexes (K\ = 106'74,
K2 = 106-48, K3 = IO0-85, K4 = 10 100 ). Note that [Cl"] is 0.559 mmol
I" 1 and that it is not necessary to know the mercury concentration in
seawater.
The total mercury concentration is given as the sum of all contributing
species. Thus

From the definition of the stability constants, we know that

Thus,

Now let D
and
note that this is a constant for a stipulated chloride concentration. Thus,
at the seawater chloride concentration ([Cl ~] = 10~0'25) this gives

The fractional (or percentage) contribution of each species can be
determined using
Fractional contribution Expression

Value for [Cl~] = 0.559 mmol \~l

Thus, at the high chloride concentration found in seawater, the mercury
speciation is dominated by the tetra- and tri-chloro species.
If the total mercury content for a given seawater sample is known, the
concentration of each species can be readily calculated, e.g.:

This is a simplified example that has not considered other species
(bromides, hydroxides, etc.) that might be important in seawater.
The speciation of constituents in solution can be calculated if the
individual ion activities and stability constants are known. This information is relatively well known with respect to the major constituents in
seawater, but not for all trace elements. Some important confounding
variables create considerable difficulties in speciation modelling. Firstly,
it is assumed that equilibrium is achieved, meaning that neither biological interference nor kinetic hindrance prevents this state. Secondly,

seawater contains appreciable amounts (at least in relation to the trace
metals) of organic matter. However, the composition of the organic
matrix, the number of available binding sites, and the appropriate
stability constants are poorly known. Nevertheless, speciation models
can include estimates of these parameters. Organic material can form
chelates with relatively high stability constants and dramatically decrease
the free ion activity of both necessary and toxic trace elements. Organisms may make use of such chemistry, producing compounds either to
sequester metals in limited supply or to detoxify contaminants. Thirdly,
surface adsorption of dissolved species onto colloids or suspended
particles may remove them from solution. As with organic matter, an
exact understanding of the complexation characteristics of the suspended particles is not available, but approximations can also be
incorporated into speciation models.
Elements may be present in a variety of phases other than in true
solution. Colloidal formation is particularly important for elements such
as Fe and Mn, which produce amorphous oxyhydroxides with very great
complexation characteristics. Adsorption processes cannot be ignored in
biogeochemical cycling. Particles tend to have a much shorter residence
time in the water column than do dissolved constituents. Scavenging of
trace components by falling particles accelerates deposition to the
sediment sink.
Several elements in seawater may undergo alkylation via either
chemical or biological mechanisms.21 Type I mechanisms involve
methyl radical or carbonium ion transfer and no formal change in the
oxidation state of the acceptor element. The incoming methyl group may
be derived for example from methylcobalamin coenzyme, S-adenosylmethionine, betaine, or iodomethane. Elements involved in Type I
mechanisms include Pb, Tl, Se, and Hg. Other reaction sequences
involve the oxidation of the methylated element. The methyl source can
be a carbanion from methylcobalamin coenzyme. Oxidative addition
from iodomethane and enzymatic reactions has also been suggested.
Some elements that can undergo such methylation processes are As, Sb,
Ge, Sn and S. Methylation can enhance the toxicity of some elements,
especially for Pb and Hg. The environmental mobility can also be
affected. Methylation in the surface waters can enhance volatility and
so favour evasion from the sea, as observed for S, Se, and Hg.
Methylation within the sediments may facilitate transfer back into
overlying waters.
Elements may exhibit multiple oxidation states in seawater. Redox
21

P. Craig, 'Organometallic Compounds in the Environment', ed. P.S. Craig, Longman, Harlow,
1986, p. 1.

processes can be modelled in an analogous manner to the ion pairing and
complexation outlined previously. The information is often presented
graphically in the form of a predominance area diagram, that is a plot of
pe versus pH showing the major species present for the designated
conditions. Although a single oxidation state might be anticipated from
equilibrium considerations, there are several ways in which multiple
oxidation states might arise. Biological activity can produce non-equilibrium species, as evident in the alkylated metals discussed above.
Whereas MnIV and Cu11 might be expected by thermodynamic reasoning, photochemical processes in the surface waters can lead to the
formation of significant amounts of Mn11 and Cu1. Fe111 is the favoured
redox state of Fe in seawater, but it is relatively insoluble and exists
predominantly in a colloidal phase. Photochemical reduction to Fe11,
which only slowly oxidizes to Fe111, might act as a very important
mechanism rendering Fe bioavailable to marine organisms.
Goldberg and co-workers have presented impressive information
(given that seawater concentrations are as low as 1.5 pg I" 1 for Ir and
2 pg I" 1 for Ru) on the speciation, including redox state, of platinum
group metals as a means of interpreting distributions in seawater and
marine sediments.18 Pt and Pd are stabilized in seawater as tetrachlorodivalent anions. Their relative abundance of 5 Pt.l Pd agreeing with a
factor of five difference in /J4. Pt is enriched in ferromanganese nodules
following oxidation to a stable (iv) state, behaviour not observed for Pd.
Rh exists predominantly in the heptavalent state, but accumulates in
reducing sediments as lower valence sulfides. Au and Ag are present
predominantly in solution as monovalent forms. Ag, but not Au,
accumulates in anoxic coastal sediments.
3 SUSPENDED PARTICLES AND MARINE SEDIMENTS
3.1 Description of Sediments and Sedimentary Components
The sediments represent the major sink for material in the oceans. The
main pathway to the sediments is the deposition of suspended particles.
Such particles may be only in transit through the ocean from a
continental origin or be formed in situ by chemical and biological
processes. Sinking particles can scavenge material from solution.
Accordingly, this section introduces the components found in marine
sediments, but emphasises processes that occur within the water column
that lead to the formation and alteration of the deposited material.
Marine sediments cover the ocean floor to a thickness averaging
500 m. The deposition rates vary with topography. The rate may be
several mm per year in nearshore shelf regions, but is only 0.2-7.5 mm

Table 6

The four categories of marine sedimentary components with examples of
mineral phases

Classification

Mineral example

Lithogenous

Quartz
Microcline
Kaolinite
Montmorillonite
Mite
Chlorite
Fe-Mn minerals
Carbonate fluoroapatite
Barite
Pyrite
Aragonite
Dolomite

Hydrogenous

Biogenous

Calcite
Aragonite
Opaline silica
Apatite
Barite
Organic matter

Cosmogenous

Cosmic spherules
Meteoric dusts

Chemical formula

From Harrison and de Mora.22

per 1000 years on the abyssal plains. Oceanic crustal material is formed
along spreading ridges and moves outwards eventually to be lost in
subduction zones, the major trenches in the ocean. Because of this
continual movement, the sediments on the seafloor are no older than
Jurassic in age, about 166 million years.
The formation of marine sediments depends upon chemical, biological, geological, and physical influences. There are four distinct processes
that can be readily identified. Firstly, the source of the material is
obviously important. This is usually the basis on which sediment
components are classified and will be considered below in more detail.
Secondly, the material and its distribution on the ocean floor are
influenced by its transportation history, both to and within the ocean.
Thirdly, there is the deposition process that must include particle
formation and alteration in the water column. Finally, the sediments
may be altered after deposition, a process known as diagenesis. Of
particular importance are reactions leading to changes in the redox
state of the sediments.
22

R. M. Harrison and S. J. de Mora, 'Introductory Chemistry for the Environmental Sciences', 2nd
Edn., Cambridge University Press, Cambridge, 1996, p. 373.

The components in marine sediments are classified according to
origin. Examples are given in Table 6. Lithogenous (or terrigenous)
material comes from the continents as a result of weathering processes.
The relative contribution of lithogenous material to the sediments will
depend upon proximity to the continent and the source strength of
material derived elsewhere. The most important components in the
lithogenous fraction are quartz and the clay minerals (kaolinite, illite,
montmorillonite, and chlorite). The distribution of the clay minerals
varies considerably. Illite and montmorillonite tend to be ubiquitous in
terrestrial material, but the latter has a secondary origin associated with
submarine volcanic activity. Kaolinite typifies intense weathering
observed in tropical and desert conditions. Therefore, it is relatively
enriched in equatorial regions. On the other hand, chlorite is indicative
of the high latitude regimes where little chemical weathering occurs. The
lithogenous components tend to be inert in the water column and
represent detrital deposition. Nonetheless, the particle surfaces can act
as important sites for adsorption of organic material and trace elements.
Hydrogenous components, also known as chemogenous or halmeic
material, are those produced abiotically within the water column. This
may comprise primary material formed directly from seawater upon
exceeding a given solubility product, termed authigenic precipitation.
The best known example of authigenic material is ferromanganese
nodules found throughout the oceans. Alternatively, secondary material
may be formed as components of continental or volcanic origin become
altered by low temperature reactions in seawater, a mechanism known
as halmyrolysis. Halmyrolysis reactions can occur in the estuarine
environment, being essentially an extension of chemical weathering of
lithogenous components. Such processes continue at the sedimentwater interface. Accordingly, there are considerable overlaps between
the terms weathering, halmyrolysis, and diagenesis. Owing to the
importance that surface chemistry has on the final composition,
authigenic precipitation and halmyrolysis are considered further in
Section 3.2.
Biogenous (or biotic) material is produced by the fixation of mineral
phases by marine organisms. The most important phases are calcite and
opaline silica, although aragonite and magnesian calcite are also deposited. As indicated in Table 7, several plants and animals are involved, but
the planktonic organisms are the most important with respect to the
world ocean. The source strength depends upon the species composition
and productivity of the overlying oceanic waters. For instance, siliceous
oozes are found in polar latitudes (diatoms) and along the equator
(radiolaria). The relative contribution of biogenous material to the
sediments depends upon its dilution by material from other sources and

Table 7

Quantitatively important plants and animals that secrete calcite,
aragonite, Mg-calcite, and opaline silica

Mineral

Plants

Animals

Calcite

Coccolithophorids a

Foraminifera a
Molluscs
Bryozoans

Aragonite

Green algae

Molluscs
Corals
Pteropods a
Bryozoans

Mg-calcite

Coralline (red) algae

Benthic foraminifera
Echinoderms
Serpulids (tubes)

Opaline silica

Diatoms a

Radiolaria a
Sponges

From Berner and Berner.23
Planktonic organisms.

the extent to which the material can be dissolved in seawater. As noted
previously, both calcareous and siliceous skeletons are subject to
considerable dissolution in the water column and at the sediment-water
interface.
There are two sources that give rise to minor components in the
marine sediments. Cosmogenous material is that derived from an
extraterrestrial source. Such material tends to comprise small (i.e.
<0.5 mm) black micrometeorites or cosmic spherules. The composition
is either magnetite or a silicate matrix including magnetite. They are
ubiquitous but scarce, with relative contributions to the sediments
decreasing with an increase in sedimentation rate. Finally, there are
anthropogenic components, notably heavy metals and Sn, which can
have a significant influence on sediments in coastal environments.
As noted in Section 1.1, the principal modes of transport of particulate
material to the ocean are by rivers or via the atmosphere. Within the
oceans, distribution is further affected by ice rafting, turbidity currents,
organisms, and oceanic currents. Turbidity currents refer to the turbid
and turbulent flow of sediment-laden waters along the seafloor caused by
sediment slumping. They are especially important in submarine canyons
and can transport copious amounts of material, including coarse-grained
sediments, to the deep sea. Ice rafting can also transport substances to
the deep sea. Although, ice rafting is presently confined to the polar
latitudes (40 0N and 55 0S), there have been considerable variations in ice
23

E. K. Berner and R. A. Berner, 'The Global Water Cycle', Prentice-Hall, Englewood Cliffs, NJ,
1987, p. 397.

limits within the geologic record. Organisms are notable not only for
biogenous sedimentation, but also because they can influence finegrained lithogenous material that becomes incorporated into faecal
pellets, consequently accelerating the settling rate. Ocean currents are
important for the distribution of material with a long residence time.
Major surface currents are zonal and tend to reinforce the pattern of
aeolian supply. On the other hand, deep water currents are of little
consequence as velocities are slow relative to the settling rates.
3.2

Surface Chemistry of Particles

3.2.1 Surface Charge. Particles in seawater tend to exhibit a negative
surface charge. There are several mechanisms by which this might arise.
Firstly, the negative charge can result from crystal defects (i.e. vacant
cation positions) or cation substitution. Clay minerals are layered
structures of octahedral AlO6 and tetrahedral SiO4. Either substitution
of Mg11 and Fe11 for the Al111 in octahedral sites or replacement of SiIV in
tetrahedral location by Al111 can cause a net negative charge. Secondly, a
surface charge can result from the differential dissolution of an electrolytic salt such as barite (BaSO4). A charge will develop whenever the rate
of dissolution of cations and anions differs. Thirdly, organic material can
be negatively charged due to the dissociation of acidic functional groups.
Adsorption processes can also lead to the development of a negatively
charged particle surface. One example is the specific adsorption of
anionic organic compounds onto the surfaces of particles. Another
mechanism relates to the acid-base behaviour of oxides in suspension.
Metal oxides (most commonly Fe, Mn) and clay minerals have frayed
edges resulting from broken metal-oxygen bonds. The surfaces can be
hydrolysed and exhibit amphoteric behaviour:

The hydroxide surface exhibits a different charge depending upon the
pH. Cations other than H + can act as the potential determining ion. The
point of zero charge (PZC) is the negative log of the activity at which the
surface exhibits no net surface charge. At the PZC:

The PZC for some mineral solids found in natural waters are shown in
Table 8. Clearly, the extent to which such surfaces can adsorb metal

Table 8

The point of zero charge (PZC) for
some mineral phases

Mineral

pHpzc

(X-Al2O3
(X-Al(OH)3
y- AlOOH
CuO
Fe 3 O 4
a-FeOOH
7-Fe 2 O 3
Te(OH) 3 ' (amorphous)
MgO
(5-MnO2
0-MnO 2
SiO 2
ZrSiO 4
Feldspars
Kaolinite
Montmorillonite
Albite
Chrysotile

9.1
5.0
8.2
9.5
6.5
7.8
6.7
8.5
12.4
2.8
7.2
2.0
5
2-2.4
4.6
2.5
2.0
>12

From Stumm and Morgan.13

cations will be dependent upon the pH of the solution. At the pH typical
of sea water, most of the surfaces indicated in Table 8 would be negatively
charged and would readily adsorb metal cations.
3.2.2 Adsorption Processes. Physical or non-specific adsorption
involves relatively weak attractive forces, such as electrostatic attraction
and van der Waals forces. Adsorbed species retain their co-ordinated
sphere of water and, hence, cannot approach the surface closer than the
radius of the hydrated ion. Adsorption is favoured by ions having a high
charge density, i.e. trivalent ions in preference to univalent ones.
Additionally, an entropy effect promotes the physical adsorption of
polymeric species, such as Al and Fe oxides, because a large number of
water molecules and monomeric species is displaced.
Chemisorption or specific adsorption involves greater forces of attraction than physical adsorption. As hydrogen bonding or n orbital
interactions are utilized, the adsorbed species lose their hydrated
spheres and can approach the surface as close as the ionic radius.
Whereas multilayer adsorption is possible in physical adsorption, chemisorption is necessarily limited to monolayer coverage.
As outlined previously, hydrated oxide surfaces have sites that are
either negatively charged or readily deprotonated. The oxygen atoms
tend to be available for bond formation, a favourable process for

transition metals. Several mechanisms are possible. An incoming metal
ion, M z + , may eliminate an H + ion as:

Alternatively, two or more H h ions may be displaced, thereby forming a
chelate as:

A metal complex, ML^ + , may be co-ordinated instead of a free ion by
displacement of one or more H + ions in a manner analogous to the
above reaction. In addition, the metal complex might eliminate a
hydroxide group giving rise to a metal-metal bond as:
—X—O—H + MU + ^ - X - M L i z ' 1 ) + + OH"
It should be noted that all of these reactions are equilibria for which an
appropriate equilibrium constant can be defined and measured. These
data can then be incorporated into the speciation models discussed in
Section 2.5.
3.2.3 Ion Exchange Reactions. Both mineral particles and particulate
organic material can take up cations from solution and release an
equivalent amount of another cation into solution. This process is
termed cation exchange and the cation exchange capacity (CEC) for a
given phase is a measure of the number of exchange sites present per
100 g of material. This is operationally defined by the uptake of
ammonium ions from 1 mol \~x ammonium acetate at pH 7. The specific
surface area and CEC are given in Table 9 for several sorption-active
materials.
There are several factors that influence the affinity of cations towards a
given surface. Firstly, the surface coverage will increase as a function of
the cation concentration. Secondly, the affinity for the exchange site is
enhanced as the oxidation state increases. Finally, the higher the charge
density of the hydrated cation, the greater will be its affinity for the
exchange site. In order of increasing charge density, the group I and II
cations are:

Table 9

The specific surface area and cation-exchange capacities of several
sorption-active materials

Material

Specific surface area
(m2 g~l)

Cation-exchange capacity
(meq/100 g)

Calcite(<2/mi)
Kaolinite
Illite
Chlorite
Montmorillonite
Freshly precipitated Fe(OH) 3
Amorphous silicic acid
Humic acids from soils

12.5
10-50
30-80
50-150
300
1900

3-15
10-40
20-50
80-120
10-25
11-34
170-590

From Forstner and Wittman.24

3.2.4 Role of Surface Chemistry in Biogeochemical Cycling. Reactions
at the aqueous-particle interface have several consequences for material
in the marine environment, from estuaries to the deep sea sedimentwater boundary. Within estuarine waters, suspended particles experience
a dramatic change in the composition and concentration of dissolved
salts. A number of halmyrolysates can be formed. Clay minerals undergo
cation exchange as Mg 2+ and Na + replace Ca 2+ and K+ . Alternatively,
montmorillonite may take up K + becoming transformed to illite.
Hydrogenous components, in particular Mn and Fe oxides, may be
precipitated onto the surfaces of suspended particles. The particulate
material generally accumulates organic coatings within estuaries, which,
together with an increase in the ionic strength of the surrounding
solution, leads to the formation of stable colloids. Both the oxide and
organic coatings can subsequently scavenge other elements in the
estuary.
Within the ocean, the exchange of material from the dissolved to the
suspended particulate state influences the distribution of several elements. This scavenging process removes dissolved metals from solution
and accelerates their deposition. The effectiveness of this process is
obvious in the depth profiles of metals, especially those of the surface
enrichment type. Furthermore, the removal can be expressed in terms of
a deep water scavenging residence time as indicated in Table 10.
The scavenging mechanism can be particularly effective in the watersediment boundary region. The resuspension of fine sediments generates
a very large surface area for adsorption and ion exchange processes.
Within the immediate vicinity of hydrothermal springs, reduced species
24

U. Forstner and G. T. W. Wittman, 'Metal Pollution in the Aquatic Environment', SpringerVerlag, Berlin, 2nd Edn., 1981, p. 486.

Table 10

The deep water scavenging residence times of some trace elements in
the oceans

Element

Scavenging residence time
(yr)

Element

Scavenging residence time
(yr)

Sn
Th
Fe
Co
Po
Ce
Pa
Pb

10
22-33
40-77
40
27-40
50
31-67
47-54

Mn
Al
Sc
Cu
Be
Ni
Cd
Particles

51-65
50-150
230
385-650
3700
15 850
117 800
0.365

From Chester.1

of Mn and Fe are released and subsequently are oxidized to produce
colloidal oxyhydroxides, which have a large surface area and very great
sorptive characteristics. Finally, ferromanganese nodules form at the
sediment-water interface and become considerably enriched in a number
of trace metals via surface reactions.
Ferromanganese nodules result from the authigenic precipitation of
Fe and Mn oxides at the seafloor. Two morphological types are
recognized, depending upon the growth mechanism. Firstly, spherical
encrustations produced atop oxic deep sea sediments grow slowly,
accumulating material from seawater. These seawater nodules exhibit a
relatively low Mn: Fe ratio and are especially enriched with respect to Co,
Fe, and Pb. Secondly, discoid shaped nodules develop in nearshore
environments deriving material via diffusion from the underlying anoxic
sediments. Such diagenetic nodules grow faster than deep sea varieties,
and metals tend to be in lower oxidation states. They have a high Mn:Fe
ratio and enhanced content of Cu, Mn, Ni, and Zn. Ferromanganese
nodules have concentric light and dark bands in cross-section, related to
Fe and Mn oxides, respectively. Patterns of trace element enrichment in
the nodules are determined by mineralogy, the controlling mechanisms
being related to cation substitution in the crystal structure. Mn phases
preferentially accumulate Cu, Ni, Mo, and Zn. Alternatively, Co, Pb, Sn,
Ti, and V are enriched in Fe phases.
3.3

Diagenesis

Diagenesis refers to the collection of processes that alters the sediments
following deposition. These mechanisms may be physical (compaction),
chemical (cementation, mineral segregation, ion exchange reactions), or

biological (respiration). The latter are of particular importance as the
bacteria control pH and ps in the interstitial waters, master variables that
affect a wide range of equilibria. They influence the composition of the
interstitial water, which in turn can exert a feedback effect on the
overlying seawater. Also, they can ultimately control the mineralogical
phases that are lost to the sediment sink.
Organic material accumulates with other sedimentary components at
the time of deposition. High biological activity in surface waters and
rapid sedimentation ensures that most nearshore and continental margin
sediments contain significant amounts of organic matter. Biochemical
oxidation of this material exhausts the available O2, creating anoxic
conditions. The oxic/anoxic boundary occurs at the horizon where the
respiratory consumption of O2 balances its downward diffusion. Upon
depleting the O2, other constituents are used as oxidants leading to the
step wise depletion of NO^, NO^", and SO 4 ". Thereafter, organic matter
itself may be utilized with the concurrent production of CH4.
This series of reactions causes progressively greater reducing conditions, with consequent influences on the chemistry of several elements.
Metals are reduced and so are present in lower oxidation states. In
particular, Mn undergoes reductive dissolution from MnO2(s) to M n ^ .
As the divalent state is much more soluble, Mn is effectively remobilized
under anoxic conditions and can be released back into overlying seawater. As seen in the previous section, this is one pathway to ferromanganese nodule formation. This can also be true for other elements that
had been deposited following incorporation into the Fe and Mn oxide
phases. In contrast, some elements can be preserved very effectively in
anoxic sediments. Interstitial waters in marine sediments, in contrast to
freshwater deposits, have high initial concentrations OfSO 4 ". Bacterial
sulfate reduction proceeds via the reaction:

Thus, sulfide levels in interstitial waters increase. A number of elements
form insoluble sulfides, which under these anoxic conditions are precipitated and retained within the sediments. A notable example is the
accumulation of pyrite, FeS2, but also Ag, Cu, Pb, and Zn are enriched in
anoxic sediments in comparison with oxic ones.
4 PHYSICAL AND CHEMICAL PROCESSES IN ESTUARIES
Rivers transport material in several phases: dissolved, suspended particulate, and bedload. Physical and chemical processes within an estuary
influence the transportation and transformation of this material, thereby

affecting the net supply of material to the oceans. Several definitions and
geomorphologic classifications of estuaries have been reviewed by
Perillo.25 From a chemical perspective, an estuary is most simply
described as the mixing zone between river water and seawater characterized by sharp gradients in the ionic strength and chemical composition. Geographic distinctions can be made between drowned river
valleys, fjords and bar built estuaries. They can alternatively be classified
in terms of the hydrodynamic regime as:
1.
2.
3.
4.

salt wedge;
highly stratified;
partially stratified;
vertically well mixed.

The aqueous inputs into the system are river flow and the tidal prism.
The series above is ranked according to the diminishing importance of
the riverine flow and the increasing marine contribution. Thus, a salt
wedge estuary represents the extreme case, dominated by river flow, in
which very little mixing occurs. A fresh, buoyant layer flows outward
over denser, saline waters. In contrast, the vertically well mixed estuary is
one dominated by the tidal prism. The inflowing river water mixes
thoroughly with and dilutes the seawater, but the effective dilution
diminishes with distance along the mixing zone.
The position of the mixing zone in the estuary exhibits considerable
temporal variations. There can be a strong seasonal effect, largely due to
non-uniform river discharge. High winter rainfall leads to a winter-time
discharge maximum. However, winter precipitation as snow creates a
storage reservoir, such that the river flow maximum occurs following
snow melt in spring or even early summer if the catchment area is of high
elevation as for the Fraser River in western Canada. On shorter time
scales, the mixing zone is influenced by tidal cycles. Thus, the penetration
of seawater into the estuary depends upon the spring-neap tidal cycle
and the diurnal nature of the tides. Together these influences determine
the geographic extent of sediments experiencing a variable salinity
regime. Variations in the river discharge affects the mass loading of the
discharge, both in terms of suspended sediment and bedload material.
The hydrodynamic regime in the estuary influences the deposition of the
riverine sedimentary material and the mixing of dissolved material.
The estuary is a mixing zone for river water and seawater, the
characteristics of which differ considerably. River water is slightly
25

G. M. E. Perillo, 'Geomorphology and Sedimentology of Estuaries', ed. G. M. E. Perillo, Elsevier
Science, Amsterdam, 1995, p. 17.

acidic and of low ionic strength with a salt matrix predominantly of
Ca(HCO3)2 (— 120 mg 1~ *). In contrast, seawater has a higher pH (~ 8),
higher ionic strength (~0.7) and consists primarily of NaCl (^- 35 g 1~x).
As a consequence, the salt matrix within the estuary is dominated by the
sea-salt end member throughout the mixing zone except for a small
proportion at the dilute extreme. Salinity can be used as a conservative
index, although conductivity is better, not being subject to systematic
conversion errors in the initial mixing region.
In a plot of concentration versus some conservative index {i.e. S%o,
Cl%o, or conductivity), the theoretical dilution curve would comprise a
straight line between the river and seawater end members. A dissolved
constituent that exhibits such a distribution is said to behave conservatively in the estuary. Whereas a negative slope shows that the riverine end
member is progressively diluted during mixing with seawater, a positive
slope indicates that the seawater end member has the greater concentration. Conservative behaviour is exhibited, for example, by Na + , K + ,
and SO4~. Reactive silica may at times behave conservatively. Nonconservative behaviour can result from an additional supply of material
(causing positive deviations from the theoretical dilution curve). Elements that may show a maximum concentration at some intermediate
salinity are Mn and Ba. Alternatively, the removal of dissolved material
during mixing {i.e. negative deviation from the theoretical dilution curve)
can be caused by biological activity or by dissolved to particulate
transformations. Biological activity can cause non-conservative behaviour for nutrient elements, including reactive silica. Dissolved constituents typically transformed to the particulate phase include Al, Mn, and
Fe (see Figure 12) in some estuaries. The pH distribution is usually
characterized by a pH minimum in the initial mixing zone, resulting from
the non-linear salinity dependence of the first and second dissociation
constants of H2CO3. Notwithstanding the obvious utility of componentconservative index plots, they can be applied and interpreted only with
caution. Often it is difficult to define the exact composition of the end
members. Hence, a plot that apparently denotes non-conservative
behaviour could arise if temporal fluctuations in the concentration of
the component of interest occur on the same relative time scales as
estuarine flushing.
A component can undergo considerable physico-chemical speciation
alterations in an estuary. With respect to dissolved constituents, the
composition and concentration of available ligands changes. Depending
upon the initial pH of the riverine water, OH ~~ may become markedly
more important down the estuary. Similarly, chlorocomplexes for metals
such as Cd, Hg, and Zn become more prevalent as the salinity increases.
Conversely, the competitive influence of seawater-derived Ca and Mg for

Dissolved Fe (yunot/kg)

Taieri estuary

Salinity
Figure 12

(%o)

The distribution of dissolved Fe versus salinity in the Taieri Estuary, New
Zealand; % 21 October 1980, O 4 December 1980
(From Hunter 26 )

organic material decreases the relative importance of humic complexation for Mn and Zn.
Estuaries are particularly well known for dissolved-particulate interactions. Phase changes come about via several mechanisms. Firstly,
dissolution—precipitation processes may occur. This is especially important for the authigenic precipitation of Fe and Mn oxyhydroxides.
Secondly, components may experience adsorption-desorption reactions.
Desorption can occur in the initial mixing zone, partly in response to the
pH minimum. Adsorption, particularly in association with the Fe and
Mn phases, can accumulate material within the suspended sediments.
Thirdly, flocculation and aggregation processes can remove material
from solution. This occurs as particulate material with negatively
charged surfaces adsorb cations in the estuary. The surface charge
diminishes and as the ionic strength increases, the particles experience
less electrostatic repulsion. Eventually the situation arises whereby
particle collisions lead to aggregation due to weak bonding. This
process can be facilitated if particle surfaces are coated with organic
material (and is then known as flocculation rather than aggregation).
The three types of processes outlined above often happen simultaneously
26

K. Hunter, Geochim. Cosmochim. Ada, 1983, 47, 467.

in the estuarine environment. Such non-conservative behaviour is
typified by dissolved Fe, as shown in Figure 12. The transformation of
dissolved into particulate phases can then be followed by deposition to
the estuarine sediments. Thus, the flux of material to the ocean can be
considerably modified, particularly as such sediments may be transported landward rather than seaward.
Biological activity in the estuarine environment can also influence the
speciation of constituents, notably dissolved-particulate partitioning. A
complex regeneration cycle determines distributions and modifies the
riverine flux. This is especially the case for nutrients, and estuaries are
often termed a nutrient trap. Estuaries tend to be regions of high
biological productivity as rivers have elevated nutrient concentrations.
Moreover, several freshwater organisms die upon encountering brackish
water with consequent cell rupture and the release of contents into
solution. Regardless of the source, the nutrients stimulate phytoplankton
productivity. Whereas some of the biological material (either transported via the freshwater or produced in situ within the estuarine zone)
can be flushed out to sea, the remainder settles out of the surface water to
be deposited onto the floor of the estuary. Respiration of this debris by
benthic organisms regenerates nutrients and any contaminants that have
been accumulated, releasing them into the bottom water. In estuaries
where a two-layer flow is well defined, these nutrients are transported
upriver in the salt wedge and entrained into the exiting river water,
thereby adding to the available nutrient pool. Polluted rivers having a
high nutrient loading are subject to eutrophication due to overstimulation of biological activity. Anoxic conditions within the bottom waters
and/or underlying sediments can result, depending upon the organic
loading. As mentioned previously (Section 2.2.2), some fjords develop
anoxic conditions when bottom waters stagnate due to limited mixing.
5 MARINE CONTAMINATION AND POLLUTION
Both contamination and pollution entail the perturbation of the natural
state of the environment by anthropogenic activity. The two terms are
distinguishable in terms of the severity of the effect, whereby pollution
induces the loss of potential resources.27 Additionally, a clear causeeffect relationship must be established for a substance to be classified as a
pollutant towards a particular organism. The human-induced disturbances take many forms but the greatest effects tend to be in coastal
environments due to the source strengths and pathways. Waters and
sediments in coastal regions bear the brunt of industrial and sewage
27

E. Goldberg, Mar. Pollut. Bull., 1992, 25, 45.

discharges, and are subject to dredging and spoil dumping. Agricultural
runoff may contain pesticide residues and elevated nutrients, the latter of
which may overstimulate biological activity producing eutrophication
and anoxic conditions. The deep sea has not escaped contamination. The
most obvious manifestations comprise crude oil, petroleum products,
and plastic pollutants, but includes the long-range transport of longlived radionuclides from coastal sources. Additionally, the aeolian
transport of heavy metals has enhanced natural fluxes of some elements,
particularly lead. Three case studies are introduced below to illustrate
diverse aspects of marine contamination and pollution.

5.1 Oil Slicks
Major releases of oil have been caused by the grounding of tankers (e.g.
Torrey Canyon, South-West England, 1967; Argo Merchant, Nantucket
Shoals, USA, 1976; Amoco Cadiz, North-West France, 1978; Exxon
Valdez, Alaska, 1989) or by the accidental discharge from offshore
platforms (e.g. Chevron MP-41C, Mississippi Delta, 1970; Ixtox /, Gulf
of Mexico, 1979). Because oil spills receive considerable public attention
and provoke substantial anxiety, oil pollution must be put into perspective. Crude oil has been habitually introduced into the marine environment from natural seeps at a rate of approximately 340 x 10 6 Iy" 1 .
Anthropogenic activity has recently augmented this supply by an order
of magnitude; however, most of this additional oil has originated from
relatively diffuse sources relating to municipal runoff and standard
shipping operations. Exceptional episodes of pollution occurred in the
Persian Gulf in 1991 (910 x 106 1 y" 1 ) and due to the Ixtox I well in the
Gulf of Mexico in 1979 (530 x 106 1 y" 1 ). In contrast to such mishaps,
the Amoco Cadiz discharged only 250 x 106 Iy" 1 of oil in 1978
accounting for the largest spill from a tanker. The cumulative pollution
from tanker accidents on an annual basis matches that emanating from
natural seepage. Nevertheless, the impacts can be severe when the
subsequent slick impinges on coastal ecosystems.
Regardless of the source, the resultant oil slicks are essentially surface
phenomena that are affected by several transportation and transformation processes.28 With respect to transportation, the principal agent for
the movement of slicks is the wind, but length scales are important.
Whereas small (i.e. relative to the slick size) weather systems, such as
thunderstorms, tend to disperse the slick, cyclonic systems can move the
slick essentially intact. Advection of a slick is also affected by waves and
28

S. Murray, 'Pollutant Transfer and Transport in the Sea', ed. G. Kullenberg, CRC Press, Boca
Raton, 1982, Vol. 2, p. 169.

currents. To a more limited extent, diffusion can also act to transport the
oil.
Transformation of the oil involves phase changes and/or degradation.
Several physical processes can invoke phase changes. Evaporation of the
more volatile components is a significant loss mechanism, especially for
light crude oil. Oil slicks spread as a buoyant lens under the influence of
gravitational forces, but generally separate into distinctive thick and thin
regions. Such pancake formation is due to the fractionation of the
components within the oil mixture. Sedimentation can play a role in
coastal waters when rough seas bring dispersed oil droplets into contact
with suspended particulate material and the density of the resulting
aggregate exceeds the specific density of seawater. Colloidal suspensions
can consist of either water-in-oil or oil-in-water emulsions, which behave
distinctly differently. Water-in-oil emulsification creates a thick, stable
colloid that can persist at the surface for months. The volume of the slick
increases and it aggregates into large lumps known as 'mousse', thereby
acting to retard weathering. Conversely, oil-in-water emulsions comprise
small droplets of oil in seawater. This aids dispersion and increases the
surface area of the slick, which can accelerate weathering processes.
Chemical transformations of oil are evoked through photochemical
oxidation and microbial biodegradation. Not only is the latter more
important in nature, but strategies can be adopted to stimulate biological
degradation, consequently termed bioremediation. All marine environments contain micro-organisms capable of degrading crude oil. Furthermore, most of the molecules in crude oils are susceptible to microbial
consumption. Oil contains little nitrogen or phosphorus so microbial
degradation of oil tends to be nutrient limited. Bioremediation often
depends upon the controlled and gradual delivery of these nutrients,
while taking care to limit the concurrent stimulation of phytoplankton
activity. Approaches that have been adopted are the utilization of slowrelease fertilizers, oleophilic nutrients and a urea-foam polymer fertilizer
incorporating oil-degrading bacteria. Bioremediation techniques were
successfully applied in the clean-up of Prince William Sound and the
Gulf of Alaska following the Exxon Valdez accident. Alternative
bioremediation procedures relying on the addition of exogenous bacteria
have still to be proved. Similarly, successful bioremediation of floating
oil spills has yet to be demonstrated.
Source apportionment of crude oil in seawater and monitoring the
extent of weathering and biodegradation constitute important challenges
in environmental analytical chemistry. As the concentration of individual compounds varies from one sample of crude oil to another, the
relative amounts define a signature characteristic of the source. Compounds that degrade at the same rate stay at fixed relative amounts

throughout the lifetime of an oil slick. Hence, a 'source ratio', which
represents the concentration ratio for a pair of compounds exhibiting
such behaviour remains constant. Conversely, a 'weathering ratio'
reflects the concentration ratio for two compounds that degrade at
different rates and consequently continually changes. Oil spill monitoring programmes conventionally determine four fractions:29
1.
2.
3.
4.

volatile hydrocarbons;
alkanes;
total petroleum hydrocarbons;
polycyclic aromatic hydrocarbons (PAHs).

The volatile hydrocarbons, albeit comparatively toxic to marine organisms, evaporate relatively quickly and hence serve little purpose as
diagnostic aids. The alkanes and total petroleum hydrocarbons make
up the bulk of the crude oil. They can be used to some extent for source
identification and monitoring weathering progress. The final fraction,
the PAHs, comprises only about 2% of the total content of crude oil but
includes compounds that are toxic. Moreover, these components exhibit
marked disparities in weathering behaviour due to differences in water
solubility, volatility and susceptibility towards biodegradation. As demonstrated in Figure 13, both a source ratio (C3-dibenzothiophenes:
C3-phenanthrenes) and a weathering ratio (C3-dibenzothiophenes:
C3-chrysenes) have been defined from amongst such compounds that
enable the extent of crude oil degradation to be estimated in the marine
environment, as well as for subtidal sediments and soils.29
5.2

Plastic Debris

The increasing accumulation of litter and debris along shorelines
epitomizes a general deterioration of environmental quality on the high
seas. The material originates not only from coastal sources, but also
arises from the ancient custom of dumping garbage from ships. Drilling
rigs and offshore production platforms have similarly acted as sources of
contamination. Some degree of protection in recent years has accrued
from both the London Dumping Convention (LDC) and the International Convention for the Prevention of Pollution from Ships
(MARPOL) which outlaw such practices. However, the problem of
seaborne litter remains global in extent and not even Antarctica has
been left unaffected.30
29

G . S. Douglas, A. E. Bence, R. C. Prince, S. J. McMillen and E. L. Butler, Environ. Sci. TechnoL,
1996, 30, 2332.
30
M. R. Gregory and P. G. Ryan, 'Marine Debris: sources, impacts and solutions', ed. J. M. Coe
and D. B. Rogers, Springer, New York, 1996, p. 49.

Increasing Oil Weathering

D3/C3 Weathering Ratio

North Sea,
mean • 0.80 ± 2 SD

Alaska North Slope crude,
mean =1.190 ± 2 SD

Iranian Crude
mean =2.48 ± 2 SD

D3/P3 Source Ratio
Figure 13 Plot of weathering ratio ( C3-dibenzothiophenes:C3-chrysenes ) versus source
ratio ( C3-dibenzothiophenes:C3-phenanthrenes) for fresh and degraded oil
samples from three different crude oil spills
(From Douglas et al )

The debris consists of many different materials, which tend to be nondegradable and endure in the marine environment for many years. The
most notorious are the plastics {e.g. bottles, sheets, fishing gear, packaging materials, and small pellets), along with glass bottles, tin cans, and
lumber. This litter constitutes an aesthetic eyesore on beaches, but more
importantly can be potentially lethal to marine organisms. Deleterious
impacts on marine birds and mammals result from entanglement and
ingestion. Lost or discarded plastic fishing nets remain functional and
can continue 'ghost fishing' for several years. This is similarly true for
traps and pots that go astray. Plastic debris settling on soft and hard
bottoms can smother benthos and limit gas exchange with pore waters.
Despite the negative effects of seaborne plastic debris, this material can
have positive consequences, serving as new habitats for opportunistic
colonizers.
5.3

Tributyltin

Tributyltin (TBT) provides an interesting case study of a pollutant in the
marine environment.31 Because TBT compounds are extremely poison31

'Tributyltin: case study of an environmental contaminant', ed. S. J. de Mora, Cambridge
University Press, Cambridge, 1996, p. 301.

ous and exhibit broad-spectrum biocidal properties, they have been
utilized as the active ingredient in marine anti-fouling paint formulations. Its potency and longevity ensures good fuel efficiencies for ship
operations and guarantees a long lifetime between repaintings. TBTbased paints have been used on boats of all sizes, from small yachts to
supertankers, ensuring the global dispersion of TBT throughout the
marine environment, from the coastal zone to the open ocean.
Notwithstanding the benefits, TBT's extreme toxicity and environmental persistence has resulted in a wide range of deleterious biological
effects on non-target organisms. TBT is lethal to some shellfish at
concentrations as low as 0.02 ^g TBT-Sn I" 1 . Lower concentrations
result in sub-lethal effects, such as poor growth rates and reduced
recruitment leading to the decline of shellfisheries. The most obvious
manifestations of TBT contamination have been shell deformation in
Pacific oysters (Crassostrea gigas), and the development of imposex (i.e.
the imposition of male sex organs on females) in marine gastropods. The
latter effect has caused dramatic population decline of gastropods at
locations throughout the world. Laboratory experiments and field
observations of deformed oysters and imposex indicate that adverse
biological effects occur at concentrations below those detectable. Thus, a
'no-effect' concentration has yet to be demonstrated. TBT has been
observed to accumulate in fish and various marine birds and mammals,
with as yet unknown consequences. Although it has not been shown to
pose a public health risk, one study reported measurable butyltin
concentrations in human liver.32
The economic consequences of the shellfisheries' decline led to a rapid
political response globally. The first publication suggesting TBT to be
the causative agent only appeared in 1982,33 but already the use of TBTbased paints has been banned in some countries, including New Zealand.
Other nations have imposed partial restrictions, its use being permitted
only on vessels > 25 m in length or on those with aluminium hulls and
outdrives. This has certainly had the effect to decrease the TBT flux to
the marine environment as manifested in sedimentary TBT profiles.
Oyster aquaculture in Arcachon Bay benefitted immediately, with a
notable decline in shell deformations and TBT body burdens and the
complete recovery of production within two years (Figure 14). Comparable improvements in oyster conditions have been reported for Great
Britain and Australia. Similarly, there have been many reported
instances of restoration of gastropod populations at previously impacted
locations. However, large ships continue to act as a source of TBT to the
32
33

K. Kannan and J. Falandysz, Mar. Pollut. Bull, 1997, 34, 203.
C. Alzieu, M. Heral, Y. Thibaud, M. J. Dardignac, and M. Feuillet, Rev. Trav. Inst. Marit., 1982,
45, 100.

Oyster Production (tons)

Year
Figure 14 Annual oyster (Crassostrea gigas) production in Arcachon Bay, 1978-85;
restrictions on TBT use were applied starting January 1982
(Data from Alzieu34)

marine environment. It should be of concern that imposexed gastropods
have been observed at sites {e.g. North Sea and Strait of Malacca) where
the source of TBT can only be attributed to shipping.
TBT exists in solution as a large univalent cation and forms a neutral
complex with Cl" or OH~. It is extremely surface active and so is readily
adsorbed onto suspended particulate material. Such adsorption and
deposition to the sediments limits its lifetime in the water column.
Degradation, via photochemical reactions or microbially mediated pathways, obeys first order kinetics. Several marine organisms, as diverse as
phytoplankton to starfish, debutylate TBT. Stepwise debutylation produces di- and mono-butyltin, which are much less toxic in the marine
environment than is TBT. As degradation rates in the water column are
of the order of days to weeks, they are slow relative to sedimentation.
TBT accumulates in the sediments where degradation rates are much
slower, with the half-life being of the order of years.35 Furthermore,
concentrations are highest in those areas, such as marinas and harbours,
which are most likely to undergo dredging. The intrinsic toxicity of TBT,
its persistency in the sediments and its periodic remobilization by
anthropogenic activity are likely to retard the long-term recovery of the
marine ecosystem.
34
35

C. Alzieu, Mar. Environ. Res., 1991, 32, 7.
C. Stewart and S. J. de Mora, Environ. Technoi, 1990, 11, 565.

Recapitulating, the unrestricted use of TBT has ended in many parts
of the world but significant challenges remain. For the most part, the
coastal tropical ecosystems remain unprotected and the sensitivity of its
indigenous organisms relatively poorly evaluated. TBT endures in
sediments globally, with concentrations usually greatest in environments
most likely to be perturbed. The widespread introduction of TBT into
seawater continues from vessels not subject to legislation. Organisms in
regions hitherto considered to be remote now manifest TBT contamination and effects. Such observations imply further restrictions on the use
of organotin-based paints are required. However, the paramount lesson
learned from TBT should be that potential replacement compounds
must be properly investigated prior to their introduction in order to
avoid another global pollution experiment.
QUESTIONS
1. What are the main sources and sinks for dissolved and particulate
metals entering the ocean?
2. Define residence time and outline the factors that influence the
residence time of an element in the ocean. Provide examples of
oceanic residence times for elements that span the time scale.
3. Define salinity. What are the notable variations in sea-salt, in
terms of concentration and composition, in the world ocean?
4. Using data available in Table 4, calculate the concentration of
H2CO3 in seawater at 50C in equilibrium with atmospheric CO2 at
370 ppm.
5. Calculate the speciation of H2S in seawater at pH8.3 at 25 0C
given that pKx = 7.1 and pK2 = 17.0.
6. Explain how DMS of marine origin might affect global climate.
7. What are the major nutrients in the ocean? Describe their
concentration profiles and account for any differences according
to chemical behaviour.
8. Describe the different types of depth profile that various metals
exhibit and explain how differences in their profiles originate.
9. Assuming that 7 = 1 for all species, calculate the speciation of lead
in seawater at 25 0C given the following values for stability
constants for the chlorocomplexes (K1 = 10782, P2 = 101088,
P3= 101394J4= 1016-30). Note that [CP] is 0.559 m m o i r 1 .
10. With specific reference to question 9, outline the potential limitations of using equilibrium models to explain chemical behaviour
of trace metals in seawater.
11. Describe the classification of marine sediments and give examples
of each sediment type.

12. Why do suspended particles exhibit a surface charge and how
could this characteristic moderate the composition of an ocean's
waters?
13. With respect to estuarine chemistry, describe conservative and
non-conservative behaviour and provide examples of cationic and
anionic species for each category.
14. What natural processes are responsible for weathering an oil slick?
15. Why is tributyltin considered such a potent pollutant in the
marine environment?

CHAPTER 5

Land Contamination and Reclamation
B. J. ALLOWAY

1 INTRODUCTION
In many parts of the world, especially in the more technologically
developed countries, soil, which forms the surface layer of the land, has
been exposed to varying degrees of contamination from a wide variety of
chemicals for many years, especially since the Second World War. Soil
constitutes one of the three key environmental media, the other two
being air and water, but the soil differs significantly from these two media
because it is predominantly solid and has the capacity to retain many
types of pollutants. These retention (sorption) mechanisms cause soil to
function as sinks for contaminants. In this way soil acts as a filter, which
reduces or prevents contaminants reaching the groundwater. Furthermore, the diverse populations of micro-organisms make soils powerful
bioreactors which can degrade many hazardous organic chemical contaminants in addition to the normal biogeochemical cycles which are a
vital component of terrestrial ecosystems.
Some contaminants which are sorbed and not degraded will gradually
accumulate and could reach concentrations that are potentially harmful
to the functioning of the soil-plant system or to consumers of crops
which take up these chemicals from contaminated soils. In contrast,
pollutant concentrations in water and air become diluted through
dispersion in these fluid media and, although larger volumes of these
media can be affected by pollution, the effect is usually of limited
duration.
The range of chemicals found contaminating soils is vast and can
include, for example: heavy metals, such as lead from paints and motor
vehicle exhausts, polycyclic aromatic hydrocarbons (PAHs) formed
during the incomplete combustion of organic materials such as wood
and coal, synthesized organic chemicals such as chlorinated solvents or

pesticides from leakages or direct application, radionuclides in fallout
from the testing of nuclear weapons or accidents at nuclear power
stations, and contaminants (such as Cd) reaching agricultural land as
impurities in fertilizers and manures. Some of the most contaminated
land is found at the sites of former or current industrial operations.
These can include metalliferous mines and smelters, chemical works,
gas manufacturing plants, petroleum refineries, and scrap yards. At
these types of sites, the contaminants may be raw materials, processrelated chemicals such as catalysts and electrodes, manufactured
chemical products, and a wide range of wastes.
The distinction between 'pollution' and 'contamination' is used by
some authors to indicate the relative severity of adverse effects. 'Contamination' is sometimes used for a situation where a substance resulting
from human activity is present in the environment but not causing any
obvious harmful effects. In contrast, the more pejorative term 'pollution'
is often used when a substance is present and having a harmful effect.
Unfortunately, although this distinction is very convenient, it often does
not hold true when more detailed studies are made of contamination in
soils or other components of the environment. Although not so obvious,
toxic or other harmful effects can often be found in situations which have
been previously classed as contamination. However, it is the convention
to refer to all cases of pollution or contamination of soils/land as
contamination. Therefore 'contaminated land' covers the whole range
from the presence of low concentrations of a chemical from an extraneous source right through to a case of severe toxicity in plants and/or
hazards to animals and humans.
In the UK, contaminated land is defined as 'land which represents an
actual or potential hazard to health or the environment as a result of
current or previous use'. A more severe case is presented by 'derelict
land' which is defined as 'land so damaged by industrial or other
development that it is incapable of beneficial use without treatment'.
Although it has been customary to distinguish between contamination
and pollution, the former being where a substance is recognized as being
present but not causing a recognizable problem and the latter where a
problem (usually toxicity) is recognized, it is the convention in reference
to land that the term 'contaminated' is used for all situations. It has been
estimated that around 40,000 ha in England and Wales constitute
'contaminated land' and that almost all the land surface in the UK is
contaminated to at least a slight extent. Atmospheric deposition of
contaminants is ubiquitous although it will be of higher magnitude
close to, or downwind of, major sources, but soils become contaminated
by other means including direct application of materials containing
contaminants and by flooding.

In view of the widespread occurrence of land contamination, it is
important that soils at sites suspected of being contaminated are sampled
appropriately and analysed to determine the nature of the contaminants
and the concentrations involved in order that a risk assessment can be
made. In addition to the risk of toxicity to ecosystems, crops, livestock,
and humans, damage to materials and services are also important
considerations. These include the sulfate attack of concrete, degradation
of plastics used in piping and insulation of electric cables by hydrocarbons of various types, and the risk of fires or explosions from gases
and other combustible materials. The risks to humans include respiratory and other disorders from the inhalation of toxic fumes and dusts,
direct ingestion of toxic compounds in contaminated drinking water or
on particles of soil, absorption of chemicals through the skin, and the
consumption of food plants which have accumulated excessive concentrations of potentially harmful substances. In general, severe contamination, where acutely toxic conditions prevent the growth of most types of
food crops and other vegetation, presents less of a health risk (with
regard to ingestion) than conditions where a significant accumulation of
potentially toxic substances such as cadmium by food crops can occur
without causing obvious toxic symptoms. In the latter situation, crops
are likely to be consumed without the chronic toxicity risk being
appreciated, whereas crop plants which die or display symptoms of
stress are less likely to be consumed.
In addition to contamination, which occurs as a result of human
activity (anthropogenic), anomalously high concentrations of potentially
harmful inorganic substances can occur in soils as a result of geochemical
processes without any human involvement. Although not strictly 'contamination' or pollution, these geochemical anomalies can cause toxicity
and material degradation problems. Examples include: marine black
shale rocks which can contain relatively high concentrations of harmful
elements such as cadmium, lead, uranium, and zinc and sulfides.
This chapter aims to show the breadth of the subject of soil contamination and its reclamation and to introduce some of the fundamental
concepts relating to the behaviour of contaminants in soils. However, the
space available does not permit the presentation of much detail, and the
reader is recommended to consult other more specialist texts for further
information.
2 SOIL: ITS FORMATION, CONSTITUENTS, AND
PROPERTIES
Soil is the complex biogeochemical material which forms at the interface
between the earth's crust and the atmosphere, and differs markedly in

physical, chemical, and biological properties from the underlying weathered rock from which it has developed. Soil comprises a matrix of
mineral particles and organic material, often bound together as aggregates, and populated by a wide range of micro-organisms, soil animals,
and plant roots. The spaces between particles and aggregates form a
system of pores, which are filled with aqueous soil solution and gases.
The physical, chemical, and biological properties of soils are highly
relevant to considerations of contaminated land and its reclamation.

2.1

Soil Formation

The soil profile (a vertical section from the surface to the underlying
weathered rock) is the unit of study in pedology (which is the study of the
origin, occurrence and classification of soils). Natural pedogenic processes bring about the differentiation of the soil material into distinct
horizontal layers, called horizons, and the characteristics and assemblages of these horizons forms the basis of soil morphological classification. A typical soil profile with a summary of the distribution of
contaminants is shown in Figure 1.
The type of soil that forms at any site is determined by climate, the
nature of the parent rock material on which it forms, the landscape

Horizon
Litter
Organic matter
Organo-mineral A
horizon

Desiccation crack
Worm
Behaviour of contaminants
burrow Atmospheric deposits and elements
O cycled through plants accumulate here
Accumulation of metal and organic
A contaminants after humification of litter

Eluvial horizon E

Root E Some contaminants adsorbed on clay
channel particles moved down profile

Illuvial horizon B

ation of contaminants adsorbed
B Accumul
on clays

Parent material C

Contaminants in this region are either highly
C soluble or moved down in particles which
fall down desiccation cracks

Bedrock R
Figure 1 Diagram of a soil profile showing horizon nomenclature and the general
distribution of contaminants

(especially whether it is a free-draining or water collecting site), the
vegetation, and the time period over which the soil has been forming.
These factors control the intensity and type of pedogenic processes
operating at any site and determine the nature of the soil profile that
develops. The major pedogenic processes include: leaching, elution,
podzolization, calcification, ferralitization, laterization, salination, solodization, gleying, and peat formation and several of these usually
interact in any soil although one is visually predominant.1 With the
exception of the last two, most pedogenic processes involve the vertical
movement of solutes and material (usually the products of weathering)
either down or up the soil profile, depending on the overall balance of
precipitation and evaporation. Movement is predominantly downward
in humid regions and upward in arid and semi-arid areas where
evaporation exceeds precipitation. Gleying is the creation of reducing
conditions in soils, due to intermittent or permanent waterlogging, and
leads to the reduction and dissolution of oxides of Fe and Mn.
The pedogenic processes occurring within a soil will have a major
influence on the behaviour of contaminants. For example, processes
involving leaching down the profile will tend to transport soluble
contaminants toward the water table. However, this behaviour is
modified by reactions of the contaminant chemicals with the solid
matrix of the soil. Soil organic matter, especially colloidal humus, has a
strong capacity to sorb a wide range of substances, including heavy
metals and many different types of organic contaminants.
2.2

Soil Constituents

It is important to note that true (pedological) soils have many features in
common, and several predictions can be made about the behaviour of
contaminants within them. However, in some types of contaminated
land, especially that at long-established industrial sites (either active or
derelict), the material forming the ground under and around buildings
may differ markedly from a pedological soil. This is because many
industrial sites have undergone continual development where manufacturing processes, products, and wastes have changed over many years.
Frequently this involved demolishing old buildings and using old
construction material and wastes as foundations for new operations.
Thus, much of the material in the top 2 or 3 metres may comprise
building rubble, underground pipes and tanks as well as contaminant
chemicals. It is essentially 'made land' and bears little similarity to
undisturbed (pedological) soils. However, although some of these con1

E. M. Bridges, 'World Soils', Cambridge University Press, Cambridge, 1978.

stituents may differ from the normal range of soil minerals and humus,
the principles affecting the physical and chemical behaviour of contaminants are still the same although it will be necessary to measure the
properties of these 'non-soil' constituents at each site. Soil containing
large amounts of concrete or cement rubble may be highly permeable
and so prone to leaching, and it is also likely to have a high pH from the
alkaline constituents of the cement. The alkaline cement is itself a soil
contaminant.
2.2.1 The Mineral Fraction. The soil clay fraction comprises platelike particles < 2 pan in diameter and is formed mainly of clay minerals
with some small particles of other minerals such as Fe, Al, and Mn
oxides. Clay minerals are secondary minerals which have been synthesized from the weathering products of primary minerals. Clay minerals
are sheet silicates formed from two basic components: a silica sheet
formed of Si—O tetrahedra, and a gibbsite sheet comprising Al—OH
octahedra. Isomorphous substitution of either Al 3 + for Si 4 + , or Mg2 +
for Al3 + in the structural sheets of some clay minerals gives rise to excess
negative charges on their surface. This permanent excess negative charge
is independent of the soil pH and provides clay minerals with their ability
to adsorb cations. This adsorption on clay minerals and organic matter is
referred to as cation exchange and a soil's capacity for this is referred to
as its cation exchange capacity (CEC) which is measured in centimoles
charge per kg (cmolc k g " l = 10~2 mol cation exchange capacity per kg
of soil). The most commonly occurring types of clay minerals are:

Kaolinite

—a highly stable 1:1 (silica: gibbsite) mineral, which is
non-swelling with a relatively small surface area (5100m 2 g~ 1 ) and a low adsorptive capacity (cation
exchange capacity [CEC] 2-20 cmolc kg~ l ).
Smectites —are 2:1 (silica:gibbsite) minerals which swell on wetting
and shrink on drying, they have a large surface area
(700-800 m2 g" 1 ) due to the access of soil solution to all
lamellae surfaces which, together with isomorphous
substitution, contributes to their relatively high adsorptive capacity (CEC 80-120 cmolc kg" 1 ). Some of the
better known members of this group of minerals include
montmorillonite and bentonite.
Mite
—a 2:1 clay like the smectites with a lower adsorptive and
swelling/shrinking capacity and properties intermediate
between kaolinite and smectites (surface area 100—
200 m2 g" \ a CEC 10-40 cmolc kg" *).

Vermiculite—a 2:1 mineral with a very high degree of isomorphous
substitution of Mg 2+ for Al 3+ in the gibbsite sheet
giving it a high CEC (100-150 cmolc kg" *).
Other ubiquitous secondary minerals in soils include oxides of Fe and Al
which are the ultimate residual weathering products of soils. They are
high stable under the oxidizing conditions found in freely draining soil
environments and the predominant brown colour of soils is due to the
presence of Fe oxides. The amount of these oxides present in a soil
depends on the mineralogy of the parent rock, the degree of weathering,
and the oxidation-reduction (redox) conditions. The ferromagnesian
minerals such as olivine, augite, and biotite mica found in basic igneous
rocks tend to be the richest primary sources of Fe. Poorly drained and
waterlogged soils have reducing conditions which cause the dissolution
of oxides and tend to be greyish in colour. The main forms of Fe oxide
are ferrihydrite (Fe2O3.2FeOOHL^H2O) the freshly deposited form, and
goethite (a-FeOOH). The Al oxides include amorphous Al(OH)3 which
slowly crystallizes to gibbsite (7-Al(OH)3). These oxides tend to occur as
precipitates within the clay-sized fraction of soil minerals (<2 /mi). The
hydrous oxides have pH-dependent surface charges; in general they are
positively charged under acid conditions and negatively charged under
alkaline conditions. Manganese oxides have important adsorptive properties but are much less abundant than those of Fe and Al.
Many other minerals may be found in soils, the most ubiquitous being
resistant quartz grains which form much of the sand-sized fraction of
soils and are relatively unreactive. Calcite (CaCO3) is a major soil
mineral constituent in soils developed on limestones in humid areas and
in many soils in semi-arid regions. It buffers the soil to a pH usually
between 7 and 8.5 and also contributes to the sorptive properties of soils
by chemisorption and precipitation of metal carbonates (e.g. CdCO3).
Traces of incompletely weathered primary minerals can also be found in
soils.
2.2.2 Soil Organic Matter. The presence of organic matter and living
organisms is a characteristic of soil and helps distinguish it from regolith
(weathered rock). All soils, in the pedological sense of the word, contain
organic matter in their surface horizons but the amount and type may
vary considerably. Soils in arid environments tend to contain only small
amounts of organic matter (< 1%), whereas those in humid temperate
regions have higher contents (1-10%). The organic matter present in
soils can be classified as either humic or non-humic. Humic compounds
are highly polymerized, colloidal, products of microbial decomposition
of plant material, especially lignins. Non-humic material includes un-

decomposed or partially decomposed fragments of plant tissues and soil
organisms. These latter include a wide range of species of bacteria, fungi,
protozoa, actinomycetes, and algae, and many species of mesofauna,
such as earthworms, which fill an important ecological niche in the
comminution of plant litter and its incorporation into the soil.
Although organic matter only tends to form a small percentage of the
mass of the soil (usually less than 10%), it has a very great influence on
soil chemical and physical properties especially with regard to the
behaviour of contaminants. Humus has a high, but pH-dependent,
CEC (up to 200 cmolc kg" 1 ) and a strong sorptive capacity for many
different types of contaminants, including non-polar organic molecules
and heavy metals.2
2.3

Soil Properties

Space does not permit a detailed discussion of the full range of soil
properties here and, therefore, only those properties which have a
significant effect on the behaviour of soil contaminants will be discussed.
Readers are recommended to consult other specialized text books (e.g.
References 3-5) for more details of the chemical, physical, and biological
properties of soils.
2.3.1 Soil Permeability. The voids between soil particles and aggregates form a continuous system of pores within the soil profile. Pores
with a diameter greater than 30 jam tend to drain under gravity and are
normally filled with air in dry weather. Pores smaller than 30 jam
diameter tend to retain water against gravity and much of this may be
available to plant roots. Soluble contaminants infiltrating the soil profile
will be subject to internal drainage within the profile, diffusion between
regions of different solute concentration and, in most cases, adsorption
onto the organo-mineral colloidal complex. In addition to the interparticle pores, larger (macropores) also exist; these are created by worm
burrows, root channels, desiccation cracks, or deep soil disturbance and
lead to more rapid movement of contaminants down the profile, either in
solution, or adsorbed on suspended soil particles.
Dense non-aqueous phase liquids (DNAPLs) such as solvents tend to
infiltrate relatively rapidly through soil profiles and reach the groundwater. The permeability of soil can vary widely as a result of the nature of
2

S. J. Ross, 'Soil Processes: A Systematic Approach', Routledge, London, 1989, p. 444.
N . Brady, 'The Nature and Property of Soils', Macmillan, New York, 10th Edn., 1990.
4
D . L. Rowell, 'Soil Science: Methods and Applications', Longman Scientific and Technical,
Harlow, 1994, p. 350.
5
R. E. White, 'Introduction to the Principles and Practice of Soil Science', Blackwell, Oxford, 3rd
Edn., 1997.
3

the soil constituents, the effective diameter of the system of pores, and
soil management. Clay-rich soils tend to have relatively low permeabilities whereas sandy or fissured soils are more permeable. Compaction of
soils by machinery, cultivations, or excavations when the soil is too wet,
can reduce its permeability and may lead to waterlogging and the onset
of anoxic conditions.
2.3.2 Soil Chemical Properties
pH. The pH of the soil is the most important physico-chemical
parameter affecting plant growth and the behaviour of ionic contaminants in soils. Soil pH is a measurement of the concentration of H + ions
in the soil solution present in the soil pores which is in equilibrium with
the negatively charged surfaces of the soil particles. Unless stated
otherwise, soil pH values are usually measured in distilled water but the
use of dilute electrolytes {e.g. CaCl2) more closely reflects the field
situation.
Soil pHs are normally within the range 4-8.5 although the extreme
range found over the world is 2-10.5. In general, soils in humid regions
tend to have pH values between 5 and 7 and those in arid regions
between 7 and 9. The median pH for more than 3000 soils from all over
the USA was reported to be 6.1 (range 3.9-8.9)6 and for nearly 6000
samples in the UK the median pH was 6.0 (3.1-9.2).7 In temperate
regions, such as the UK, the optimum pH for arable soils is 6.5 and for
grassland it is 6.0. Soil pH conditions can be raised by liming with
CaCO3 and can be reduced by applying elemental sulfur although the
latter is rarely practised.
In general, bacteria do not tolerate very acid conditions, so if it is
intended to promote the microbial degradation of organic contaminants the soil should be maintained at a pH between 6 and 8. The
mobility and bioavailability of most divalent heavy metals are greater
under acid conditions and therefore liming is a way of reducing their
bioavailability (except for Mo because the TsAoO\~ anion is more
available at high pH).
Redox conditions. The balance of oxidation-reduction conditions in
soils mainly affects the speciation of elements such as C, N, O, S, Fe, and
Mn although Ag, As, Cr, Cu, Hg, and Pb are also affected. The redox
conditions in a soil also reflect the oxygen supply for plant roots and soil
micro-organisms. Respiration by plant roots, soil fauna and microorganisms consumes a relatively large amount of oxygen. In situations
6

C . G. S. Holmgren, M. W. Meyer, R. L. Chaney, and R. B. Daniels, J. Environ. Qual., 1993, 22,
335-348.
7
S . P. McGrath and P. Loveland, 'Soil Geochemical Atlas of England and Wales', Blackie
Academic and Professional, Glasgow, 1992.

of waterlogging, or exclusion of air by over compaction, micro-organisms with anaerobic respiration predominate, causing a change in the
products of decomposition of organic matter (volatile fatty acids and
ethylene, etc.) and in the speciation of susceptible metals. Microbially
mediated methylation of some metallic and metalloid contaminants
including As, Hg, Sb, Se, and Tl, tends to occur under anoxic conditions
and can be an important mechanism for both the loss of the element
from the soil and its conversion to more toxic and bioavailable forms.
Redox equilibria are controlled by the aqueous free electron activity
and can be expressed either as ps (negative log of electron activity) or
En (millivolt difference in potential between a Pt electrode and a H
electrode). The conversion factor for the two expressions is: £"H (niV) =
5.2 p £ . 8
The combined effect of redox and pH on the behaviour of hydrous
oxides in soils can be summarized by stating that either low pH or
negative redox values generally result in the dissolution of oxides of Fe
and Mn and, conversely, small increases in either pH or redox values can
lead to the precipitation of ferrihydrite or Mn oxides. Sulfate anions are
reduced to sulfide below pe — 2.0 and this can lead to the formation of
sulfides of a wide range of metals. These sulfides tend to be insoluble and
act as a temporary sink of the metal until redox conditions change and
the sulfides are oxidized. The oxidation of sulfides, such as iron pyrites
(FeS2) results in a marked increase in the acidity of the soil.
The adsorption of contaminants in soils. The concentration of solutes in
the soil solution is determined by the interaction of adsorption and
desorption processes. When contaminants are deposited on the soil
surface they either react with the colloids in the soil aggregates at the
surface or are washed into the soil profile in rain, irrigation water, or
snow melt. Soluble contaminants will infiltrate the topsoil and enter the
systems of pores, whereas insoluble and hydrophobic organic molecules
may either bind to sites on the soil surface and become incorporated into
the topsoil by movement of soil particles during cultivation, or, in the
case of dense non-aqueous phase liquids (DNAPLs) move rapidly down
the profile through large pores and fissures.
Several different adsorption reactions can occur between the surfaces
of organic and mineral colloids and the contaminants. The extent to
which the reactions occur will be determined by the composition of the
soil (especially the clay mineral, Fe and Mn oxide, and organic matter
contents), the soil pH, and the nature of the contaminants. The more
strongly adsorbed contaminants are less likely to be leached down the
soil profile and will tend to have lower bioavailabilities. Ionic contami8

W. L. Lindsay, 'Chemical Equilibria in Soils', John Wiley & Sons, Chichester, 1979, p. 449.

nants such as metals, inorganic anions, and certain organic molecules
including the bipyridyl herbicides Paraquat and Diquat are adsorbed
onto surface charges on soil colloids. Non-polar organic molecules,
which includes most of the persistent organic pollutants and pesticides
are adsorbed onto humic polymers by both chemical and physical
adsorption mechanisms.
Adsorption of cations and anions in soils. Ion exchange (or non-specific
adsorption) refers to the exchange between the counter-ions balancing
the surface charge on the soil colloids and the ions in the soil solution. In
the case of cations, it is the negative charges on soil colloids that are
responsible for adsorption.
Soil organic matter has a higher CEC at pH 7 than other soil colloids
and therefore plays a very important part in all adsorption reactions in
most soils, even though it is normally present in much smaller amounts
than clays. However, sandy soils with low contents of both organic
matter and clay minerals tend to have low adsorptive capacities and are a
greater danger for contaminants infiltrating through the soil profile to
the water table.
The negative charges on the surfaces of soil colloids are of two types:
(a) permanent charges resulting from the isomorphous substitution of a
clay mineral constituent by an ion with a lower valency, and (b) the pH
dependent charges on oxides of Fe, Al, Mn, and Si and organic colloids
which are positive at pH values below their isoelectric points and
negative above their isoelectric points. Hydrous iron and aluminium
oxides have relatively high isoelectric points (pH 8) in vitro but lower
values within the soil system (ca. pH 7) and so tend to be positively
charged under most conditions, whereas clay and organic colloids are
predominantly negatively charged under alkaline conditions. With most
colloids, increasing the soil pH, at least up to neutrality, tends to increase
their CEC. Humic polymers in the soil organic matter fraction become
negatively charged due to the dissociation of protons from carboxyl and
phenolic groups.
The concept of cation exchange implies that ions will be exchanged
between the soil solution and the zone affected by the charged colloid
surfaces (double diffuse layer). The relative replacing power of any ion
on the cation exchange complex will depend on its valency, its diameter
in hydrated form, and the type and concentration of other ions present in
the soil solution. With the exception of H + , which behaves like a
trivalent ion, the higher the valency, the greater the degree of adsorption.
Ions with a large hydrated radius have a lower replacing power than ions
with smaller radii. For example, K + and Na + have the same valency but
K + will replace Na + owing to the greater hydrated size of the Na + ion.

The commonly quoted relative order of replaceability of metal cations
on the cation exchange complex is:

Anion adsorption occurs when anions are attracted to positive charges
on soil colloids. As stated above, hydrous oxides of Fe and Al are usually
positively charged below pH 7 or 8 and so tend to be the main sites for
anion exchange in soils. In general, most soils tend to have far smaller
capacities for anion exchange than for cation exchange. Some anions,
such as nitrates and chlorides are not adsorbed to any marked extent but
others, such as orthophosphates tend to be strongly adsorbed.
Specific adsorption is a stronger form of adsorption, involving several
heavy metal cations and most anions with surface ligands in forming
partly covalent bonds with lattice ligands on adsorbents, especially
oxides of Fe, Mn and Al. This adsorption is strongly pH-specific and
the metals and anions which are most able to form hydroxy complexes
are adsorbed to the greatest extent. The order for the increasing strength
of specific adsorption of selected heavy metals is:

In addition to cation exchange reactions, soil organic matter can
sequester metals by complexation mechanisms, especially chelation. The
general order of complexation of metals is:

2.3.5

Adsorption and Decomposition of Organic Contaminants.

The

adsorption of non-ionic and non-polar organic contaminants (including
many pesticides) occurs mostly on soil humic material. Since the highest
content of organic matter occurs in the surface horizons of soils, there is
a tendency for most organic contaminants to be concentrated in the
topsoil. Migration of organic contaminants down the profiles of soils
occurs to the greatest extent in highly permeable sandy, gravelly, or
fissured soils with low organic matter contents. High concentrations of
dissolved organic compounds (DOC) of humic origin can cause
enhanced mobility and leaching of organic (and metallic) contaminants
in soils due to the binding of the contaminant to the soluble ligand. Soils
vary in their contents of DOC but applications of sewage sludge, animal
manure, and composts result in increased concentrations.

In the case of organic pesticides, most are relatively insoluble and do
not move down the soil profile, but exceptions to this are the organophosphate insecticides, phenoxyacetic acid herbicides, and bipyridylium
herbicides, which can be leached. The phenoxyalkanoic acid herbicides,
such as 2,4-D, 2,4,5-T, and MCPA, exist as anions at normal values of
soil pH and are adsorbed to a limited extent by hydrous oxides and by
hydrogen bonding to humic polymers. However, this sorption is less
marked than that which occurs with the cationic bipyridyl herbicides
(paraquat and diquat) which are inactivated on contact with soil
colloids.2
Many pesticide and other organic contaminants (see Section 4.2) that
reach water courses from soils have often been washed into the water
adsorbed onto soil particles in runoff and not leached down through the
profile.2 In some cases where organic wastes containing organic contaminants, such as sewage sludges and some composts, are applied to soils,
they act as both a source and a sink of contaminants (e.g. PAHs, PCBs,
and heavy metals). Positive correlations are frequently found between the
concentrations of pesticide residues and the soil organic matter content.
The propensity for a non-polar organic contaminant molecule to be
sorbed by soil organic matter can be predicted by determining its
octanol-water distribution coefficient (A^ow).9 Substances with low K0^
values (< 10) tend to be hydrophilic (water soluble) whereas those with
higher values are more hydrophobic and more strongly bound to organic
matter (and lipids in living organisms). For example, the herbicide 2,4-D
(2,4-dichloro-phenoxyacetic acid) is relatively soluble in water
(900 mg P 1 ) and has a low octanol-water distribution coefficient (log
Kow = 1.57), whereas the insecticide DDT has a low water solubility
(0.002 mg P 1 ), a high log A^ow value (5.98), and accumulates in soil
organic matter and lipids in living organisms.
Organic contaminants may be lost from the soil by physical processes
including volatilization, photolytic degradation, and leaching, and by
microbial degradation. Soil micro-organisms can become adapted to
degrade many different types of persistent contaminant molecules by the
secretion of extracellular enzymes. However, the C—Cl bond does not
occur in naturally occurring compounds and hence the chlorinated
organic molecules are more difficult to degrade. Nevertheless, several
species of micro-organism have been found to have developed the ability
to degrade chlorinated compounds, but there is normally a time lag while
they adapt to the new molecules. During this period there is localized
selection and multiplication of strains that have a mutation enabling
them to break the bonds of the contaminant molecule and also to
9

M. D. LaGrega, P. L. Buckingham, and J. C. Evans, 'Hazardous Waste Management', McGrawHill, New York, 1994, p. 1146.

tolerate the toxicity of either the contaminant or its intermediate
decomposition products.
Adsorption of non-polar organic contaminants onto soil organic
matter will not occur in the presence of oils. Microbially synthesized
surfactants can help to accelerate the rate of degradation of hydrocarbon
oils in contaminated soils. For many organic contaminants, adsorption
onto soil colloids and the presence of water are important factors
promoting decomposition by micro-organisms. Ross lists the types of
degradation of organic molecules as: (a) non-biological degradation,
including hydrolysis, oxidation and reduction, and photodecomposition;
and (b) microbial decomposition, often involving specially adapted
micro-organisms.2 This type of decomposition normally follows a first
order (exponential) type of reaction after an initial lag period while the
micro-organisms become adapted to the substrate. In most cases, nonbiological degradation processes, such as photodecomposition and
volatilization, can occur at the same time as microbially catalysed
reactions.
The range of factors affecting the degradation of organic contaminants by micro-organisms include: soil pH, temperature, supply of
oxygen and nutrients, the structure of the contaminant molecules, their
toxicity and that of their intermediate decomposition products, the water
solubility of the contaminant, and its adsorption to the soil matrix (and
therefore the organic matter content of the soil).2 Adsorption of organic
molecules tends to decrease with increasing temperature because most
adsorption reactions are exothermic. Volatilization losses tend to be
greatest at high temperatures.
The persistence of organic contaminants in soils is determined by the
balance between adsorption onto soil colloids, uptake by plants, and
transformation or degradation processes. Organochlorine molecules are
regarded as being highly persistent with a duration in the soil of 2-10 or
more years. PAHs have been shown to persist for more than 20 years in
soils treated with sewage sludge, but different PAH molecules vary in the
rate at which they are decomposed.10

3 SOURCES OF LAND CONTAMINANTS
Soils receive contaminants from a wide range of sources, including:
(1) Atmospheric fallout from:
—fossil fuel combustion (oxides and acid anions of S and N; heavy
metals; PAHs)
10

S. R. Wild, K. S. Waterhouse, S. P. McGrath, and K. C. Jones, Environ. ScL Technoi, 1990, 24,
17-6-1711.

—Pb, PAHs, etc. from automobile exhausts
—metal smelting operations (As, Cd, Cu, Cr, Hg, Ni, Pb, Sb, Tl,
Zn)
—metal-using industries, including foundries (Cd, Cu, Pb, Zn)
—chemical industries (organic micropollutants, Hg)
—waste disposal by incineration (Cd, TCDDs, TCDFs)
—radionuclides from reactor accidents and the atmospheric
testing of nuclear weapons
—large fires (PAHs; Pb, Cr, etc. from paint; TCDDs; TCDFs)
(2) Agricultural chemicals:
—herbicides (organic molecules—some containing TCDD; B, and
As compounds)
—insecticides (chlorinated hydrocarbons, e.g. DDT, BHC)
—fungicides (Cu, Zn, Hg, and organic molecules)
—acaricides {e.g. Tar Oil')
—fertilizers {e.g. Cd and U impurities in phosphates)
(3) Waste Disposal (intentional/unintentional input to soil):
—farm manures (As, Cu, and Zn in pig and poultry manures)
—sewage sludges (heavy metals; organic pollutants; PAHs; pathogens)
—composts from domestic wastes (metals; organics)
—mine wastes (coal mines—SO4~; metalliferous mines As, Cd,
Cu, Pb, Zn, Ba, U, etc.)
—seepage of leachate from landfills (metals; Cl""; PCBs; etc.)
—ash from fossil fuel combustion, incinerators, bonfires and
accidental fires (PAHs; TCDDs; metals; etc.)
—burial of diseased livestock on farmland
(4) Incidental Accumulation of contaminants:
—corrosion of metal in contact with soil {e.g. Zn from galavanized
metal; Cu and Pb from roofing, scrapyards, etc.)
—wood preservatives from fencing (PCP; PAHs in creosote; As;
Cr; Cu)
—leakage from underground storage tanks (petroleum; aviation
fuel; chlorinated solvents)
—warfare and military training (hydrocarbons; PAHs from fires,
explosives, and their degradation products; metals from munitions and vehicles)
—sports and leisure activities (Pb, Sb, and As from shotgun
pellets; Pb from fishing weights; Pb, Cd, Ni, and Hg from
discarded batteries; hydrocarbons from spilt petrol and lubricating oil)
(5) Derelict industrial sites—wide range of contaminants from production, waste disposal, and building demolition, e.g.:

—Gas works—phenols; tars (PAHs); cyanides; and As
—Electrical industries—Cu, Pb, Zn; PCBs; solvents
—Tanneries—Cr
—Scrapyards—metals; PCBs; hydrocarbons
A more comprehensive list of the most common contaminating uses of
land includes: waste disposal sites, gas works, oil and petroleum
refineries, petrol stations, electricity generating stations, iron and steel
works, non-ferrous metals processing, metal products fabrication and
metal finishing, chemical works, glass making and ceramics, textile
plants, leather tanning works, timber and timber products treatment
works, manufacture of integrated circuits and semi-conductors, food
processing, sewage works, asbestos works, docks and railway land, paper
and printing works, heavy engineering installations, installations processing radioactive materials, and burial of diseased farm livestock.11
4 CHARACTERISTICS OF SOME MAJOR GROUPS OF LAND
CONTAMINANTS
4.1 Heavy Metals
Metals, such as Ag, Cd, Cr, Cu, Hg, Mn, Mo, Ni, Pb, Sb, Tl, U, V, and
Zn tend to be relatively strongly adsorbed by soil constituents. Their
mobility and bioavailability depend on the soil conditions but some
metals, such as Cd and Zn which tend to be less strongly sorbed than Pb
and Cu, can be leached down soil profiles, especially under acid
conditions. Industrial contaminated land often contains a suite of heavy
metal contaminants present in high concentrations. Examples include
chlor-alkali industries where Hg electrodes are used in the electrolysis of
sea water to produce Cl2 gas and NaOH. Both air and soil contamination
will have occurred at these sites from emissions of volatile Hg and
spillages of liquid Hg. Mines, smelters, foundries, paint factories, and
scrap yards are likely to have been heavily contaminated but examples of
less obvious non-point sources of heavy metals are: sewage sludge
disposal to land (range of metals), Pb in shotgun pellets used for game
bird and clay pigeon shooting, and Cd in phosphatic fertilizers.
A very wide range of heavy metal concentrations is found in sewage
sludges but the general trend in technologically advanced countries is
that concentrations have decreased over the last 30 years with the
introduction of cleaner technologies and pollution prevention procedures. In the UK the median (50 percentile) Cd content of sludges
n

House of Commons Environment Committee, '1st Report: Contaminated Land', HMSO,
London, 1990,VoI. 1.

decreased by 64% (from 9 /ig Cd g l to 3.2 /xg Cd g l) over the period
1983 to 1993. The maximum concentrations of metals in sludges reported
in the literature are as follows, but most sludges will have very much
lower values (jig g" l in dry matter): Ag (< 960), As (< 30), Cd (< 3410),
Cr (<40600), Cu (50-8000), Hg (<55), Mn (60-3900), Mo (<40), Ni
(<5300), Pb (<3600), Sb (<34), Se (<10), Sn (<700), V (<400), Zn
(<49000).12 Shotgun pellets are composed predominantly of Pb (9098%) but can also contain Sb (up to 8%), As, and Ni. More than 1000 t
of Pb may be dispersed into soils annually in some countries. Phosphatic
fertilizers are used on agricultural soils in all advanced agricultural
systems but these can contain considerable amounts of Cd owing to
concentrations of up to 640 mg Cd kg" 1 P occurring in the phosphate
rock used for their manufacture.13

4.2

Organic Contaminants

There are more than 20000 organic contaminants known already and
this number will increase as analytical methods are further refined and
more studies are made of materials containing complex mixtures of
organic pollutants, such as industrial wastes, sewage sludges and landfill
leachates.
Petroleum hydrocarbons tend to be very ubiquitous contaminants of
soils, arising from spillages, leaking tanks, and intentional mixing of
petroleum refinery sludges with soil in the process of landfarming, which
exploits the degradation brought about by soil micro-organisms. Many
of the lower molecular weight hydrocarbon compounds are relatively
easily degraded by soil organisms but it takes time and the contaminated
soil could pose a fire risk during the period it is contaminated.
Pesticides can be soil contaminants as a result of persistence after use
on crops, runoff from treated land, accidental spillages, or pesticide
manufacture. Although there are over ten thousand commercial pesticide formulations of around 450 compounds in use, they can be classified
under a relativity narrow range of molecular groups. These compounds
are used because of their toxicity towards specific pests but there is a risk
of harm to soil organisms, other beneficial plants, animals, and humans.
Usually, less than 10% (often around 1%) of the pesticide applied
reaches its target, the rest is dispersed in the environment. Humans can
be affected through the food chain or in contaminated drinking water
12

'Heavy Metals in Soils', ed. B. J. Alloway, Blackie Academic and Professional, Glasgow, 2nd
Edn., 1995.
M. E. Sumner and M. J. McLaughlan, 'Contaminants and the Soil Environment in the
Australasia-Pacific Region', ed. R. Naidu, R. S. Kookana, D. P. Oliver, S. Rogers, and M. J.
McLaughlan, Kluwer Academic Publishers, Dordrecht, 1996, p. 125.

which the pesticide, or its decomposition products, reach either by
leaching to ground water or in runoff into surface waters.
Three main groups of insecticides are currently used in agriculture:
organochlorines, organophosphates, and carbamates. The organochlorines have been widely used for up to 50 years and are the most persistent
of all groups of pesticides. Their persistence decreases in the order:
DDT > dieldrin > lindane (BHC) > heptachlor > aldrin
with half lives of eleven years for DDT down to four years for aldrin.2
Organophosphates are highly toxic to humans and other mammals but
are less persistent in the soil than organochlorines (six month half lives
for parathion, diazinon, and demeton). Carbamates are used to control a
wide range of pests including molluscs, fungi, and insects but have a
similar persistence to organophosphates. Aldicarb (or Temik) is a highly
toxic carbamate that is used as both an insecticide and a nematicide on
potato and sugar beet crops. It is readily oxidized in the soil but its
oxidation products are also highly toxic and readily leached and can
cause ecological and human health problems.14 Some examples of the
molecular structures of pesticides and other organic contaminations are
shown in Figure 2.
There are six major groups of compounds used as herbicides: phenoxyacetic acids, toluidines, triazines, phenylureas, bipyridyls, and glycines.
The most important phenoxyacetic acids are 2,4-D and 2,4,5-T which
have a persistence of up to eight months but are of particular environmental significance because they can be contaminated with dioxins
(TCDDs). 'Agent Orange', the defoliant used by the US forces in
Vietnam, comprised a mixture of 2,4-D and 2,4,5-T and caused widespread contamination of soils by dioxin. The toluidines and triazines are
fairly strongly adsorbed and have a persistence of up to twelve months
although atrazine contamination of groundwaters is a serious problem in
many intensive arable farming areas. The phenylureas tend to be fairly
soluble and are rapidly leached. In contrast, the bipyridyls such as
paraquat and diquat are cationic and strongly adsorbed on soil colloids
(Kd values of <4.2 x 10~~4 on montmorillonite) and are therefore very
persistent. Fungicides comprise a more diverse group of compounds
including inorganics, such as copper and mercury compounds, and a
wide range of organic compounds.
Apart from pesticides, which are synthesized intentionally, a wide
range of chlorinated compounds including dioxins (TCDDs) and dibenzodifurans (TCDFs) have been widely dispersed in the environment as a
14

F. A. M. de Haan, 'Scientific Basis for Soil Protection in Europe', ed. H. Barth and P. L'Hermite,
Elsevier, Amsterdam, 1987, p. 211.

2,4,5-T

ATRAZINE

2,4,5-trichlorophenoxyacetic acid
PARAQUAT(dichloride)

1 >1'-dimethyl-4,4l-bipyridylium
dichloride

2-chloro-4-ethylamino-6-isopropylamino-1,3,5-triazine
DDT

1,1 -di(4-chlorophenyl)-trichloroethane

PARATHION

0,0 -diethyl-0-(4-nitrophenyi)-phosphorothioate
TCDD

2,3,7,8-tetrachlorodibenzodioxin
Figure 2

TCDF

2,3,7,8-tetrachlorodibenzofura

Examples of the molecular structures of pesticides and organic contaminants

result of their accidental synthesis at relatively high temperatures, their
highly stable structure, and their slow rate of degradation. Typical
situations in which they are formed include: synthetic reactions where
the temperature conditions become too hot (such as in the manufacture
of the herbicide 2,4,5-T) and the incineration of rubbish containing PVC
and other sources of chlorides and aromatic compounds. Polychlorinated biphenyls (PCBs) are very stable and were originally manufactured
for use in electrical transformers and capacitors and as plasticizers.
Although no longer in use, they are found in soils around industrial and

domestic waste tips and electronic component factories as well as in
remote sites which they have reached by transport in air. Incineration
temperatures need to be >1200°C with adequate oxygen in order to
ensure the destruction of stable molecules like PCBs. Solvents used for
degreasing electrical components such as trichloroethane are important
atmospheric and groundwater contaminants (DNAPLs) and may also
occur in soils around factories (where they could pose a fire or explosion
risk).15
Polycyclic aromatic hydrocarbons (PAHs) are very important persistent organic pollutants in the environment. Some of them, such as
benzo( 1000 fig g~l are sometimes
found.16
The organic contaminants frequently found in sewage sludges include:
—halogenated aromatics (PCBs—polychlorinated biphenyls, polychlorinated terphenyls, PCNs—polychlorinated naphthalenes, and
polychlorobenzenes)
—-aromatic amines and nitrosamines
—phenols and halogenated aromatics containing oxygen
—polyaromatic and heteroaromatic hydrocarbons (PAHs) and halogenated aliphatics
15

B. J. Alloway and D. C. Ayres, 'Chemical Principles of Environmental Pollution', Blackie
Academic and Professional, London, 2nd Edn., 1997.
16
D. Sauerbeck, 'Scientific Basis for Soil Protection', ed. H. Barth and P. L'Hermite, Elsevier,
Amsterdam, 1987, p. 181.

—aliphatic and aromatic hydrocarbons
—phthalate esters
—pesticides
Of all the organic contaminants listed here, the PAHs and PCBs are
currently considered to constitute the greatest hazard to human health.16
The range of heavy metal concentrations found in sewage sludges is
given above in Section 4.1.
In addition to contaminant chemicals, sewage sludges may contain
some pathogens which were not destroyed during the sewage treatment
process. Many of these will perish when exposed to extremes of
temperature and UV light but care should be taken with this serious
potential risk.
A relatively high proportion of the sludge produced in many countries
is applied to agricultural land as a means of disposal (67% in the UK).
This sludge has many useful properties for agriculture (source of N and P
and physical soil conditioner) but its use is limited by its concentrations
of persistent contaminants. Sewage sludges are frequently used in the
landscaping of derelict land where they act as a growth medium for
plants grown for amenity purposes rather than food. The nutrients they
contain may also have a beneficial effect on soil micro-organism populations which could help to stimulate the biodegradation of persistent
organic pollutants.
5 POSSIBLE HAZARDS FROM CONTAMINATED LAND
Soil contamination can restrict the options available for the reuse of land
because of the potential hazards posed by the contaminants. These can
include:
Hazard
(1) Direct ingestion of
contaminated soil (mainly
children or animals)
(2) Inhalation of dusts and
vapours
(3) Uptake by plants of
contaminants hazardous to
animals and people through the
food chain

Examples of Contaminants
As, Cd, Hg, Pb, CN~, coal tars,
phenols
toluene, benzene, xylene, various
solvents, radon, Hg, asbestos,
metal-rich particles
As, Cd, Pb, Tl, PAHs

(4) Phytotoxicity
(5) Deterioration of building
materials and services
(6) Fires and explosions

SO4", Cu, Ni, Zn, CH4
SO4 ~~, SO3 ~, Cl ~, coal tar,
mineral oils, organic solvents
CH4, S, coal dust, oils, tar,
petroleum, rubber, high calorific
value wastes (e.g. old landfills)
(7) Dermal contact (e.g. during
coal tar, phenols, radionuclides,
demolition, children at play)
PAHs, TCDDs, asbestos, etc.
(8) Contamination of water
CN", SO 4 ", DNAPLs (e.g.
solvents), soluble metals,
pesticides, fuels
(Adapted from References 17 and 18).

6 METHODS OF SITE INVESTIGATION
In investigations of land suspected of being significantly contaminated,
the type of contaminants suspected, their source, the transport mechanism (e.g. wind or direct placement), and the nature of the site are all
important factors determining the choice of sampling procedure. For
example, agricultural land suspected of being contaminated by fallout
from a point source of contamination several kilometres away would be
investigated differently from a former industrial site where the remaining
buildings and other structures indicate the areas most likely to be heavily
contaminated. Site investigations involve the collection (by boring or
digging) of samples. These are subsequently sent to an appropriately
accredited laboratory for determination of a range of contaminants and
selected soil parameters (e.g. pH) and the process can be very expensive.
Therefore, as with most matters, the sampling regime is a compromise
between efficiency and cost. It is convenient to divide the investigation of
contaminated sites into several categories:
(a) Large areas, including agricultural land, contaminated by atmospheric fallout from distant, or non-point sources: investigations
on land of this type will usually need to use a grid pattern of
sampling with most emphasis on topsoil (0-15 cm, especially 0—
5 cm) but samples will need to be collected from greater depths
17

M. J. Beckett and D. L. Sims, 'Contaminated Soil', ed. J. W. Assink and W. J. van der Brink,
Martinus Nijhoff, Dordrecht, 1986, p. 285.
^Interdepartmental Committee on the Redevelopment of Contaminated Land, 'Guidance on the
assessment and redevelopment of contaminated land', Guidance Note 59/83, Department of the
Environment, London, 1987.

(15-30, 30^45 cm) to determine whether contaminants have been
translocated down the profile. The normal practice for agricultural land in the UK is to take samples with a screw auger in a W
pattern with 25 cores per 5 ha block. The cores for each depth
sample are bulked together to give one sample. If the soil appears
to differ in colour or texture in part of the area being considered,
each distinct area should be (bulk) sampled separately.
Preliminary sampling surveys of areas affected by fallout from
discrete sources are often carried out using transects where
samples are collected at regular intervals in straight lines both up
and downwind from the source. The siting of the transects is
often made along compass bearings but these may have to be
modified to take account of the prevailing wind direction,
accessibility of the area, and the constraints of coastlines and
topography.
(b) Discrete areas of industrial or other obvious contaminating
activity and land adjacent to them: in this situation it is vitally
important to know the purpose for which the site was used, the
processes which were employed, the raw materials and the
location of their storage and handling facilities, the products
manufactured, any possible byproducts formed, the wastes from
enterprises on the site, and the waste disposal practices adopted.
This information should be collected first in a preliminary desk
study of old plans and other available records, before any
attempt is made to take samples on the site. This is because the
information on the location of potential areas of major contamination ('hot spots') within the site, such as storage and effluent
tanks, bulk raw materials stores, and waste heaps is required in
order to design the sampling programme. These potential 'hot
spots' should be carefully examined and sampled separately
from any systematic survey of the whole site. Any pre-existing
pits or other subterranean structures that are found should be
sampled, but care should be taken with regard to safety
procedures.
The objective of carrying out a site investigation at a derelict
industrial site is to locate the most severely contaminated areas ('hot
spots'). These areas could be as small as 50 m2 and are therefore
difficult to locate in any systematic sampling procedure. Comparisons
of stratified random grids, regular square grids, and herringbone
patterns have shown that the latter is generally the most efficient for
locating target areas which are circular, pear-shaped plumes, or ellipses
but with the exception of linear/elliptical targets the regular square grid

Table 1 Example of numbers of sampling points for contaminated sites of
different areas using the recommendation of BSlDDlJ5
Area of site

Recommended number

(ha)

of sampling points

0.5
1.0
1.5

15
25
85

Minimum contaminated area to
provide one sample (at P < 0.05)

(m2)

905
1129
1732

After Hobson.21

is adequate for most purposes.19 Sampling on a grid is the most widely
used approach but the number of samples and size of grid appear to be
mainly a matter of professional judgement. The British Standards Draft
Code of Practice DD17520 recommends minimum numbers of sampling
points for sites of different areas; an example is given in Table 1.
Another approach used by the Interdepartmental Committee for the
Redevelopment of Contaminated Land is that the grid spacing should be
kept to a size which could be effectively handled if it happened to be
missed during sampling (but discovered later).22 This generally implies
that grids of 10-25 m will be used for small sites and 25-50 m for larger
sites. AlOm grid on a 1 ha site would give a 64% probability of finding a
contaminated patch of 100 m2 (which is the size ofalOm x 10m grid).
However, studies on a gasworks site have shown that a 25 m grid gave
results which were not significantly different to much smaller grids of
down to 6.25 m. This obviously has major financial implications, as of
course does the possibility of litigation and very high costs of cleaning up
a 'hot spot' area which may have been missed in the site investigation.
Ideally, soil samples should be taken from trial pits (with appropriate
safety precautions) rather than boreholes. This allows observations to be
made about the stratification of materials and other relevant points. The
excavation may release gases, vapours, or liquids and so vigilance is
necessary and gas sampling may need to be carried out as a precaution.23
Boreholes can be drilled more rapidly than pits can be excavated thus
allowing the site to be surveyed in a shorter time, and they are the only
19

R . Bosnian, L. V o o r t m a n , J. Harmsen, a n d C. Coggan, 'Guidance on the Procedure for the
Investigation of U r b a n a n d Industrial Sites with Regard to Soil C o n t a m i n a t i o n — I S O 10381 Part
5' (Working Draft), International Organisation for Standardisation, Geneva, 1996, p . 64.
20
British Standards Institution, 'Draft for Development, D D 1 7 5 : 1988 Code of Practice for the
Identification of Potentially Contaminated L a n d a n d Its Investigation', BSI, L o n d o n , 1988.
21
D . M . H o b s o n , 'Contaminated Land: Problems a n d Solutions', ed. T. Cairney, Blackie Academic
and Professional, L o n d o n , 1993, p . 29.
22
Interdepartmental Committee o n the Redevelopment of Contaminated Land, Guidance N o t e 18/
79, D e p a r t m e n t of the Environment, L o n d o n , 1983.23
E . E. Finnecy a n d K . W . Pearce, 'Understanding O u r Environment', ed. R. E. Hester, Royal
Society of Chemistry, L o n d o n , 1st Edn., 1986, p . 172.

feasible way of sampling below 3 or 4 m at large numbers of sites. In
practice, a combination of pits and boreholes is often used. Careful
investigation of the ground before the sampling is essential; areas where
the soil material differs in colour, texture, moisture content, apparent
organic content, and even smell should be noted and included in the
sampling.
Samples collected by digger, power augering, and boreholes should be
placed in sealed containers to avoid loss of volatiles, although samples
collected for analysis of volatile compounds are usually placed in
purpose made containers with a space above the sample to allow
accumulation of volatiles. This is referred to as head-space analysis and
the volatile constituents are usually determined by gas chromatography
either on its own (GC) or linked to a mass spectrometer (GC-MS). Care
should be taken to avoid contamination of samples during collection,
packaging, transportation, and processing for analysis. Samples of soil
are normally air dried or oven dried at 30 0C but care should be taken
over the possibility of harmful vapours being released from the samples.
The chemical and microbiological analysis of samples collected from
sites suspected of being contaminated is beyond the scope of this chapter.
However, modern techniques, including inductively coupled optical
emission spectrometry (ICP-OES) and atomic absorption spectrometry
(AAS) for inorganic contaminants such as heavy metals, and gas
chromatography (GC) and gas chromatography linked to a mass
spectrometer (GC-MS) for organic contaminants, provide relatively
rapid and efficient analytical procedures which can cope with large
numbers of samples. In addition, a new generation of rapid, on-site
instruments are being developed which avoid the delay that occurs when
samples are collected and sent off to a laboratory for analysis. Some of
these methods can be used for on-line, real time monitoring, such as
sensors installed for monitoring wells or those used for monitoring
volatile emissions from a site. These developments in analytical chemistry enable site investigations to be carried out more rapidly but the
sampling procedure and location of the sampling points and the choice
of analytical determinants is still vitally important.
7 INTERPRETATION OF SITE INVESTIGATION DATA
Once a site has been investigated and shown to be contaminated, a risk
assessment needs to be undertaken in order to decide what course of
action to take. For example, in a case where agricultural land has been
found to be slightly contaminated, it may be decided that regular
monitoring is all that is required and no further remedial action, except
possibly to investigate the source of the contamination. In the case of

sewage sludge application to land, there will be a gradual build-up of
contaminants, such as heavy metals. Nevertheless, so long as the
national maximum permissible limits are not exceeded, normal agricultural use of the land can continue. In the case of a currently used
industrial site that has been shown to have been contaminated, if the
contaminants are not being transferred off the site and are not currently
causing a hazard either to human health or to ecosystems, it may be
acceptable for that site to be left untouched so long as further contamination is prevented and regular monitoring is carried out. One possibility is to install a containment system which prevents leaching and other
Table 2

UK Department of the Environment Trigger Concentrations for
Environmental Contaminants1*'23 (total concentrations except where
indicated)
Threshold

Contaminant

Proposed uses

Contaminants which may pose hazards to health
As
Gardens, allotments
parks, playing fields, open space
Cd
Gardens, allotments
parks, playing fields, open space
Cr (hexavalentb) Gardens, allotments
parks, playing fields, open space
Cr
Gardens, allotments
parks, playing fields, open space
Pb
Gardens, allotments
parks, playing fields, open space
Hg
Gardens, allotments
parks, playing fields, open space
Se
Gardens, allotments
parks, playing fields, open space

(Trigger
concentrations
}jigg~x)
Action

10
40
3
15
25
1000
600
1000
500
2000
1
20
3
6

Contaminants which are phytotoxic but not normally hazardous to health
B (water soluble) Any uses where plants grown
3
Cu (total)
Any uses where plants grown
130
(extractable0)
50
Ni (total)
Any uses where plants grown
70
(extractable)
20
Zn (total)
Any uses where plants grown
300
(extractable)
130
a

Action concentration yet to be specified.
HexavaIent Cr extracted by 0.1 M HCl adjusted to pH 1 at 37.5 0C.
c
Extracted in 0.05 M EDTA
b

Table 3

UK Department of the Environment Trigger Concentrations for
Contaminants associated with former coal carbonization sites23
Threshold

(Trigger concentrations fig g~l)

Contaminant

Proposed use

PAHs

Gardens, allotments
50
Landscaped areas
1000
Gardens, allotments
200
Landscaped areas, open space
500
Buildings, hard cover
5000
Gardens, allotments
5
Landscaped areas
5
Gardens, allotments, landscaped areas
25
Buildings, hard cover
100
Gardens, allotments
250
Landscaped areas
250
Buildings, hard cover
250
All uses
50
Gardens, allotments, landscaped areas 2000
Buildings
2000
Hard cover
2000
All uses
250
All uses
5000
Gardens, etc.
pH < 5

Coal tar

Phenols
Free cyanide
Complex
cyanides
Thiocyanate
Sulfate

Sulfide
Sulfur
Acidity

Action

500
10 000
200
1000
500
500
1000
5000
Nil
Nil
10000
50000
Nil
1000
5000
pH < 3

movement of contaminants out of the contained volume of soil, but the
engineering task of installing this can be expensive. However, where a
derelict site is to be redeveloped, it will be necessary to remediate the site to
an approved quality standard. This may be a universal standard, such as
those used in the Netherlands where, at least until recently, all land should
be cleaned to a standard appropriate for any future use, including the most
demanding, such as domestic gardens where vegetables may be grown and
children may eat some soil. Elsewhere, a 'fitness for purpose' approach is
used where more stringent quality standards are used for domestic
gardens than for industrial sites where food crops will not be grown.
Examples of the critical concentrations of contaminants at sites which
are to be redeveloped are given in Tables 2 and 3. For comparison, some
indicative values from the soil quality standards used in the Netherlands
include: Target Values (jig g~l) for As 29, Cd 0.8, Cr 100, Cu 36, Hg 0.3,
Ni 35, Pb 85, and Zn 140. The intervention concentrations (at which
action must be taken) for these elements are (fig g" 1 ): As 50, Cd 20, Cr
800, Cu 500, Hg 10, Ni 500, Pb 600, and Zn 3000.24
24

Ministry of Housing, Physical Planning and Environment, Directorate General for Environmental Protection (Netherlands), Environmental Standards for Soil and Water, Leidschendam, 1991.

It must be mentioned that in some cases, 'natural attenuation' of
contamination may be occurring at some sites and this implies that either
the movement of a contaminant has been restricted by natural permeability boundaries and sorption systems or that certain organic contaminants are being degraded without any further human intervention. This
situation is obviously important to recognize but may not be of wide
practical significance, except perhaps if it is occurring in part of an
actively used site where a change of land use is not likely and therefore
sufficient time may be available for the processes to work. In all cases of
land contamination, it is very important that, as far as possible, further
active contamination should be prevented. This may mean emptying
leaking tanks, or changing manufacturing or material handling processes. Unfortunately, with some industrial processes, such as metalliferous smelting, it is impossible to completely eradicate atmospheric
pollution although it can be kept to a minimum.
8 RECLAMATION OF CONTAMINATED LAND
The treatment of contaminated land is a very large and rapidly developing subject area with many new technologies being developed. The
methods used can be classed as ex situ and in situ and examples are as
follows.
8.1 Ex situ Methods
8.1.1 'Dig and Dump'. This is the colloquial name for the process of
excavation of the contaminated soil and its disposal at a licensed landfill.
A variation on this is to excavate and incinerate the severely contaminated soil but this would be much more expensive than landfilling.
Incineration to temperatures in excess of 10000C (up to 2500 °C) with
adequate oxygen is the most effective way of destroying highly persistent
pollutants such as PCBs and TCDDs (dioxins). The mineral residue of
the soil would be converted to a fused silica-rich ash unsuitable for
landscape purposes, which would probably need to be disposed of in a
landfill. Another variation is to solidify the contaminated soil, usually
with cement, or another suitable binder. This is convenient for heavy
metal contamination since it renders the metals immobile and the soil
can possibly be used for road construction.
8.1.2 Soil Cleaning. In this method contaminated soil is excavated (as
above) and transported to a cleaning facility which could be a soilwashing plant or a bioreactor. One of the following treatments is then
carried out.

(a) The soil is washed with selected extractants such as acids, chelating
agents or surfactants to remove certain inorganic or organic
contaminants.
(b) Bioreactors usually comprise a vessel in which a suspension of soil
and water can be stirred and conditions optimized for the
degradation of organic contaminants. Augmentation of the
microbial population present in the soil is possible but it is
usually found that the indigenous organisms can carry out the
degradation.
(c) Biodegradation of stripped contaminated soil in beds or windrows, called 'biopiles' is also carried out. This has the same
objectives as other types of bioremediation but is ex situ in that
the soil is moved to a site where it can be more easily turned and
kept at the optimal temperature.
The cleaned soil usually comprises course mineral particles which cannot
be used for supporting plant growth unless mixed with a source of
organic matter {e.g. compost) and a supply of plant nutrients.

8.2

In situ Methods

The development of in situ methods over the last 20 years has reflected a
great deal of exciting innovative science and technology. This has been
encouraged by policies of government organizations in countries such as
the USA and the Netherlands and several successful techniques have
been developed with others still undergoing investigation. There is a
strong financial incentive to develop techniques which do not necessitate
the expense of digging out contaminated soil and transporting it to a safe
disposal site, such as a licensed landfill. However, in some cases this latter
course of action is sometimes still required where the hazard is too great
to leave the site for a long period while less intensive remediation takes
place. Some of the in situ methods are described below.
8.2.1 Physico-chemical Methods
Covering. Covering of the contaminated soil by a layer of clean soil so
that plants can be grown in the uncontaminated 'cover loam'. The depth
of covering is usually at least a metre but, as with all aspects related to
contaminated land, cost is once again a major consideration. A plastic or
geotextile membrane is often used to separate the contaminated material
from the overlying cover soil. It is important not to disturb the underlying contaminated soil (by excavations or planting deep rooting trees);
however, there is a possibility that some contaminants may migrate

upwards through the soil profile during periods of prolonged dry
weather when evapotranspiration is greater than precipitation.
Dilution. In cases of mild to moderate contamination, it may be
possible to dilute the concentration of contaminants by deep ploughing
and mixing the underlying uncontaminated soil with the contaminated
topsoil. This can be done in the case of heavy metals such as Pb which are
sorbed in a relatively immobile and unavailable form, but the method is
unsuitable for mobile or highly toxic chemicals.
Reduction of availability. The plant availability of the contaminants,
such as heavy metals can be minimized by liming or adding adsorptive
minerals to the soil. However, this does not reduce the risk to children
who might intentionally or accidentally eat soil.
By raising the pH of most soils to 7, most cationic heavy metals,
including: Cd, Cu, Cr, Ni, Pb and Zn can be rendered more strongly
adsorbed and less available for uptake by plants. Manipulation of the
redox conditions is possible, for example by flooding or drainage, but is
not usually a practicable option, except for the creation of wetlands to
render contaminants such as heavy metals unavailable due to the
formation of insoluble sulfides. However, the waterlogged conditions
must be permanent because the sulfides would oxidize and become
soluble and plant available, as in the case of Cd in contaminated paddy
soils in Japan which go through a wetting and drying cycle.
Washing. In situ soil washing can be carried out by setting up a
sprinkler irrigation system which can be used to deliver various extracting solutions to soil. These could include dilute acids or chelating agents
for leaching out heavy metals such as Cd, but the site must have suitable
underdrainage to remove the contaminant-carrying leachate away to a
safe disposal route. This often necessitates the construction of a system
of drainage pipes which deliver the leachate to a sump where it is treated
to remove the contaminant metals. This can be done by exchange or
chelating resins or by precipitation.
Soil vapour extraction. Soils contaminated with highly volatile compounds, such as solvents can be ameliorated by a technique known as soil
vapour extraction. This is only suited to soils with a relatively high
permeability comprising large pores and/or fissures and involves inserting a system of perforated pipes in the contaminated layer of soil so that
air can be drawn through by suction pump. The extracted air containing
the volatile compounds is then passed through a column of activated
carbon to sorb and remove the organic contaminants.

Air stripping. This technique is used to remove volatiles contaminating
groundwater and may need to be carried out alongside other techniques
of soil cleaning. Contaminated groundwater extracted from a well is
brought to the surface and has air bubbled through it to enhance the
volatilization of the organic contaminant. This is usually carried out in a
tower containing a system of baffles which facilitate the aeration. The
volatile compounds are removed from the sparging air by activated
carbon filters.
8.2.2 Biological Methods
Bioremediation. The most widely used method so far has been the in
situ bioremediation of soils contaminated with organic chemicals, such
as petroleum hydrocarbons, using the indigenous soil micro-organisms.
By optimizing the soil conditions for the micro-organisms (mainly
bacteria) through adjusting the pH, temperature (plastic tunnels), and
nutrient supply and by regular cultivation to promote aeration, it has
been shown that many organic contaminants can be effectively degraded.
In the Netherlands there is a government backed research programme on
the in situ bioremediation of contaminated soils and waters. The
acronym for this programme is NOBIS and the work carried out so far
has demonstrated that in situ bioremediation is an effective clean-up
method.
Bioventing. Bioventing is a method which combines the principles of in
situ bioremediation with those of soil vapour extraction. This technique
aims to optimize the biodegradation of certain organic contaminants,
including volatile and semivolatile compounds, halogenated organics,
and PAHs, through the addition of oxygen and nutrients to the soil to
stimulate the indigenous micro-organisms but it is also possible to
augment these by the addition of cultures of bacteria or fungi which
have the ability to degrade certain recalcitrant molecules. The air and
nutrients are blown into a system of perforated pipes in the soil and soil
vapour is also extracted by suction from a series of boreholes (air wells).
The area of contaminated land may be covered by an impermeable cap to
prevent air being drawn in from the surface so that all movement of air is
through the contaminated zone.25
Crop growth. The amelioration of soils contaminated with heavy
metals, such as Cd or Zn could possibly be brought about by growing
hyper-accumulating crop genotypes that take up relatively high concentrations of these and other metals. Although only in the development
stage, this technique is expected to provide an environmentally sound
25

Royal Commission on Environmental Pollution, 19th Report, 'Sustainable Uses of Soil', HMSO,
London, 1996.

clean-up procedure. The harvested plant material containing the accumulated contaminant metals would need to be disposed of safely,
possibly by burning and landfilling the metal-rich ash.
8.3

Specific Techniques for Gasworks Sites

As an example of the number of techniques which are either currently
available or which are being developed for cleaning up contaminated
soils, Brown and co-workers26 provide a list of methods applicable to the
remediation of former gasworks sites, where the predominant organic
contaminants are PAHs; these methods include:
(a) Co-burning—Highly contaminated soils and tars are mixed with
combustible materials and burnt in a modified boiler as a fuel.
(b) Thermal desorption—Contaminated soils are heated to volatilize
certain organic contaminants such as tars.
(c) Coal tar removal—Pulverized coal slurry is used to adsorb tars.
The coal slurry is separated from the cleaned soil and used as a
fuel.
(d) Surfactant flushing—Soil is washed with mixtures of surfactants
to remove PAHs. Surfactant-based foams can also be used for
this.
(e) In situ ozonation—Ozone gas is used to oxidize PAHs directly.
(f) Enhanced biodegradation—Various methods are used, including:
growing the biodegrading bacteria in a unipolar magnetic field;
using surfactants to increase the exposure of sorbed PAHs to
bacterial degradation; using fungi to degrade the more recalcitrant
complex PAHs which are very resistant to bacterial activity;
and, finally, adding chemical oxidants to render PAHs more
biodegradable.
9 CASE STUDIES
This section gives some brief examples of actual contaminated land cases
which illustrate the various principles outlined earlier.
9.1

Gasworks Sites

The manufacture of gas (coal gas) from coal commenced in the early
years of the nineteenth century and most industrialized communities
had their own gasworks to provide fuel for lighting and heating. It is
26

R. A. Brown, M. Jackson, and M. Loucy, Special issue: 'International Symposium and Trade Fair
on the Clean-up of Manufactured Gas Plants', Land Con tarn. Reclam., 1995, 3, 4, 2.1-2.2.

estimated that in the UK there may have been up to 5000 sites
contaminated from this land use (both major gasworks and small gas
manufacturing plants to serve a large factory). The Netherlands has
234 gasworks sites needing remediation, and there are 41 in Canada.
For almost 200 years gasworks sites (also known as manufactured gas
plants—MGPs) have been closed down and the land used for other
purposes. However, it is only in the last few decades that the scale of
the contamination and its potential hazard have been recognized. At
several sites where former gasworks were developed for housing many
years ago, it has been necessary to carry out major retrospective cleanup procedures owing to the possible hazards to the residents. One
example of this is the Kralingen works, near Rotterdam in the
Netherlands. In addition to the gasworks sites themselves, other sites
such as old waste dumps have become contaminated with highly toxic
wastes from gas manufacture and there may be a total of many
thousand of sites (up to 100000) in the UK affected in some way by
gas manufacture.27
The gas manufacturing process involved the dry heating of coal in an
air-free environment, which produced crude gas, tar, coke, and cinders.
The crude gas was subsequently purified by passing through tar separators, condensers, wet purification to remove NH3, HCN, phenols, and
creosols, and, finally, dry purification with ferric oxide to remove sulfur
and cyanide compounds. These purification processes resulted in several
highly toxic compounds being accumulated at gasworks sites. The tars
contain PAHs, hydrocarbons, phenols, benzene, xylene, and naphthalene and the ferric oxide was converted to Fe4(Fe(CN)6)3 (hexacyanoferrate—'prussian blue') and sulfides but residues of oxide also contained
Pb, Cu, and As. These materials can be found on old gasworks sites and
at former waste dumps near to gasworks. The tars were utilized for
various by-products. Being dense liquids they tended to accumulate in
voids and pits within the site.
The PAHs are persistent organic pollutants and are very toxic. Some
are carcinogens and they can enter the body by ingestion, dermal
contact, or inhalation. As a result of the ubiquitous occurrence of
gasworks-contaminated land and the importance of PAHs as contaminants, many of the in situ bioremediation techniques (outlined earlier)
have been developed to deal with these chemicals. Cyanides are potentially very toxic but, fortunately, the characteristic blue colour (colloquially referred to as 'blue billy') is a good indicator of their presence;
unfortunately this makes it attractive to children to handle. A green
coloured Fe cyanide compound is also found at these sites—Fe2(CN)6
27

A. O. Thomas and J. N. Lester, ScL Total Environ., 1994, 152, 239-260.

('Berlin green').28 In addition to solid forms, soluble cyanide compounds
are also found in the groundwater at these sites along with PAHs,
phenols, and several other organic pollutants. In addition to these
characteristic contaminants, other common contaminants such as asbestos are also found at old gas works sites.27
A survey of eight former gasworks sites in the UK showed the
following maximum values for key contaminants (fig g" 1 ): sulfate
< 250 000, elemental S < 250 000, sulfide <4000, free CN <64, total
CN < 8000, total phenols < 1000, toluene extract (tars/PAHs) < 250 000,
and total heavy metals: As <250, Cd <64, Pb <4000, and Cr <250.23
9.2

Soil Contaminated by Landfilling and Waste Disposal

Love Canal, near Niagara Falls in New York State, USA is the name
given to a section of unfinished canal which was used in the 1940s by a
chemical company as a disposal site for approximately 20 000 tonnes of
chemical wastes including intermediates from the manufacture of chlorinated pesticides. The covered waste tip was subsequently used for the
construction of a school and houses. Although there had been various
reports of irritating fumes affecting children, the problem came to a head
after heavy rainfall in the winter of 1977-78 which caused a considerable
rise in the water table. Some drums broke though the soil cap over the
waste and the basements of houses were penetrated by harmful chemicals
in the groundwater. The United States Environmental Protection
Agency (EPA) carried out a series of investigations in 1978 and found
82 different contaminant chemicals in the groundwater, of which 27 were
on the EPA's Priority List and 11 of these were known to be carcinogenic. The site was declared a disaster area (the first declared contamination disaster in US history) and 239 families were evacuated initially
although many more were evacuated later after further investigations.
Interestingly, the EPA later found 24 industrial sites more severely
contaminated than this in the USA. By the early 1980s, 170000 waste
disposal sites in the USA had been documented by the EPA, of which
2100 had received industrial wastes.
In 1980, in the Netherlands, a major pollution problem was discovered
in the village of Lekkerkirk which had been built in the early 1970s on
reclaimed land beside the River Lek. The surface of this land had been
raised by up to 3.5 m by layers of household and demolition waste,
covered by 0.7 m of sand. However, this elevation was still lower than
that of the adjacent river and so the groundwater tended to flow upwards
28

J. Shefchek, I. Murarka, and A. Battaglia, Special issue: 'International Symposium and Trade
Fair on the Clean-up of Manufactured Gas Plants', LandContam. Reclam., 1995, 3, 4, 2.18-2.20.

through the waste towards the surface. In 1978, the surface soil was
found to be contaminated by volatile compounds which caused the
deterioration of buried plastic drinking water pipes and contaminated
the water supply as well as causing toxicity in garden plants and noxious
odours. In 1980, investigations revealed that 75% of the houses in the
village were affected and these had to be evacuated and the residents
found alternative housing. The evacuated houses were supported on
hydraulic jacks while 87 000 m3 of fill mainly contaminated with paint
solvents (toluene and lower boiling point solvents), resins, and heavy
metals (Cd, Hg, Pb, Sb and Zn) were removed and subsequently disposed
of by incineration. The maximum concentrations found were {fig g" 1 ):
toluene <1000, lower boiling point solvents <3000, and metals: Cd
<97,Hg < 8.2, Pb < 740, and Zn <1670.
9.3

Heavy Metal Contamination from Metalliferous Mining and
Smelting

Zinc was mined around the village of Shipham in Somerset, England
during the eighteenth and nineteenth centuries and very high concentrations of Pb, Zn, and Cd were left in the soils and waste heaps around the
village.29 In the late 1970s it was realized that houses built on former
mining land between 1951 and 1960 had high concentrations of
potentially toxic elements such as Cd in the garden soil and that there
was a risk to the health of the householders, who regularly consumed
vegetables and fruit grown in these gardens. Some of the highest Cd
concentrations reported in agricultural and horticultural soils (up to
470/igCdg" 1 ) were found, together with 108-6540 fig Pb g" 1 and
250-37200 fig Zn g" 1 . Although the mean Cd concentration in almost
1000 samples of vegetables and fruit was nearly 17 times higher than the
national average concentration of 0.015 fig Cd g" 1 , no obvious adverse
health effects were found in the group of 500 volunteers (about half of the
village population) who underwent detailed health checks. Nevertheless,
these soils will remain indefinitely contaminated and the residents of
houses with contaminated gardens have been advised not to consume
home-grown produce. However, the uptake of Cd at this site was
proportionally low owing to a relatively high concentration of free
CaCO3 in the soil which buffered the pH at values above 7.0 and sorbed
much of the potentially bioavailable metals.
In contrast to Shipham, in the paddy rice growing district of the Jinzu
Valley in Toyama Province of Japan after the Second World War, elderly
women who had borne several children were found to be suffering from a
severe skeletal disorder called Itai-itai (which literally translates as 'ouch
29

H . Morgan, The Shipman Report—Special Issue, ScL Total Environ., 1988, 75, 1, 43.

ouch'!) because of the pain experienced. This disease has been found to
be caused by excess Cd and a relative deficiency of Ca and protein in the
diet. The Cd and other metal contaminants reached the soil from a Zn-Pb
mine and smelter upriver from the paddy fields. Concentrations of Cd in
the soils were much lower than those in Shipham (ca. 5 /ig Cd g~l) but the
availability of the metal was much higher due to the alternating waterlogging and drying out of the paddy soils which initially led to precipitation of Cd as CdS in the gleyed soil followed by a rapid increase in Cd
mobility and bioavailability when the soils returned to an oxygenated state
and the sulfide dissolved releasing Cd, Zn, and SO4-. Unlike in Shipham,
the subsistence farmers of the Jinzu Valley consumed a diet largely grown
on the contaminated paddy land. The main effects were only in women
who had given birth to several babies and this is associated with the
dynamics of Ca in the skeleton of pregnant women. These cases reveal that
many other factors need to be considered, as well as the type and amount
of contamination, in order to assess the impacts on human health.
9.4

Heavy Metal Contamination of Domestic Garden Soils in Urban
Areas

Urban areas tend to have higher concentrations of heavy metal contaminants such as Pb than rural areas due to the greater amounts of
atmospheric deposition. However, garden soils anywhere can be contaminated from bonfire ash when old painted wood has been burnt,
when metal containing rubbish (such as old Pb-sheathed electric cable)
are buried in the soil, when Pb-containing petrol is spilt, or when
composts, manures, and even mineral fertilizers (especially Cd-containing phosphatic fertilizers) are applied to soil. A survey of heavy metals in
soils from 3550 domestic gardens in several towns and cities in England
and Wales reported the following maximum concentrations (jig g" 1 ): 30
Pb 14 125 (geometric mean 230), Cd 17 (geometric mean 1.2), Cu 16800
(geometric mean 53) and Zn 14 568 (geometric mean 260). In Greater
London, the concentrations tended to be highest of all. In 579 gardens in
Greater London, geometric mean concentrations of heavy metals were:
Pb 647, Cd 1.3, Cu 73.0, and Zn 424. Heavy metal concentrations were
higher in the gardens of older houses (> 35 years), houses within 500 m
of commercial garages (Pb in exhausts), and those in close proximity to
demolition sites, waste tips, or metallurgical industries and those within
10 m of a road. In addition to metals, garden soils are often contaminated with atmospherically deposited PAHs, pesticides, solvents, and
other contaminants.
30

I. Thornton, E. Culbard, S. Moorcroft, J. Watt, M. Wheatley, and M. Thompson, Environ.
Technol. Lett., 1983,6, 137.

9.5

Land Contamination by Solvents, PCBs, and Dioxins Following a
Fire at an Industrial Plant

A fire in Cheshire, UK, at a works (Chemstar) that recovered solvents
from chemical wastes, resulted in a high level of soil contamination by
solvents, benzene, and PCBs from the wastes and dioxins formed from
constituents in the wastes during the combustion process. Concentrations (/zg g" 1 ) of up to 208 benzene, 304 total solvents, 1160 PCBs, and
168 toxic equivalents of dioxin and furans (2,3,7,8-TCDD) were found.
These dioxin/furan concentrations are some of the highest reported in
the literature.31 For comparison, in a well documented case of dioxin
contamination at Times Beach, Missouri, USA, where contaminated
waste oil had been sprayed onto a horse arena to reduce dust problems in
the dry summer, concentrations of 33 fig g~l toxic equivalents 2,3,7,8TCDD caused the death of 48 horses, and many other types of animals
also died.32 At the Chemstar site the main animals to be affected were
guinea pigs, which are well known to be highly sensitive to dioxin
toxicity.31
10 CONCLUSIONS
Soils can be contaminated by a vast range of chemicals, and the
behaviour of the contaminant in the soil depends on both its own
chemical properties and the physical and chemical properties of the soil.
Heavy metals tend to persist in soil for a long time, amounting to
centuries in some cases. Organic contaminants have the potential to be
degraded by soil micro-organisms but the rate at which this occurs can
range from days to years or decades. Some of the highly persistent
chemicals such as PCBs, dioxins and PAHs may persist for more than 20
years. However, it is generally found that most of the persistent organic
pollutants are not readily taken up by plants. Ingestion of contaminated
soil is likely to be a more important exposure route than accumulation
into crops.
The predominant factors affecting the sorption of contaminants by a
soil are the pH and mineralogy for heavy metals and the organic matter
content for organic chemicals. Industrially contaminated land can
contain large amounts of non-pedological material, such as lumps of
construction material, and also numerous voids and fissures, which
could facilitate movement towards the water table. It is important to
remember that it is impossible to avoid some form of contamination in
31
32

T. Craig and R. Grzonka, Land Contam. Rectam., 1994, 2, 1, 19-26.
S. E. Manahan, 'Environmental Chemistry', Brooks/Cole (Wadsworth), Monterey, CA, 1984, p.
612.

soils in technologically advanced countries. This is because atmospheric
deposition (sometimes following long-distance transboundary transport)
and contaminants in agricultural materials will result in at least slight
accumulations of materials as a result of human activity.
Questions
1. Explain why soil and sediments act as sinks for organic and
inorganic contaminants, unlike the other environmental media, air
and water, and consider the importance of the following with
regard to the retention of pollutants:
(i) soil pH
(ii) soil organic matter
(iii) clay minerals and oxides of Fe, Al and Mn.
2. Compare the expected behaviour of cadmium and lead contaminants in soils and their bioavailability to food crops.
3. Consider the possible fate of organic pollutants entering the soil
and discuss the importance of sorption, volatilization, leaching,
and degradation.
4. Compare the factors controlling the mobility and bioavailability of
an organic contaminant, such as a PAH (e.g. benzo(a)pyrene), and
a heavy metal such as Cu.
5. What lessons can be learnt from the Love Canal and Lekkerkirk
cases with regard to the disposal of chemical wastes and risk
assessment.
6. Former gasworks sites exist in many towns and cities in industrially
developed countries. Discuss their potential for redevelopment and
the main contamination hazards associated with these sites. Consider how a site investigation should be carried out at a former
gasworks and the main contaminants which should be analysed for.
7. Discuss the importance of soil quality criteria and guidelines values
for safe concentrations of possible contaminants. Compare the
arguments for a multifunctionality approach compared with a
fitness for purpose approach.
8. 'A knowledge of soil science is an important requirement when
planning an investigation and possible clean-up of a site which is
suspected of being significantly contaminated.' Discuss.

CHAPTER 6

Environmental Cycling of Pollutants
ROY M. HARRISON

INTRODUCTION: BIOGEOCHEMICAL CYCLING

The earlier chapters of this book have followed the traditional subdivision of the environment into compartments (e.g. atmosphere, oceans,
etc.). Whilst these sub-divisions accord with human perceptions and
have certain scientific logic, they encourage the idea that each compartment is an entirely separate entity and that no exchanges occur between
them. This, of course, is far from the truth. Important exchanges of mass
and energy occur at the boundaries of the compartments and many
processes of great scientific interest and environmental importance occur
at these interfaces. A physical example is that of transfer of heat between
the ocean surfaces and the atmosphere, which has a major impact upon
climate and a great influence upon the general circulation of the atmosphere. A chemically based example is the oceanic release of dimethyl
sulfide to the atmosphere, which may, through its decomposition
products, act as a climate regulator (see Chapter 4).
Pollutants emitted into one environmental compartment will, unless
carefully controlled, enter others. Figure 1 illustrates the processes
affecting a pollutant discharged into the atmosphere.1 As mixing
processes dilute it, it may undergo chemical and physical transformations before depositing in rain or snow (wet deposition) or as dry gas or
particles (dry deposition). The deposition processes cause pollution of
land, freshwater, or the seas, according to where they occur. Similarly,
pollutants discharged into a river will, unless degraded, enter the seas.
Solid wastes are often disposed into a landfill. Nowadays these are
carefully designed to avoid leaching by rain and dissemination of
pollutants into groundwaters, which might subsequently be used for
potable supply. In the past, however, instances have come to light where
1

W. H. Schroeder and D. A. Lane, Environ. Sci. TechnoL, 1988, 22, 240.

Dry
transformations
Air
concentrations

Wet
transformations

Transport
and diffusion

Initial
mixing

Scavenging

Total emissions

Manmade

Natural

Scavenging
Dry
deposition

Wet
deposition

Figure 1 Schematic diagram of the atmospheric cycle of a pollutant1
(Reprinted from Environmental Science and Technology by permission of the
American Chemical Society)

insufficient attention was paid to the potential for groundwater contamination, and serious pollution has arisen as a result.
Another important consideration regarding pollutant cycling is that of
degradability, be it chemical or biological. Chemical elements (other
than radioisotopic forms) are, of course, non-degradable and hence once
dispersed in the environment will always be there, although they may
move between compartments. Thus, lead, for example, after emission
from industry or motor vehicles, has a rather short lifetime in the
atmosphere, but upon deposition causes pollution of vegetation, soils,
and waters.2 On a very long time-scale, lead in these compartments will
leach out from soils and transfer to the oceans, where it will concentrate
in bottom sediments.
Some chemical elements undergo chemical changes during environmental cycling which completely alter their properties. For example,
nitrate added to soil as fertilizer can be converted to gaseous nitrous
oxide by biological denitrification processes. Nitrous oxide is an unreactive gas with a long atmospheric lifetime which is destroyed only by
breakdown in the stratosphere. As will be seen later, nitrogen in the
environment may be present in a wide range of valence states, each
conferring different properties.
Some chemical compounds are degradable in the environment. For
example, methane (an important greenhouse gas) is oxidized via carbon
2

R. M. Harrison and D. P. H. Laxen, 'Lead Pollution: Causes and Control', Chapman & Hall,
London, 1981.

monoxide to carbon dioxide and water. Thus, although the chemical
elements are conserved, methane itself is destroyed and were it not
continuously replenished would disappear from the atmosphere. The
breakdown of methane is an important source of water vapour in the
stratosphere, illustrating another, perhaps less obvious, connection
between the cycles of different compounds.
Degradable chemicals which cease to be used will disappear from the
environment. PCBs are no longer used industrially to any significant
degree, having been replaced by more environmentally acceptable alternatives. Their concentrations in the environment are decreasing,
although because of their slow degradability (i.e. persistence), it will
take many years before their levels decrease below analytical detection
limits.
The transfer of an element between different environmental compartments, involving both chemical and biological processes, is termed
biogeochemical cycling. The biogeochemical cycles of the elements lead
and nitrogen will be discussed later in this chapter.

1.1

Environmental Reservoirs

To understand pollutant behaviour and biogeochemical cycling on a
global scale, it is important to appreciate the size and mixing times of the
different reservoirs. These are given in Table 1. The mixing times are a
very approximate indication of the time-scale of vertical mixing of the
reservoir.3 Global mixing can take very much longer as this involves
some very slow processes. These mixing times should be treated with
considerable caution as they oversimplify a complex system. Thus, for
example, a pollutant gas emitted at ground level mixes in the boundary
layer (ca. 1 km) on a time-scale typically of hours. Mixing into the free
troposphere (1-10 km) takes days, whilst mixing into the stratosphere
(10-50 km) is on the time-scale of several years. Thus, no one time-scale
describes atmospheric vertical mixing, and the same applies to other
reservoirs. Such concepts are useful, however, when considering the
behaviour of trace components. For example, a highly reactive hydrocarbon emitted at ground level will probably be decomposed in the
boundary layer. Sulfur dioxide, with an atmospheric lifetime of days,
may enter the free troposphere but is unlikely to enter the stratosphere.
Methane, with a lifetime of several years, extends through all of the three
regions.
3

P. Brimblecombe, 'Air Composition and Chemistry', Cambridge University Press, Cambridge, 2nd
Edn., 1996.

Table 1

Size and vertical mixing of various reservoirs
(from Brimblecombe3)
Mass
(kg)

Biosphere 3
Atmosphere
Hydrosphere
Crust
Mantle
Core
a

4.2
5.2
1.4
2.4
4.0
1.9

x 10 1 5
x 10 18
x 10 21
x 10 22
x 10 2 4
x 10 2 4

Mixing time
(years)
60
0.2
1600
> 3 x 10 7
>108

Plants, animals, and organic matter are included but coal and
sedimentary carbon are not. The mixing time of carbon in living matter
is about 50 years.

It should be noted from Table 1 that the atmosphere is a much smaller
reservoir in terms of mass than the others. The implication is that a given
pollutant mass injected into the atmosphere will represent a much larger
proportion of total mass than in other reservoirs. Because of this, and the
rather rapid mixing of the atmosphere, global pollution problems have
become serious in relation to the atmosphere before doing so in other
environmental media. The converse also tends to be true, that once
emissions into the atmosphere cease, or diminish, the beneficial impact is
seen on a relatively short time-scale. This has been seen in relation to lead,
for instance, where lead in Antarctic ice (derived from snow) has shown a
major decrease resulting from diminishing emissions from industry and
use of leaded petrol4 (Figure 2). Improved air quality in relation to CFCs
will take longer to achieve because of the much longer atmospheric
lifetimes (> 100 years) of some of these species (see Chapter 1).
1.2

Lifetimes

A very useful concept in the context of pollutant cycling is that of the
lifetime of a substance in a given reservoir. We can think in terms of
substances having sources, magnitude S, and sinks, magnitude R. At
equilibrium:
R= S
An analogy is with a bath; the inflow from a tap (S) is equal to
the outflow (R) when the bath is full. An increase in S is balanced by
an increase in R. If the total amount of substance A in the reservoir
4

J.-P. Candelone, S. Hong, C. Pellone, and C. F. Boutron,/. Geophys. Res., 1995,100,16605-16616.

Concentration of lead (pg/g)

LEAD

Date of deposition of snow or ice
Figure 2

Changes in lead concentrations in snowjice deposited in central Greenland from
1773 to 1992
(Adapted from Candelone et al.A)

(analogy = mass of water in the bath) is A9 then the lifetime, x is defined
by:
(1)
In practical terms, the lifetime is equal to the time taken for the
concentration to fall to 1/e (where e is the base of natural logarithms)
of its initial concentration, if the source is turned off. If the removal
mechanism is a chemical reaction, its rate may be described as
follows:
(2)
(In this case d[^4]/d/ describes the rate of loss of A if the source is switched
off; obviously with the source on, at equilibrium d[^4]/d/ = 0). The latter
part of equation (2) assumes first order decay kinetics, i.e. the rate of
decay is equal to the concentration of A9 termed [A], multiplied by a rate
constant, k. As discussed later this is often a reasonable approximation.
Taking equation (1) and dividing both numerator and denominator by
the volume of the reservoir, allows it to be rewritten in terms of
concentration. Thus:
(3)

since S' = R
(4)

Thus the lifetime of a constituent with a first order removal process is
equal to the inverse of the first order rate constant for its removal.
Taking an example from atmospheric chemistry, the major removal
mechanism for many trace gases is reaction with the hydroxyl radical,
OH*. Considering two substances with very different rate constants5 for
this reaction, methane and nitrogen dioxide:
(5)
molecs"1

(6)

molecs"1

(7)

Making the crude assumption of a constant concentration of OH*
radical6 (more justifiable for the long-lived methane, for which fluctuations in OH* will average out, than for short-lived nitrogen dioxide),

where
Worked example

What are the atmospheric lifetimes of CH4 and NO2 if the diurnally
averaged concentration of OH* radical is 1 x 106 molec cm~3?

Then from equation (4)

= 5.1 years for CH4
5

B. J. Finlayson-Pitts and J. N. Pitts Jr., 'Atmospheric Chemistry', John Wiley and Sons,
Chichester, 1986.
6
C . N. Hewitt and R. M. Harrison, Atmos. Environ., 1985,19, 545.

By analogy, for nitrogen dioxide, the lifetime,
T = 20 hours
This general approach to atmospheric chemical cycling has proved
useful in many instances. For example, measurements of atmospheric
concentration, [A]9 for a globally mixed component may be used to
estimate source strength, since

and

where V is the volume of atmosphere in which the component is mixed.
Source strengths estimated in this way, for example for the compound
methyl chloroform, CH3CCl3, known to destroy stratospheric ozone,
may be compared with known industrial emissions to deduce whether
natural sources contribute to the atmospheric burden.
1.2.1 Influence of Lifetime on Environmental Behaviour. Some knowledge of environmental lifetimes of chemicals is very valuable in predicting their environmental behaviour. In relation to the atmosphere, there is
an interesting relationship between the spatial variability in the concentrations of an atmospheric trace species and its atmospheric lifetime.3
Compounds such as methane and carbon dioxide with a long lifetime
with respect to removal from the atmosphere by chemical reactions or
dry and wet deposition (see Section 2 of this chapter) show little spatial
variability around the globe, as their atmospheric lifetime (several years)
exceeds the time-scale of mixing of the entire troposphere (of the order of
a year). On the other hand, for a short-lived species such as nitrogen
dioxide, removal by chemical means or dry or wet deposition occurs
much more quickly than atmospheric mixing and hence there is very
large spatial variability, with concentrations sometimes exceeding
100 ppb in urban areas, whilst remote atmosphere concentrations can
be at the level of a few ppt. By analogy, short-lived species also show a
much greater hour-to-hour and day-to-day variation in concentration at
a given measuring point than long-lived species for which local sources
impact only to a modest degree on the existing background concentration.
This illustration using the atmosphere can be taken somewhat
further in relation to other environmental media. Lifetimes of highly

soluble species such as sodium and chloride in the oceans are long
compared to the mixing times and therefore variations in salinity across
the world's oceans are relatively small (see Chapter 4). Where soils are
concerned, mixing times will generally far exceed lifetimes and extreme
local hot spot concentrations can be found where soils have become
polluted.
Lifetime also influences the way in which we study the environmental cycles of pollutants. In the case of reactive atmospheric
pollutants, it is the reaction rate, or rate of dry or wet deposition,
which determines the lifetime. We are therefore concerned mainly with
the rates of these processes in determining the atmospheric cycle. In
the case of longer-lived species, such as persistent organic compounds
like PCBs and dioxins, chemical reaction rates are rather slow and
these compounds can approach equilibrium between different environmental media such as the atmosphere and surface ocean or the
atmosphere and surface soil, with evaporation exceeding deposition
during warmer periods and wet and dry deposition replacing the
contaminant into the soils or oceans in cooler weather conditions.
Both the kinetic approach dealing with reaction rates and the thermodynamically based approach considering partition between environmental media will be introduced in this chapter. In general the kinetic
or reaction rate approach will be most appropriate to the study of
short-lived reactive substances, whilst the equilibrium approach will be
more applicable to long-lived substances.
2
2.1

RATES OF TRANSFER BETWEEN ENVIRONMENTAL
COMPARTMENTS
Air-Land Exchange

The land surface is an efficient sink for many trace gases. These are
absorbed or decomposed on contact with plants or soil surfaces. Plants
can be particularly active because of their large surface area and ability
to absorb water-soluble gases. The deposition process is crudely
described by the deposition velocity, vd,
_,
vd(cm s ]) =

Flux (figm~2 s~l)
—:——
:——

Atmosphenc concentration (fig m~ J )

The term flux is analogous to a flow of material, in this case expressed as
micrograms of substance depositing per square metre of ground surface
per unit time. In the case of rough surfaces the square metre of area refers
to the area of a hypothetical horizontal flat surface beneath the true

Table 2

Some typical values of deposition velocity

Po IIu tan t

Surface

Deposition velocity
(cm s ~ l )

SO 2
SO 2
SO 2
SO 2
O3
O3
O3
HNO 3
CO
Aerosol ( < 2.5 fim)

Grass
Ocean
Soil
Forest
Dry grass
Wet grass
Snow
Grass
Soil
Grass

1.0
0.5
0.7
2.0
0.5
0.2
0.1
2.0
0.05
0.1

surface rather than the sum of the area of all the rough elements such as
plant leaves which make up the true surface.
Since the deposition process itself causes a gradient in atmospheric
concentration, vd is defined in relation to a reference height, usually
1 m, at which the atmospheric concentration is measured. For reasons
described later, vd is not a constant for a given substance, but varies
according to atmospheric and surface conditions. However, some
typical values are given in Table 2, which exemplify the massive
variability.
For some trace gases, for example, nitric acid vapour, dry deposition
represents a major sink mechanism. In this case the process may have a
major impact upon atmospheric lifetime.
Worked example

Dry deposition is frequently the main sink for ozone in the rural
atmospheric boundary layer. What is the lifetime of ozone with respect
to this process?
Assuming a typical dry deposition velocity of 1 cm s~l and a boundary layer height of 1000 m, (H),
Flux
Mixing depth (m)

where

By analogy with equation (4),

_h_
~ vd
= 1000/0.01 s
= 28 hours

Thus, taking the boundary layer as a discrete compartment, the lifetime
of ozone with respect to dry deposition is around 1 day. The lifetime in
the free troposphere (the section of the atmosphere above the boundary
layer) is longer, being controlled by transfer processes in and out, and
chemical reactions. The stratospheric lifetime of ozone is controlled by
photochemical and chemical reaction processes.
Dry deposition processes are best understood by considering a
resistance analogue. In direct analogy with electrical resistance theory,
the major resistances to deposition are represented by three resistors in
series. Considering the resistances in sequence, starting well above the
ground, these are as follows:
(i) ra, the aerodynamic resistance describes the resistance to transfer
downwards towards the surface through normally turbulent air;
(ii) rb, the boundary layer resistance describes the transfer through a
laminar boundary layer (approximately 1 mm thickness) at the
surface;
(iii) rs, the surface (or canopy) resistance is the resistance to uptake by
the surface itself. This can vary enormously, from essentially zero
for very sticky gases such as HNO 3 vapour, which attaches
irreversibly to surfaces, to very high values for gases of low
water solubility which are not utilized by plants (e.g. CFCs).
Since these resistances operate essentially in series, the total resistance, R,
which is the inverse of the deposition velocity, is equal to the sum of the
individual resistances.
(8)
Some trace gases have a net source at the ground surface and diffuse
upwards; an example is nitrous oxide.
Whether the flux is downward or upward, it is driven by a concentration gradient in the vertical, dc/dz. The relationship between flux, F, and

concentration gradient is:

where K2 is the eddy diffusivity in the vertical (a measure of the
atmospheric conductance). Fluxes, and thus deposition velocities, can
be estimated by measurement of a concentration gradient simultaneously
with the eddy diffusivity.7 It is usually assumed that trace gases transfer
in the same manner as sensible heat {i.e. convective heat transfer, not
radiative or latent heat) or momentum. Thus the eddy diffusivity for
either of these parameters is measured usually from simple meterological
variables (gradients in temperature and wind speed).
A few substances are capable of showing both upward and downward
fluxes. An example is ammonia. Ammonium in the soil, NH4+, is in
equilibrium with ammonia gas, NH3
(9)
when atmospheric concentrations of ammonia exceed equilibrium concentrations at the soil surface (known as the compensation point), the net
flux of ammonia is downwards. When atmospheric concentrations are
below the equilibrium value, ammonia is released into the air.8
2.2

Air-Sea Exchange

The oceans cover some two-thirds of the earth's surface and consequently provide a massive area for exchange of energy (climatologically
important) and matter (an important component of geochemical cycles).
The seas are a source of aerosol {i.e. small particles), which transfer to
the atmosphere. These will subsequently deposit, possibly after chemical
modification, either back in the sea (the major part) or on land (the
minor part). Marine aerosol comprises largely unfractionated seawater,
but may also contain some abnormally enriched components. One
example of abnormal enrichment occurs on the eastern coast of the
Irish Sea. Liquid effluents from the Sellafield nuclear fuel reprocessing
plant in west Cumbria are discharged into the Irish Sea by pipeline. At
one time, permitted discharges were appreciable and as a result radioisotopes such as 137Cs and several isotopes of plutonium have accumulated in the waters and sediments of the Irish Sea. A small fraction of
these radioisotopes were carried back inland in marine aerosol and
7
8

J. A. Garland, Proc. R. Soc. London, Ser. A, 1977, 354, 245.
S . Yamulki, R. M. Harrison, and K. W. T. Goulding, Atmos. Environ., 1996, 30, 109-118.

IRISH
SEA

Figure 3

Concentrations ofplutonium in soils of West Cumbria (239+24° pu to 15 cm
depth; pCi cm"2). The point marked S indicates the position of the Sellafield
reprocessing works
(From Cawse10)

deposited predominantly in the coastal zone.9 Whilst the abundance of
137
Cs in marine aerosol was reflective only of its abundance in seawater
(an enrichment factor—see Chapter 4—of close to unity), plutonium was
abnormally enriched due to selective incorporation of small suspended
sediment particles in the aerosol. This has manifested itself in enrichment
ofplutonium in soils on the west Cumbria coast,10 shown as contours of
239+ 24opu deposition (pCi cm" 2 ) to soil in Figure 3.
9

R. S. Cambray and J. D. Eakins, Nature, 1982, 300, 46.
P. A. Cawse, UKAEA Report No. AERE-9851, 1980.

Dry deposition velocity Vd. (cms" 1 )

Mean windspeed 10 m s"1
Mean windspeed / 5 m s"1

Diameter of unit density particle dp(jjm)
Figure 4

Calculated values of deposition velocity to water surfaces as a function of particle
size and wind speed

The seas may also act as a receptor for depositing aerosol. Deposition velocities of particles to the sea are a function of particle size,
density, and shape, as well as the state of the sea. Experimental
determination of aerosol deposition velocities to the sea is almost
impossible and we have to rely upon data derived from wind tunnel
studies and theoretical models. The results from two such models
appear in Figure 4, in which particle size is expressed as aerodynamic
diameter, or the diameter of an aerodynamically equivalent sphere of
unit specific gravity.11'12 If the airborne concentration in size fraction of
diameter dt is ch then
Total flux
where vg(dt) is the mean value of deposition velocity appropriate to the
size fraction dt. Measurements show that whilst most of the lead, for
example, is associated with small, sub-micrometre particles, the larger
particles compose the major part of the flux.
Airborne concentrations of particulate pollutants are not uniform
over the sea. The spatial distribution of zinc over the North Sea13
averaged over a number of measurement cruises appears in Figure 5.
ll

S . A . Slinn and W. G. N. Slinn, Atmos. Environ., 1980,14, 1013.
R. M. Williams, Atmos. Environ., 1982, 16, 1933.
13
C. R. Ottley and R. M. Harrison, Eurotrac ASE Annual Report, Garmisch-Partenkirchen, 1990.
12

Z n

Figure 5

Spatial distribution of zinc concentrations (in ng m
during 1989

) in air over the North Sea

(From Ottley and Harrison )

Spatial patterns of other metals and many artificial pollutants are
similar, reflecting the impact of land-based source regions, with concentrations falling toward the north and centre of the sea.
Because of its position and relatively high pollution loading, the North
Sea is a focus of considerable interest. An inventory of inputs of trace
metals {e.g. Pb, Cd, Zn, Cu, etc.) accords similar importance to riverine
inputs and atmospheric deposition.14 Controls have now been applied to
many source categories and total inputs of the metals indicated have in
general declined appreciably. One particular example is lead, for which
most European countries introduced severe controls on use in gasoline
(petrol) during the 1980s and atmospheric concentrations have fallen
accordingly. Although the data are less clear, it might be anticipated that
concentrations in river water will also decline as a result of reduced
inputs from direct atmospheric deposition and in runoff waters from
highways and land surfaces.
As explained in Chapter 4, the sea may be both a source and a sink of
trace gases. The direction of flux is dependent upon the relative
concentration in air and seawater.15 If the concentration in air is Ca, the
14

R. F. Critchley, Proc. Int. Conf. Heavy Metals in the Environment, Heidelberg, Germany; CEP
Consultants, Edinburgh, 1983, p. 1109.
15
P. S. Liss and L. Merlivat, 'The Role of Air-Sea Exchange in Geochemical Cycling', ed. P. BuatMenard, Reidel, Dordrecht, 1986, p. 113.

equilibrium concentration in seawater, CW(equ) is given by
(10)
where His the Henry's Law constant. The Henry's Law constant can be
expressed as follows:

where ps = saturation vapour pressure and S^q = equilibrium solubility
in water.
Worked example

For benzene

Calculate H for benzene at 250C.
For benzene

If Cw is the actual concentration of the dissolved gas in the surface
seawater and

the system is at equilibrium and no net transfer occurs. If, however, there
is a concentration difference, AC, where
(H)

there will be a net flux. If

the water is sub-saturated with regard to the trace gas and transfer occurs
from air to water. Conversely, gas transfers from supersaturated water to

the atmosphere if

The rate at which gas transfer occurs is expressed by
(12)
where AT(T)W is termed the total transfer velocity. This can be broken
down into component parts as follows:
(13)
where ka and &w are the individual transfer velocities for chemically
unreactive gases in air and water phases, respectively and a ( = Reactive/
^inert) is a factor which quantifies any enhancement of gas transfer in the
water due to chemical reaction. The terms rw and ra are the resistances to
transfer in the water and air phases respectively and are directly
analogous to the resistance terms in equation (8). For chemically reactive
gases, usually ra » rw and atmospheric transfer limits the overall flux.
For less reactive gases the inverse is true and ^( T ) w = A:w; the resistance in
the water is the dominant term.
Much research has gone into evaluating kw and Kmw, both in
theoretical models, and in wind tunnel and field studies. The results are
highly wind speed dependent due to the influence of wind upon the
surface state of the sea. The results of some theoretical predictions and
experimental studies16 for CO 2 (a gas for which kw is dominant) are
shown in Figure 6.
In addition to dry deposition, trace gases and particles are also
removed from the atmosphere by rainfall and other forms of precipitation (snow, hail, etc.), entering land and seas as a consequence. Wet
deposition may be simply described in two ways. Firstly,
Concentration in rain (mg kg"1)
Scavenging ratio = —
r—
:
:
:
Concentration
in air
(mg kg"1)
Typical values of scavenging ratio 17 lie within the range 300-2000.
Scavenging ratios are rather variable, dependent upon the chemical
nature of the trace substance (particle or gas, soluble or insoluble, etc.)
16
17

A. J. Watson, R. C. Upstill-Goddard, and P. S. Liss, Nature, 1991, 349, 145.
R. M. Harrison and A. G. Allen, Atmos. Environ., 1991, 25A, 1719.

(10~ 5 m s H )

Transfer velocity (cm h"1)

Beaufort

Wind speed (m s"1)
Figure 6 Air-sea transfer velocities for carbon dioxide at 200C as a function of wind speed
at 10 metres (m s~* or Beaufort Scale). The graph combines experimental data
(points) and a theoretical line
(From Watson et aL16) (Reprinted by permission from Nature (London), 349,
145; Copyright © 1991 Macmillan Magazines Ltd.)

and the type of atmospheric precipitation. Incorporation of gases and
particles into rain can occur both by in-cloud scavenging (also termed
rainout) and below-cloud scavenging (termed washout).
Numerical modellers often find it convenient to describe wet deposition by a scavenging coefficient, actually a first order rate constant for
removal from the atmosphere. Thus, for trace substance A,

where A is the washout coefficient, with units of s" 1 . A typical value of
A for a soluble substance is 10" 4 S" 1 although actual values are difficult
to measure and are highly dependent upon factors such as rainfall
intensity.

Table 3

A comparison of the concentration of major elements
in 'average' riverine paniculate material and surficial
rocks
Concentrations (g kg" l )

Element

Riverine paniculate
material

Surficial rocks

Al
Ca
Fe
K
Mg
Mn
Na
P
Si
Ti

94.0
21.5
48.0
20.0
11.8
1.1
7.1
1.2
285.0
5.6

69.3
45.0
35.9
24.4
16.4
0.7
14.2
0.6
275.0
3.8

Adapted from Martin and Meybeck18

3 TRANSFERS IN AQUATIC SYSTEMS
When rain falls over land some drains off the surface directly into surface
water courses in surface runoff. A further part of the incoming rainwater
percolates into the soil and passes more slowly into either surface waters
or underground reservoirs. Water held in rock below the surface is
termed groundwater, and a rock formation which stores and transmits
water in useful quantities is termed an aquifer. Water which passes
through soil or rock on its way to a river is chemically modified during
transit, generally by addition of soluble and colloidal substances washed
out of the ground. Some substances are removed from the water; for
example river water often contains less lead than rainwater; one
mechanism of removal is uptake by soil.
River waters carry both dissolved and suspended substances to the sea.
The concentrations and absolute fluxes vary tremendously. The suspended solids load is largely a function of the flow in the river, which
influences the degree of turbulence and thus the extent to which solids are
held in suspension and resuspended from the bed, once deposited. Table
3 shows a comparison of'average' riverine suspended particulate matter
and surficial rock composition18 for the major elements. Elements
resistant to chemical weathering or biological activity (e.g. aluminium,
titanium, iron, phosphorus) show some enrichment in the riverine solids,
18

J. M. Martin and M. Meybeck, Mar. Chem., 1979, 7, 177-206.

Table 4

A verage concentrations of the major
constituents dissolved in rain and river water
Concentrations (mg dm~ 3 )

Constituent

Rain water

River water

Na +
K+
Me 2 +
C^+
Fe
Al
CP
SO2T
HCO 3 SiO2
pH

1.98
0.30
0.27
0.09

6.3
2.3
4.1
15
0.67
0.01
7.8
11.2
58.4
13.1

3.79
0.58
0.12
5.7

Adapted from Garrels and Mackenzie19

whilst more soluble elements are subject to weathering and are depleted
in the solids, being transported largely in solution (sodium, calcium).
Some pollutant elements such as the metals lead, cadmium, and zinc tend
to be highly enriched in the solids relative to surficial rocks or soils due to
artificial inputs.
The dissolved components of river water typically exhibit significantly
higher concentrations than in rainwater19 (Table 4), due to leaching from
rocks and soils. Some insight into the processes governing river water
composition may be gained from Figure 7. Starting from the point of
lowest dissolved salts concentrations, the ratio of Na/(Na + Ca)
approaches one. This is similar to rainwater, and is termed the precipitation dominance regime. It is typified by rivers in humid tropical areas of
the world with very high rainwater inputs and little evaporation. As the
dissolved solids concentration increases the ratio Na/(Na H- Ca)
declines, indicating an increasing importance for calcium in the rock
dominance regime. Here, increased weathering of rock provides the
major source of dissolved solids. As dissolved solids increase further,
the abundance of calcium decreases relative to sodium as the water
becomes saturated with respect to CaCC>3, and this compound precipitates. Waters in the evaporation/precipitation regime are typified by
rivers in very arid parts of the world (e.g. River Jordan) and the major
seas and oceans of the world.20'21
19

R. M. Garrels and F. T. MacKenzie, 'Evolution of Sedimentary Rocks', ed. W. W. Norton, New
York, 1971.
R. J. Gibbs, Science, 1970, 170, 1088.
21
R. M. Harrison and S. J. de Mora, 'Introductory Chemistry for the Environmental Sciences',
Cambridge University Press, Cambridge, Second Edn., 1996.

major oceans

(a)

Black
Caspian
Jordan

Baltic

Colorado

Total dissolved solids (mg dm* 3 )

Volga
Yukon
Columbia*

Mississippi
Ganges
Congo

On
rioco
Negro

(b)

Tefe

sea
water
evaporation
crystallisation

series

rock dominance

series

precipitation
dominance

Weight Na*
Weight (Na*•Co2*)
Figure 7

The chemistry of the Earth's surface waters: (a) typical values of the ratio
Na + I(Na+ + Ca2+) as a function of dissolved solids concentration for various
major rivers and oceans; (b) the processes leading to the observed ratios

(From Gibbs20) copyright © 1970, American Association for the
Advancement of Science

The flux of material in a river to the sea is expressed by:
FluxCgs""1) = Volumetric discharge (Hi3S"1) x Concentration (gm~ 3 )
In total, the rivers of the world carry around 4.2 x 1012 kg per year of
dissolved solids to the oceans and 18.3 x 1012 kg per year of suspended
solids.

Depth ( c m )

Year deposited

background

Lead concentration

(jug g" 1 )

Figure 8 Lead profile in a lake sediment in relation to depth and the year of incorporation
(From Davies and Galloway 22 )

In slow-moving water bodies such as lakes and ocean basins, suspended solids falling to the bottom produce a well stratified layer of
bottom sediment. This is stratified in terms of age with the oldest
sediment at the bottom (where when suitably pressurized it can form
rock) and the newest at the top, in contact with the water. If burrowing
organisms do not provide too much disturbance (termed bioturbation),
the sediment can preserve a record of depositional inputs to the water
body. An example is provided by Figure 8 in which lead is analysed in a
sediment core dated from its radioisotope content.22 The concentration
rises from a background in around the year 1800, corresponding to the
onset of industrialization. Considerably increased deposition is seen after
1930 due to the introduction of leaded petrol. Whilst some of the lead
input is via surface waters, the majority probably arises from atmospheric deposition.
4

BIOGEOCHEMICAL CYCLES

A general model of a biogeochemical cycle appears in Figure 9. Although
biota are not explicitly included, their role is a very important one in
mediating transfers between the idealized compartments of the model.
For example, the role of marine phytoplankton in transferring sulfur
from the ocean to the atmosphere in the form of dimethyl sulfide has
been highlighted in Chapter 4. Biota play a major role in determining
atmospheric composition. Photosynthesis removes carbon dioxide from
22

A. O. Davies and J. N. Galloway, 'Atmospheric Pollutants in Natural Waters', ed. S. J. Eisenreich,
Ann Arbor, MI, 1981, p. 401.

atmosphere

land

freshwater

oceans

marine sediment/
rock
Figure 9

Schematic diagram of the major fluxes and compartments in a biogeochemical
cycle: (1) runoff; (2) streamflow; (3) degassing; (4) particle suspension; (5)
wet and dry deposition; (6) sedimentation; (7) remobilization

the atmosphere and replenishes oxygen. In a world without biota,
lightning would progressively convert atmospheric oxygen into nitrogen
oxides and thence to nitrate which would reside in the oceans. Biota also
exert more subtle influences. In aquatic sediments, micro-organisms
often deplete oxygen more quickly than it can be replenished from the
overlying water, producing anoxic conditions. This leads to chemical
reduction of elements such as iron and manganese, which has implications for their mobility and bioavailability.
Biological reduction processes in sediments may be viewed as the
oxidation of carbohydrate (in its simplest form CH2O) with accompanying reduction of an oxygen carrier. In the first instance, dissolved
molecular oxygen is used. The reaction is thermodynamically favoured,
as reflected by the strongly negative AG.

When all of the dissolved oxygen is consumed, anaerobic organisms take
over. Initially, nitrate-reducing bacteria are favoured

Once the nitrate is utilized, sulfate reduction takes over

Finally, methane-producing organisms dominate in a sediment depleted
in oxygen, nitrate, and sulfate

Thus highly anoxic waters are commonly sources of hydrogen sulfide,
H2S, from sulfate reduction and of methane (marsh gas). The formation
of sulfide in sediments has led to precipitation of metal sulfides over
geological time, causing accumulations of sulfide minerals of many
elements, e.g. PbS, ZnS, HgS, etc.
4.1 Case Study 1: The Biogeochemical Cycle of Nitrogen
Nitrogen has many valence states available and can exist in the environment in a number of forms, depending upon the oxidizing ability of the
environment. Figure 10 indicates the most important oxidation states
and the relative stability (in terms of free energy of formation).23 The
oxides of nitrogen represent the most oxidized and least thermodynamically stable forms. These exist only in the atmosphere. Ammonia can exist
in gaseous form in the atmosphere but rather rapidly returns to the soil
and waters as ammonium, NH41". Fixation of atmospheric N 2 by
leguminous plants leads to ammonia, NH3. In aerobic soils and aquatic
systems, NH3 and NH^ are progressively oxidized by micro-organisms
via nitrite to nitrate. The latter is taken up by some biota and used as a
nitrogen source in synthesizing amino acids and proteins, the most
thermodynamically stable form of nitrogen. After the death of the
organism, microbiological processes will convert organic nitrogen to
ammonium (ammonification) which is then available for oxidation or
use by plants. Conversion of ammonia to nitrate is termed nitrification,
whilst denitrification involves conversion of nitrate to N2.
Figure 11 shows an idealized nitrogen cycle. The numbers in boxes
represent quantities of nitrogen in the various reservoirs, whilst the
arrows show fluxes.23 It is interesting to note that substances involving
relatively small fluxes and burdens can have a major impact upon people.
Thus nitrogen oxides, NO, NO2, and N2O are very minor constituents
relative to N 2 but play major roles in photochemical air pollution (NO2),
acid rain (HNO3 from NO2), and stratospheric ozone depletion (N2O).
23

P. O'Neill, 'Environmental Chemistry', George, Allen, and Unwin, London, 1985.

Oxidation
state
Oxides of nitrogen
dinitrogen oxide
N2O
nitric oxide
NO
nitrogen dioxide
NO2

Free energy
of
formation

less stable
positive

N2
dinitrogen

N
ami
negative
NO2
nitrite
more stable
NOJ
nitrate

amino acids,
proteins

Reactions
1 fixation
2 nitrification
3 assimilation by plants

4 ammonification
5 electrification
6 nitrate-containing precipitation, often
as nitric acid in acid rain

micro-organisms play a part in reactions 1, 2 U and 5

Figure 10

Chemical forms and cycle of nitrogen
(From O'Neill23)

Nitrate from fertilizers represents a very small flux but has major
implications in terms of eutrophication of surface waters.
4.2

Case Study 2: Aspects of the Biogeochemical Cycle of Lead

Lead is a simpler case to study than nitrogen due to the small number of
available valence states. The major use of lead until recently was as
tetraalkyl lead gasoline additives in which lead is present as Pb iv . The
predominant compounds used are tetramethyl lead, Pb(CH3)4, and
tetraethyl lead, Pb(C2H5)4. These are lost to the atmosphere as vapour
from fuel evaporation and exhaust emissions from cold vehicles, but
comprise only about 1-4% of lead in polluted air.2 Leaded gasoline also
contains the scavengers 1,2-dibromoethane, CH2BrCH2Br and 1,2-

NH3/NHt

N2O/NO/NO2

fertiliser
SEDIMENTSorgank
compounds

biomass

biomass
dissolved N
compounds

inorganic compounds
SEDIMENT

reservoirs

Figure 11

fluxes

Schematic representation of the biogeo chemical cycle of nitrogen, indicating
the approximate magnitude of fluxes and reservoirs
(After O'Neill23)

dichloroethane, CH2ClCH2Cl which convert lead within the engine to
lead halides, predominantly lead bromochloride, PbBrCl, in which lead
is in the Pb11 valence state, its usual form in environmental media. About
75% of lead alkyl burned in the engine is emitted as fine particles of
inorganic lead halides. Atmospheric emissions of lead arise also from
industry; both these and vehicle-emitted lead are declining. Figure 12
shows trends in United Kingdom emissions of lead to atmosphere from
leaded petrol.24 Lead emitted to the atmosphere has a lifetime of around
7-30 days and hence may be subject to long-range transport. Concentrations of trace elements in polar ice provide a historical record of
atmospheric deposition. Measurements (Figure 2) have shown a
marked enhancement in lead accompanying the increase in leaded gasoline usage, and a major decline in recent years attributable to reduced
emissions to atmosphere.4
Atmospheric lead is deposited in wet and dry deposition. Lead is
relatively immobile in soil, and agricultural surface soils in the UK
exhibit concentrations approximately double those of background soil
which contain ca. 15-20 mg kg" 1 derived from soil parent materials,
other than in areas of lead mineralization where far greater concentrations can be found. Local perturbations to the cycle of lead can be
important. For instance, the lead content of garden soils correlates
strongly with the age of the house. This is probably due to the
24

Department of Environment, Transport and the Regions, 'Digest of Environmental Statistics',
No. 19, The Stationery Office Ltd., Edinburgh, 1997.

Figure 12

Gasoline Lead (Th.Tonnes)

Blood Lead (jjg/dl)

Blood Lead
Gasoline Lead

Year
Trends in lead use in petrol (gasoline) and of lead in the blood of the general
population in the United Kingdom, 1970-1995

deterioration of leaded paintwork on older houses and the former
practices of disposing of household refuse and fire ashes in the garden.
Lead is also of low mobility in aquatic sediments and hence the sediment
may provide a record of historical lead deposition (see Figure 8).
Plants can take up lead from soil, thus providing a route of human
exposure. Careful research in recent years has established transfer
factors, termed the Concentration Factor, CF, where
_ AConcentration of lead in plant (mg kg"1 dry wt.)
A Concentration of lead in soil (mg kg"1 dry wt.)
The value of CF for lead is lower than for most metals and is typically
within the range 10~ 3 to 10~ 2 . Much higher values had been estimated
from earlier studies which ignored the importance of direct atmospheric
deposition as a pathway for contamination. The direct input from the air
to leaves of plants is often as great, or greater than soil uptake. 24 ' 25 This
pathway may be described by another transfer factor, termed the Air
Accumulation Factor, AAF, where
3

_K _ AConcentration of lead in plant (fig g"1 dry wt.)
AConcentration of lead in air (^g m~3)

Values of AAF are plant dependent, due to differences in surface
characteristics, but values of 5-40 are typical. 25 ' 26 Thus a plant grown
25
26

R. M . Harrison and M. B. Chirgawi, ScL Total Environ., 1989, 83, 13.
R. M . Harrison and M. B. Chirgawi, Sci. Total Environ., 1989, 83, 47.

on an agricultural soil with 50 mg kg l lead will derive 0.25 mg kg l
dry weight lead from the soil (CF = 5 x 10 ~ 3 ), whilst airborne lead of
0.1/igm" 3 will contribute 2.0 figg~l (^mgkg" 1 ) of lead (AAF =
20 m3 g" 1 ). Thus in this instance airborne lead deposition is dominant.
The air lead concentration of 0.1 fig m~ 3 was typical of rural areas of the
UK until 1985. Since that time, the drastic reduction of lead in gasoline
has led to appreciably reduced lead-in-air concentrations in both urban
and rural localities.
Human exposure to lead arises from four main sources:2'27
(i) inhalation of airborne particles. The adult human respires
approximately 20 m3 of air per day. Thus for an urban lead
concentration of 0.1 jugm~3, intake is 2 fig per day. This is
rather efficiently absorbed (ca. 70%) and therefore uptake is
around 1.4 fig per day in this instance.
(ii) ingestion of lead in foodstuffs. The concentrations of lead in
food obviously vary between different foodstuffs and even
between different batches of the same food. Typical freshweight
concentrations (much of the weight of some foods is water) are
from 10 to 50 ^g Pb kg" 1 . Thus a food consumption of 1.5 kg
per day represents an intake of around 50 fig per day and an
uptake (10-15% efficient) of around 6 fig per day.
(iii) drinking water and beverages. Concentrations of lead in drinking water vary greatly, related particularly to the presence or
absence of lead in the household plumbing system. Most
households in the UK conform to the EC standard of
50/ig P 1 and a concentration of 5 fig\~l may be taken as
representative. Gastrointestinal absorption of lead from water
and other beverages is highly dependent upon food intake.
After long fasting, absorptions of 60-70% have been recorded,
14-19% with a short period of fasting before and after the
meal, and only 3-6% for drinks taken with a meal. If 15% is
taken as typical, for a daily consumption of 1.5 litres, intake is
7.5 fig and uptake 1.1 fig.

(iv) cigarette smoking exposes the individual to additional lead.
Whilst both individual exposure to lead and the uptake efficiencies of
individuals are very variable, it is evident that exposure arises from a
number of sources and control of human lead intake, if deemed to be
desirable, requires attention to all of those sources. An additional
pathway of exposure, not easily quantified, and not included above is
27

Royal Commission on Environmental Pollution, 'Ninth Report: Lead in the Environment',
HMSO, London, 1983.

ingestion of lead-rich surface dust by hand to mouth activity in young
children.
The above calculations estimate that for a typical adult in a developed
country, daily uptake of lead from air, diet, and drinking water is
respectively 1.4 ^g, 6 //g, and 1.1 ^g. Exposure to lead from all of these
sources has fallen rapidly over the past 20-30 years. Figure 12 contrasts
the temporal trends in use of lead in petrol (gasoline) and blood leads in
the general population of the UK over the period when much of this
decline took place. It is interesting to note that from 1971 to 1985 use of
lead in petrol was relatively steady, but blood leads declined by a factor
of more than two over this period mainly as a response to reductions in
dietary exposure, particularly associated with the cessation of use of
leaded solder to seal food cans. A dramatic reduction in gasoline lead
usage occurred at the end of 1985 when the maximum permissible lead
content of petrol was reduced from 0.4 g 1~~l to 0.15 g 1~*, and there has
been a steady reduction in lead use since, with the increased market
penetration of unleaded fuel. Despite the ability of a vehicle emitting lead
to cause direct lead exposure through the atmosphere, as well as indirect
exposure through contamination of food and water, the lack of any
obvious step change in blood lead associated with the reduction of lead
in petrol shows clearly that at that time leaded petrol was not a major
source of exposure for the general population.
5. ENVIRONMENTAL PARTITIONING OF LONG-LIVED
SPECIES
For chemical species sufficiently long lived to approach some form of
equilibrium between environment media, partition coefficients are an
extremely useful means of expressing their likely ultimate distribution.
The best known of these is ATOW, the octanol-water partition coefficient
which is defined as follows:
concentration in octan-1-ol
concentration in water
Since for many of the compounds such as PCBs, dioxins, and PAH to
which this concept is applied, the value of Kov/ is relatively large, it is
often expressed as its logarithm, log ATOW. This is taken as a measure of
the bioaccumulative tendencies of an organic compound as it approximates to the lipid/water partition coefficient. It is predominantly
dependent on water solubility as the variation of solubility for the
various organic compounds in octan-1-ol is relatively modest. For
classes of compounds such as the dioxins, the value of Kow typically

increases with the relative molecular mass of the organic compound. In
its simplest use Kow might be used to predict the likely concentration of
an organic chemical in fish tissues relative to that in the surrounding
water.
A further useful partition coefficient is Koc which expresses the
partition of a chemical between water and natural organic carbon and
has units of dm3 kg" 1 . The utility of Koc is in describing the likely
partitioning of a chemical into soil organic matter or uptake by plants
and animals. Koc is closely correlated to J^ow but is dependent on the kind
of organic carbon considered.
A branch of numerical modelling termed fugacity modelling uses
partition coefficients such as Koc, K0^ and the Henry's Law constant
describing the partition between air and water to predict the distribution
of persistent organic chemicals in a model environment. Whilst the
approach is not sufficiently sophisticated to give exact predictions of
concentrations in environmental media, it is nonetheless very valuable in
predicting in general terms the behaviour of chemicals within the
environment and comparing the partitioning of related compounds
with different physico-chemical properties.
Questions
1. Discuss what is meant by a biogeochemical cycle, describing the
major facets and illustrating the processes involved with examples.
2. Discuss the temporal trends in lead emissions and concentrations in
the environment and how environmental media can be used to
elucidate historical trends in environmental lead.
3. Explain what is meant by an environmental lifetime and derive an
expression for environmental lifetime in terms of a chemical rate
constant. Compare and contrast the typical atmospheric lifetimes
of methane, nitrogen dioxide and the CFCs and explain how this
relates to the atmospheric distribution and properties of these
compounds.
4. Explain the processes by which trace substances can exchange
between the atmosphere and the oceans and show how rates of
exchange can be calculated. Give examples of substances whose
exchange between these media is important.
5. Explain why the waters in rivers in different parts of the world have
differing composition and relate this to the climatology of the
region. Explain carefully what is meant by dissolved and suspended
solids and explain how both arise.
6. Explain the environmental pathways followed by lead emissions
from road traffic after emission to the atmosphere and explain how

this can lead to pollution of a range of environmental media.
Indicate the quantitative ways in which such transfer can be
expressed.
7. Estimate atmospheric lifetimes for the following:
(a) methane, if the globally and diurnally averaged concentration
of hydroxyl radical is 5 x 105 cm" 3 .
(b) nitrogen dioxide in the middle of a summer day when the
concentration of hydroxyl radical is 8 x 106cm~3.
(c) nitrogen dioxide at nighttime if the sole mechanism of removal
is dry deposition with a deposition velocity of 0.1 cms" 1 , and
the mixing depth is 100 m.
8. If the atmospheric concentration of sulfur dioxide is 10 ppb,
calculate the following:
(a) the atmospheric concentration expressed in /ig m~ 3 at one
atmosphere pressure and 250C.
(b) the deposition flux to the surface if the deposition velocity is
1.0 cm s" 1 .
(c) the atmospheric lifetime with respect to dry deposition for a
mixing depth of 800 m.
(d) the atmospheric lifetime with respect to oxidation by hydroxyl
radical if the diurnally averaged OH* radical concentration is
8 x 105 cm""3 and the rate constant for the SO2-OH* reaction
is 9 x 10~ 13 cm 3 molec~ 1 s" 1 .
(e) the lifetime with respect to wet deposition if the washout
coefficient is 10" 4 S" 1 .
9. Explain the thermodynamic controls on biological reduction processes in aquatic sediments and explain how these influence the
chemical forms of nitrogen in the environment.

CHAPTER 7

Environmental Monitoring Strategies
C. NICHOLAS HEWITT AND ROBERT ALLOTT

1 OBJECTIVES OF MONITORING
The gathering of information on the existence and concentration of
substances in the environment, either naturally occurring or from
anthropogenic sources, is achieved by measurement of the substance or
phenomenon of interest. However, single measurements of this type
made in isolation are virtually worthless, since temporal and spatial
variations cannot be deduced. Rather, it is necessary to monitor the
parameter of interest by repeated measurements made over time and
space, with sufficient sample density, temporally and spatially, that a
realistic assessment of variations and trends may be made.
Monitoring of the environment may be undertaken for a number of
reasons and it is important that these be defined before sampling takes
place. The generalization that 'monitoring is done in order to gain
information about the present levels of harmful or potentially harmful
pollutants in discharges to the environment, within the environment
itself, or in living creatures (including ourselves) that may be affected by
these pollutants'1 may be expanded as follows:
(a) Monitoring may be carried out to assess pollution effects on
humans and their environment, and so to identify any possible
cause and effect relationships between pollutant concentrations
and, for example, health effects, or environmental changes.
(b) Monitoring may be carried out in order to study and evaluate
pollutant interactions and patterns. For example source appor1

Department of the Environment, 'The Monitoring of the Environment in the United Kingdom',
Report by the Central Unit on Environmental Pollution, HMSO, London, 1974.

(c)

(d)
(e)
(f)

tionment2 and pollutant pathway studies usually rely on environmental monitoring.
Monitoring may be carried out to assess the need for legislative
controls on emissions of pollutants and to ensure compliance with
emission standards. An assessment of the effectiveness of pollution legislation and control techniques also depends upon subsequent monitoring.
In areas prone to acute pollution episodes, monitoring may be
carried out in order to activate emergency procedures.
Monitoring may be carried out in order to obtain a historical
record of environmental quality and so provide a database for
future use in, for example, epidemiological studies.
Monitoring may also be necessary to ensure the suitability of
water supply for a proposed use (industrial or domestic) or to
ensure the suitability of land for a proposed use (for example for
housing).

A basic problem in the design of a monitoring programme is that each
of the above reasons for carrying out monitoring demands different
answers to a number of questions. For example, the number and location
of sampling sites, the duration of the survey, and the time-resolution of
sampling will all vary according to the use to which the collected data are
to be put. Decisions on what to monitor, when and where to monitor,
and how to monitor are often made much easier once the purpose of
monitoring is clearly defined. Therefore it is most important that the first
step in the design of a monitoring programme should be to set out the
objectives of the study. Once this has been done then the programme
may be designed by consideration of a number of steps in a systematic
way (see Figure 1) such that the generated data are suitable for the
intended use. It is important also that the data produced by a monitoring
programme should be continuously appraised in the light of these
objectives. In this way, limitations in the design, organization, or
execution of the survey may be identified at an early stage.
The aim of this chapter is to present and discuss the most important
and relevant considerations that must be taken into account in the design
and organization of a monitoring exercise. It is not intended to be a
manual or practical guide to monitoring; rather it is hoped that it
highlights the types of approaches that may be used and some of the
problems likely to be encountered. The inclusion of case studies and
references direct the reader to the more specific practical information
which is available elsewhere.
2

P . K. Hopke, 'Receptor Modelling in Environmental Chemistry', John Wiley & Sons, New York,
1985.

objectives
site selection and number

parameters to be measured

duration of survey

sampling methods
equipment selection
analytical techniques
calibration methods
data recording
data analysis
data presentation
information dissemination
Figure 1 Steps in the design of a monitoring programme

2 TYPES OF MONITORING
The Earth's surface is comprised of four distinct media; the atmosphere,
the hydrosphere, the biosphere, and the land. Pollutants can occur in any
of the solid, liquid, or gaseous phases. However, the environment is not a
simple system and consequently each of the four media may contain
pollutants in each of the three phases. Monitoring may therefore be
necessary for a particular pollutant in a specific phase in a particular
environmental compartment (e.g. sulfur dioxide in air) or it may
encompass two or more phases and/or media (e.g. dissolved and
particulate phase metals in water). Pollutants in the environment
originate from a multitude of different types of sources and the
identification of these is a prerequisite to the design of a monitoring
programme.
First, pollutant sources may be classified by their spatial distribution
as point sources, line sources, or area sources. Point sources include
industrial chimneys, liquid waste discharge pipes, and localized toxic
waste dumps on land. Line sources may include highways, airline routes,

and runoff from agricultural land, while area emissions may arise from
extensive urban or industrial complexes. Sources may also be classified as
either stationary or mobile, motor vehicles being the obvious example of
the latter. Classification may also be made for air pollutant sources on
the basis of the height of discharge, i.e. at street level, building level, stack
level, or above the atmospheric boundary layer level. A further important distinction may be made between 'planned', 'fugitive', and 'accidental' emissions to the environment.
(a) Planned emissions arise when (as is invariably the case) it is
economically or technically impossible to completely remove all
the contaminants in a discharge and hence the process operation
allows pollutants to be discharged to the environment at known
and controlled rates. Obvious examples of planned emissions
include sulfur dioxide from power generation plants and lowlevel radioactive effluent during nuclear fuel reprocessing.
(b) Fugitive emissions arise when pollutants are released in an
unplanned way, normally without first passing through the
entire process. They therefore occur at a point sooner in the
process than the stack or duct designed for 'planned' emissions.
They generally originate from operations which are uneconomic
or impractical to control, have poor physical arrangements for
effluent control, or are poorly maintained or managed. An
example is the escape of heavy metal contaminated dust from a
factory on vehicle tyres, arising from poor dust control and wheel
washing arrangements.
(c) Accidental emissions result from plant failure, such as a burst filter
bag or faulty valve or from an accident involving either equipment or operator error (e.g. the Chernobyl reactor accident).
Accidental emissions can give rise to very high mass emission
rates and ambient concentrations but they normally occur only
infrequently.
Classification of the sources of pollutants in this way allows the
distinction of two differing approaches to their monitoring. On the one
hand, samples may be taken of the effluent before discharge to, and
dispersion in, the environment (source monitoring). Alternatively,
samples may be taken of the ambient environment into which discharges occur, for example of the air or receiving waters, without
consideration of source strengths and rates. Obviously neither one of
these approaches alone can necessarily provide all the data required to
resolve a particular problem and often it is desirable to complement one
with the other.

2.1

Source Monitoring

2.1.1 General Objectives. Source monitoring may be carried out for a
number of reasons:
(a) Determination of the mass emission rates of pollutants from a
particular source, and assessment of how these are affected by
process variations.
(b) Evaluation of the effectiveness of control devices for pollution
abatement.
(c) Evaluation of compliance with statutory limitations on emissions
from individual sources.
2.1.2 Stationary Source Sampling for Gaseous Emissions. A common
feature of many industrial processes is that effluent output rates exhibit
cyclical patterns. These may be related to working shift arrangements or
be a function of the operations involved, but both require that source
testing or monitoring be planned accordingly. Process operations should
be reviewed so that discharges during the period of sampling are
representative of the plant output in order to ensure that the samples
themselves are representative of the effluent, and that the final pollutant
analysis will be a representative measure of the entire output.
Two requirements may be specified for valid source monitoring. First,
the sample should accurately reflect the true magnitude of the pollutant
emission at a specific point in the stack at a specific instant of time. This
requirement is met by adequate sampling instrument design. Secondly,
enough measurements should be obtained over time and space so that
their combined result will accurately represent the entire source emission.
This requires consideration of the emissions both in time and in space,
across the entire cross-section of the stack.
In a circular flue, sampling at the centroids of equal area annular
segments will ensure that emission variations across the stack crosssection are quantified. In a rectangular flue sample points should be
located at the centroids of smaller equal area rectangles. Generally eight
or twelve such sampling points are adequate to compensate for any
deficiencies in the location of the sampling site with respect to the length
of the stack and to non-ideal flow conditions at the site caused by bends,
inlets, or outlets. If it is a particulate pollutant which is being sampled
within the stack, it is important that an isokinetic sampling regime is
maintained.
2.1.3 Mobile Source Sampling for Gaseous Effluents. Vehicle and
aircraft emissions are heavily dependent upon the engine operating
mode (i.e. idling, accelerating, cruising, or decelerating) and the results

obtained by sampling must be considered specific to the type of operating
cycle used during the test. Emission tests are usually performed with the
vehicle on a dynamometer equipped with inertia fly wheels to represent
the vehicle weight and brake loading on a level road.
2.1.4 Source Monitoring for Liquid Effluents. Liquid wastes and
effluents often tend, like gaseous effluents, to be inhomogeneous and
care is needed in selecting sampling positions. Having considered the
location of the site in relation to plant operation (e.g. should the site be
before or after a particular stage of the process or treatment) it is
desirable that a region of high turbulence and/or good mixing be
chosen. As for gaseous emissions, several samples may have to be taken
across the cross-section of a pipe or channel. Sampling from vertical
pipes is less liable to be affected by deposition of solids than sampling
from horizontal pipes, and a distance of approximately 25 pipediameters downstream from the last inflow should ensure that mixing of
the two streams is essentially complete.3 If suitable homogeneous regions
for sampling cannot be found, particularly where suspended materials
are present, samples may have to be taken from several positions along
the effluent stream.
Where the composition of a liquid effluent is known to vary with time,
grab samples may be collected at set intervals, either manually or by use
of an automatic sampler. An alternative approach is to sample at
intervals varying with the flow rate so that a more representative
composite may be obtained.
2.1.5 Source Monitoring for Solid Effluents. Solid effluents may arise
from a number of different processes, including sludge after sewage
treatment, ash residue from municipal incinerators, or low-grade
gypsum from desulfurization plants attached to coal-fired power stations. In general, solid wastes are even less homogeneous than either
liquid or gaseous effluents. Therefore, great effort must be made to
ensure that samples are representative of the bulk waste (see Section 3.3).
Monitoring of sewage sludge is particularly common due to sludge
acting as an efficient sorption material for heavy materials. Typically,
80-100% of the input lead in a sewage treatment plant is incorporated
into the sludge, resulting in sludge lead concentrations of 100-3000 mg
g""1. Consideration must therefore be given to the concentrations of
pollutants in the material before it is used as fertilizer, incinerated,
dumped at sea, or used as landfill. The determination of the metal
balance of a sewage treatment works may be necessary when considering
the fate of the treated effluent and solid waste.
3

A. L. Wilson, 'Examination of Water for Pollution Control', ed. M. J. Suess, Pergamon, London,
1982,VoI. 1.

In most countries guidelines exist to control the disposal of sewage
sludges to land, usually based primarily upon the zinc, copper, and nickel
content of the sludge. Hence considerable quantities of other metals,
including lead, may be added to land over a normal 30 year disposal
period. In the UK the disposal of lead-rich sewage sludges to land is
controlled where direct ingestion by animals of contaminated grass or
soil can occur.
Until fairly recently most trace metal analysis of environmental
samples was designed to give a measure of the total elemental concentration in the sample, as it was felt that this gave an adequate measure of the
pollution load for that metal. However, in the past two decades it has
become apparent that total metal concentrations are often not sufficient
and that information based upon some form of physico-chemical
speciation scheme is required. This may include, for example, solubility
of the pollutant in acids of different strengths, the size distribution of
particles, and the association with organic compounds. This is because
the physical, chemical and biological responses to a pollutant will vary
according to its physical and chemical speciation. One disadvantage of
this type of analysis is that it is complicated and time-consuming
compared with total metal determinations. Thus speciation studies are
invariably limited to a few samples, where many (tens or even hundreds)
would be taken in a total-metal study.
Case Study 1: Organic solvent residues at a landfill site*

Landfill sites

are now recognized as sources of toxic and explosive substances,
including methane and organic chemicals. The contamination of groundwater by these toxic organic chemicals is of major environmental concern
in Europe and North America. At a landfill site studied near Ottawa,
Canada, disposal of chlorinated and non-chlorinated solvents, wood
preservatives, and small amounts of other wastes occurred between 1969
and 1980. Groundwater samples were collected from monitoring wells
within the landfill site using either piezometers or multilevel samplers
attached to peristaltic pumps. Analysis was carried out by gas chromatography-mass spectrometry (GC-MS) which enabled the identification
and quantification of a wide range of volatile organic compounds,
including dioxane (-300-2000/ig I" 1 ), diethylether (<2-658/ig I" 1 ),
trichloroethene (7-583 jug I" 1 ), and l,l,2-trichloro-l,2,2-trifluorethane
(Freon Fl 13) (< 5-2725 fig I" 1 ). The contaminant of greatest concern
was 1,4-dioxane, due to its toxicity and persistence. Freon Fl 13 was the
organic chemical found in greatest concentration. Although very persistent in the subsurface, it appeared to have undergone transformation, as
one toxic product, F-1113, was identified.
4

S . Lesage, R. E. Jackson, M. W. Priddle, and P. G. Riemonn, Environ. Sci. TechnoL, 1990, 24, 559566.

2.2

Ambient Environment Monitoring

2.2.1 General Objectives. Monitoring the environment may be carried
out for a number of reasons, as outlined in Section 1. However, whatever
the purpose of the survey the overriding consideration when designing a
programme is to ensure that the samples obtained provide adequate data
for the purpose intended. Invariably this means that samples should be
representative of conditions prevailing in the environment at the time
and place of collection. Thus, not only must the sampling location be
carefully chosen but also the sampling position at the chosen location.
The selection of a specific monitoring site requires consideration of
four steps: identification of the purpose to be served by monitoring;
identify the monitoring site type(s) that will best serve the purpose;
identification of the general location where the sites should be placed;
and final identification of specific monitoring sites.
2.2.2 Ambient Air Monitoring. Air pollution problems vary widely
from area to area and from pollutant to pollutant. Differences in
meteorology, topography, source characteristics, pollutant behaviour,
and legal and administrative constraints mean that monitoring programmes will vary in scope, content, and duration, and the types of
station chosen will also vary. However ambient monitoring sites may be
divided into several categories:
(a) source-orientated sites for monitoring individual or small groups
of emitters as part of a local survey (e.g. one particular factory).
(b) sites in a more extensive survey which may be located in areas of
highest expected pollutant concentrations or high population
density, or in rural areas to give a complete nationwide or regional
coverage.
(c) baseline stations to obtain background concentrations, usually in
remote or rural areas with no anticipated changes to land use.
Location of source-orientated monitors. Occasionally the effects and
impact of a specific pollutant source are of sufficient interest or
importance to warrant a special survey. This will usually include a site
at the point of anticipated maximum ground-level concentration, which
can be estimated from dispersion calculations (see Section 4.1 below),
and also a nearby site to characterize the 'background' conditions in the
area. Examination of meteorological records will usually be necessary in
order to choose suitable locations for the sites and several computerized
models are available for determining the areas of maximum average
impact from a point source. Calculation of expected ground-level
concentrations using the standard equations discussed in Section 4.1

(Concentration x Wind speed) / (Emission rate) / rrf2

Distance/km
Figure 2

Normalized ground-level concentrations from an elevated source for neutral
stability. The effective stack height (H) is the sum of the release height (e.g.
chimney height) and the height gained by the plume due to momentum and
buoyancy

show that the concentration rises rapidly with distance from the source
to a maximum and then falls gradually beyond the maximum,5 as shown
in Figure 2.6 This is for meteorological conditions of neutral stability and
different heights of emission (H). The ordinates in this graph represent
concentration normalized for emission rate (Q) and wind speed (U) and
the various curves are for different source heights (H metres) and
different limits to vertical dispersion (L). It is prudent therefore to
54
6

WMO Operations Manual for Sampling and Analysis Techniques for Chemical Constituents in
Air and Precipitation', World Meteorological Organization, No. 299, Geneva, 1971.
D. B. Turner, 'Workbook of Atmospheric Dispersion Estimates', National Air Pollution Control
Administration, US Environmental Protection Agency, Research Triangle Park, NC, 1970.

locate the monitoring site somewhat beyond the distance where the
maximum concentration is predicted. This allows some margin for error
by placing the monitor in a region of relatively small concentration
gradients. Obviously it is desirable to have an array of stations at
differing distances and directions from the source and typically 4-6
samplers might be considered sufficient for monitoring a single point
source.
In some cases pollutants are emitted to the atmosphere from a single
source but in a more diffuse manner than from a single stack. Calculations of mass emission rates and distance of maximum ground-level
concentration are more difficult to make for such diffuse or fugitive
emissions which, in some cases, may have significant impacts on the local
air quality.
Location of monitors in larger-scale surveys. Often it is important to
know the geographical extent of atmospheric pollution, and to have
localized information on source strengths or ground-level concentrations
within a plume is not sufficient. For example, the National Survey of Air
Pollution (NSAP) monitoring network in the United Kingdom was
established in 1961 following recognition of the need for the acquisition
of a nationwide, day-to-day, and long term bank of data of sulfur
dioxide and smoke concentrations. The original network of 1200 sites
was based upon the assumption that it was necessary to monitor in rural
areas (150 sites) and different types of urban areas such as high-density
residential areas, industrial areas, commercial areas, and smoke-controlled areas. Since the introduction of this simple scheme in 1961 many
of the original stations have ceased monitoring, others have been
replaced, and additional stations have been added. Reasons for these
changes include the need to monitor recently established smoke-control
areas, new industrial estates and redeveloped areas, and surveys around
new and projected power stations. In 1981 a rationalized long-term
network of 150 NSAP stations was established, re-designated as the UK
Smoke and Sulphur Dioxide Monitoring Network. However, about 400
existing sites were retained in the short term to provide continuation of
monitoring in urban areas where the EC air quality standards for smoke
and SO2 may be approached or exceeded. As concentrations fall to
'acceptable' levels in each urban area so these sites are discontinued.
These sites have, in the last five years, been supplemented by a number
of fully automated sites in urban and rural areas of the UK giving online, real-time data accessible to the public via the Internet, freephone,
and teletext. The pollutants monitored at these sites vary, but the most
comprehensively equipped sites determine SO2, NO, NOx, O3, TSP, CO,
and a wide range of volatile organic compounds. In contrast the simplest

sites only monitor ozone. A full description of this programme is
available elsewhere.7
Several methods are available for the rationalization of an existing
monitoring network, and in the case of the NSAP network spatial
correlation analysis was applied to determine which sites might be
discontinued without significantly losing overall coverage.
The design of monitoring networks for air pollution has been treated
in several different ways. For example monitoring sites may be located in
areas of severest public health effects, which involves consideration of
pollutant concentration, exposure time, population density, and age
distribution. Alternatively the frequency of occurrence of specific
meteorological conditions and the strength of sources may be used to
maximize monitor coverage of a region with limited sources. An estimate
of the ambient dosage (the product of the concentration and the
exposure time) computed from source emission data and diffusion
models can be used to determine where air monitoring sites should best
be located.
Location of regional-scale survey monitors. On the regional or global
scale, monitoring is usually concerned with long term changes in background concentrations of pollutants and so the principal siting requirement is that truly representative baseline or background levels may be
measured over a long term period without interference from local
sources. It has been suggested5 that baseline stations should be located
in areas where no significant changes in land use practices are anticipated
for at least 50 years within 100 km of the station, and should be away
from population centres, major highways, and air routes. Such locations
are hard to find, but Cape Grim in Tasmania is one example.
One example of a network established to monitor distant sources for
regional or global effects is the OECD 'Long-Range Transport of Air
Pollutants' project which measured chemical components in precipitation and SO2 and particulate sulfate in air. The long term measurement
of carbon dioxide concentrations in air at Mauna Loa in Hawaii is
probably the best example of 'baseline' monitoring.
Case study 2: National and regional radon surveys in the UK. Increasing
awareness of the importance of human exposure to the naturally
occurring radioactive gas radon, has resulted in several national surveys
being commissioned for the UK. 222Rn, commonly known as radon, is a
colourless odourless gas which results from the decay Of238U and in turn
decays to a series of radioactive daughters. Two other isotopes of radon
exist, 219Rn and 220Rn (thoron) within the 232Th and 235U radioactive
7

G . Davison and C. N. Hewitt, 'Air Pollution in the United Kingdom', Royal Society of Chemistry,
London, 1997.

decay series. However, it is 222Rn which has the greatest health
significance, since it has the largest proportion of alpha emitting, high
activity progeny. The inhalation of radon and particularly its daughters
is believed to be responsible for a major proportion of the overall annual
dose of ionizing radiation received by the UK population.
Several methods exist of the measurement of radon in air and may be
divided into active or passive techniques. The active techniques are
generally only used for research or special survey purposes and normally
consist of a pump which draws air through a filter trapping the radon
decay products. The alpha radiation from these daughters is measured
by scintillators or semiconductor detectors. The most common method
for monitoring radon is the passive alpha track detector and these have
been extensively used in the UK.8 Radon is allowed to diffuse into a
small container, but radon daughters present in the ambient air are
excluded. Inside the pot is placed a piece of polycarbonate, cellulose
nitrate, or allyl diglycol carbonate film. Alpha particles, formed by the
decay of radon, damage chemical bonds in the surface of the film. After a
period of exposure, the surface of the film is etched with NaOH and
tracks appear. The number density of these tracks is proportional to the
average radon concentration and may be either counted using a
computerized image analysing system under a microscope, or the tracks
filled with a scintillant (ZnS-Ag) which fluoresces when exposed to an
alpha source in proportion to the number of tracks present. The
detectors may be calibrated by exposure to known radon concentrations
for known periods of time. Exposure times of weeks to months are
required with this method.
To date, the National Radiological Protection Board (NRPB) has
surveyed about 200000 dwellings in the UK. The distribution of
observed radon concentrations was log normal, with a populationweighted mean of 20 Bq m~ 3 . On the basis of these measurements,
about 0.5% of the housing stock in the country (100000 houses) is
believed to have radon concentrations above the Government's Action
Level of 200 Bq m~~3, and so requiring remedial work to be taken. This
can be done by preventing the gas entering the building (e.g. by sealing
floors), increasing the ventilation rate (e.g. by extractor fan), or removing
the radon decay products from the air (e.g. by electrostatic precipitator).
The primary influence of bedrock geology on relative radon concentrations throughout the UK is exhibited in Figure 3. High levels of radon
in homes (reaching 8000 Bq m~3) were found in areas where the bedrock
contained high concentrations of uranium. These included granites and
8

National Radiological Protection Board, 'Exposure to Radon in UK Dwellings', NRPB-R272,
HMSO, London, 1994.

Figure 3

Relative radon concentrations in homes throughout the UK

(Reproduced with permission from the National Radiological Protection
Board)

areas of mineralization in south-west England and Scotland (up to
2000 ppm U) and some shales and limestones in north and central
England (about 800 ppm U).
Other factors which influence radon levels were also identified, including season, time of day, meteorological conditions, ventilation {e.g.
existence of double glazing), usage, and other occupancy habits. These
can result in a difference factor of two between summer and winter
concentrations. Furthermore, radon concentrations may vary between
rooms, with the highest radon levels in basements and, on average, first
floor bedrooms have concentrations two-thirds those of living rooms.
2.2.3 Environmental Water Monitoring. Pollutants enter the aquatic
environment from the air (by dry deposition or in precipitation occurring
either directly onto the water surface or elsewhere within the catchment
area), from the land (either in surface runoff or via sub-surface waters)
and directly through effluent discharges (domestic, industrial, or agricultural). The undesirable effects of pollutants in natural water may be due
to:
(a) stimulation of water plant growth—eutrophication—which ultimately leads to deoxygenation of the water and major ecological
change;
(b) their direct or indirect toxic effects on aquatic life;
(c) the loss of amenity and practical value of the water body,
particularly as a source of water for public supply.

Apart from the monitoring of sources of pollutants in liquid effluents
(Section 2.1.4) sampling may be carried out:
(a) in rivers, lakes, estuaries, and the sea in order to obtain an overall
indication of water quality;
(b) for rainwater, groundwater and runoff water (particularly in the
urban environment) to assess the influence of pollutant sources;
(c) at points where water is taken for supply, to check its suitability
for a particular use;
(d) using sediments and biological samples in order to assess the
accumulation of pollutants and as indicators of pollution.
As well as the measurement of chemical and physical parameters the
quantitative or qualitative assessment of aquatic flora and fauna is often
used to give an indication of the presence or absence of pollution, and
well recognized relationships exist between the abundance and diversity
of species and the degree of pollution. This is often used to assess the
cleanliness of natural fresh waters and is known as biological monitoring.
Location of sampling sites. There are two main causes of heterogeneous
distribution of quality in a water body. These are
(a) if the system is composed of two or more waters that are not fully
mixed (such as in themally stratified lakes or just below an effluent
discharge in a river) and
(b) if the pollutant distributes non-uniformly in a homogeneous water
body (for example oil which tends to float, and suspended solids
which tend to settle out of the water). Also chemical and/or
biological reactions may occur non-uniformly in different parts
of the system, so resulting in heterogeneous pollutant concentrations. When the degree of mixing is unknown it is advisable to
conduct a preliminary survey before deciding on sampling locations. Rapidly obtainable data of water temperature, pH, dissolved oxygen, or electrical conductance may be used in this
respect.
Sampling locations should generally be at points as representative of
the bulk of the water body as possible, e.g. away from river or lake banks
or the walls of channels or pipes, but often it will be desirable (and
necessary) to take samples from several locations in order to obtain the
required information.

concentration

time

concentration

(a) Sampling location close to discharge

time

concentration

(b) Intermediate sampling location

time
(c) Sampling location distant from discharge
Figure 4

Schematic diagram of the dependence of pollutant concentration on the distance
downstream from a cyclically varying waste discharge

When sampling from rivers and streams downstream of effluent
discharge longitudinal, transverse, and vertical sampling arrays may be
necessary to ensure that truly representative data are obtained. Studies of
some pollutants require sampling at considerable distances downstream
of effluent inputs, e.g. in investigating the sag in dissolved oxygen
content. When a temporally varying effluent discharge is under study it
may be desirable to sample as close to the point of discharge as mixing
allows in order to monitor short term variations in concentration.
However, if long term average water quality is of interest then sampling
should be carried out further downstream where longitudinal dispersion
and mixing will have smoothed out the short term variations (see Figure
4).
Sampling in estuaries presents special problems as great spatial and
temporal variability may be exhibited. The appropriate locations for
sampling will vary from estuary to estuary and will depend on the

parameters of interest, but a minimum of 50 samples per survey might be
appropriate. If one considers a compound which has an input at only one
end of an estuary and which is not removed from or added to solution
during its lifetime in the estuary then the concentration of that
compound in the estuary will be solely dependent upon the dilution
ratio between the river water and the seawater. Thus, for example, the
concentration of chloride ion or salinity is dependent only upon the
mixing of the fresh and the saline water bodies. This concept of
'conservative' behaviour is an important one which must be taken into
account when monitoring estuarine concentrations. If a graph is drawn
of the concentrations of the element of interest in an estuary against
salinity then the data points will fall on a straight line (the theoretical
dilution line) if physical mixing is the only process controlling the
concentration of that element in the water. However, if the element of
interest is added to or removed from the solution during mixing then the
data will not plot on a straight line. In the case of lakes and reservoirs
vertical stratification of pollutants may be very pronounced due to a
reduction in dissolved oxygen from the surface downwards. A minimum
of three samples is then probably necessary, at 1 m below the surface,
I m above the bottom, and at an intermediate point.
When water is abstracted from a river, lake, reservoir or aquifer,
samples should be regularly taken at the point of abstraction and at the
point where the water enters the distribution system. Several excellent
handbooks with full descriptions of water sampling and analytical
methods are available.9"12
The determination of concentrations of trace metals in natural
waters is a fundamental stage in the calculation of their budgets or
cycles, but is subject to the same problems of sample contamination as
occur for atmospheric samples from remote areas. All stages of the
analysis, from sampling collection, storage, and filtration to actual
laboratory manipulation require care to prevent contamination occurring. Indeed for many years measured levels of many trace elements in
seawater were purely an artifact of contamination during sampling
and analysis.
Case study 3: The behaviour and variations of caesium and plutonium in

estuarine waters. Plutonium and caesium isotopes, in addition to other
radionuclides, have been discharged into the Irish Sea in effluent from
9

APHA, 'Standard Methods for the Examination of Water and Wastewater', American Public
Health Association, New York, 15th Edn., 1980.
10
M. J. Suess, 'Examination of Water for Pollution Control—a Reference Handbook', 3 VoIs.,
WHO/Pergamon, Copenhagen, 1982.
II
J. W. Clark, W. Viessman, and M. Hammer, 'Water Supply and Pollution Control', Harper, New
York, 3rd Edn., 1977.
12
L. G. Hutton, 'Field Testing of Water in Developing Countries', Water Research Centre, 1983.

the nuclear fuel reprocessing plant at Sellafield since 1953. In this
study,13 the plutonium and caesium activity concentrations were monitored in the dissolved phase of estuarine waters, in order to investigate
the behaviour of these radionuclides. Water samples were collected by
boat at mid-channel, during spring and neap tides, and along the Esk
Estuary (10 km from Sellafield discharge point) and its tributaries (Irt
and Mite). Samples of 1 or 2 dm3 were collected using instantaneous
isokinetic samplers at near surface (0.5 m depth) on high water surveys
and mid-depth on tidal cycle surveys. In addition, a 10 dm3 sample of
river water was collected from the River Esk, well above the tidal limit.
Fractionation into particulate and dissolved phases was achieved by
filtration with 0.22 mm Millipore filters. Salinities were measured using a
Goldberg refractometer and activity concentrations by y-spectrometry
and a-spectrometry following standard preparation and chemical
separation procedures.
Figure 5 shows the variation in plutonium and caesium dissolved
phase activity (expressed as a percentage of the maximum 'seawater'
value for each study) with salinity. Figure 5a clearly indicates that 137Cs
activities follow a theoretical dilution line between seawater and river
water end members, exhibiting conservative behaviour. Therefore,
caesium does not appear to re-equilibrate between the particulate and
dissolved phases throughout estuarine waters of different salinities.
However, dissolved plutonium activities do not follow a theoretical
dilution line (Figure 5b). Higher activities are apparent in the dissolved
phase, for salinities below 18%o, than predicted by conservative mixing.
In addition, a wide variation of activities were recorded at high salinities.
Thus monitoring of 239 + 240Pu dissolved phase activity concentrations
has shown that plutonium is apparently being released from the
particulate phase at low salinities in the Esk Estuary.
Case study 4: Acidification of lakes. The acidification of lakes in North
America and Northern Europe (Scotland, Norway, and Sweden) has
been directly linked to anthropogenic activities, in particular sulfur
dioxide emissions from coal fired power stations. Much of the evidence
for increase in acidity has been indirect, such as loss of fish populations,
changes in aquatic plant and invertebrate communities, changes in
sediment metal concentrations, and results from empirical models of
lake acidification. A study of the Adirondack lakes in New York state
made direct comparisons of historical and recent lake survey data in
order to determine if significant acidification had occurred in that
region.14
13
14

D. J. Assinder, M. Kelly, and S. R. Aston, Environ. Technol. Lett., 1984, 5, 23-30.
C. E. Asbury, F. A. Vertucci, M. D. Mattson, and G. E. Likers, Environ. Sci. Technol., 1989, 23,
362-365.

Activity %

(a)

Activity %

Salinity %

(b)

Salinity %
Figure 5

137

Variation of (a) Cs and (b) 239-240Pu dissolved phase activity and salinity in
the Esk Estuary, UK (lines refer to level of confidence in theoretical mixing line)
(Reproduced with permission from Environ. Tech. Lett., 1984 5 27 Selper
Ltd)
' '

The historical data (1929-1934) included measurements of alkalinity
(a measure of the acid neutralizing capacity of the water), pH, and CO2
acidity. However, the pH data were considered to be unreliable due to
the use of colourimetric indicator solutions which can alter the pH of a
solution being measured. It was therefore decided to directly compare
alkalinity values from this data set with extensive data collected
between 1975 and 1985. Alkalinities for each lake were matched on the
basis of a unique 'pond number' system adopted by the surveys. Data
was excluded where lake identification was ambiguous, the lakes were
known to have been treated with lime, or it was suspected that the
electrode used for the alkalinity measurements was malfunctioning. The
final data set consisted of 274 lakes. Account was taken of the different
analytical techniques employed to determine alkalinity. The historic
method used titration to a fixed pH end point, determined by the

'faintest pink' colour of the methyl orange indicator. Modern surveys
employed the Gran technique to determine the end point. As a result,
the historical data set were corrected by an alkalinity subtraction of
54.6 /xequiv dm" 3 .
The historical data had a mean 'corrected' alkalinity of 141 /zequiv
dm~ 3 compared to a modern alkalinity of lOO/zequiv dm~3. This
indicates a mean estimated acidification of —41 /xequiv dm~3. This loss
of alkalinity was found to be significant (p < 0.01) by a Wilcoxon signed
rank test. It was concluded that significant acidification had occurred in
the Adirondack Mountain region since the 1930s. Inter-lake differences
in acidification were noted, probably due to factors such as atmospheric
deposition, lake hydrology, geology, or soil type. Sensitivity of lakes to
acidification has been related to crystalline bedrock and thin acid soils,
which are generally associated with high elevation lakes. This study
found a direct relationship between the mean lake elevation and mean
acidification for lakes in each of the major catchments.
2.2.4 Sediment, Soil, and Biological Monitoring. Soils and sediments
may become polluted by a number of routes, including the disposal of
industrial and domestic solid wastes, wet and dry deposition from the
atmosphere, and infiltration by contaminated waters.
The main pollution hazards on land have been identified as follows:1
(a) Harmful substances may get into the soil or plants and so into the
food supply.
(b) Substances may wash from the land and so pollute water supplies.
(c) Contaminants may be re-suspended and subsequently inhaled.
(d) Substances polluting the land may make it potentially dangerous
or unsuitable for future use {e.g. for housing or agriculture).
(e) Ecological systems may be damaged, with consequent loss to
conservation and amenity.
Some potentially harmful substances, such as mercury or lead, are
naturally present in soils but at concentrations which are not normally
deleterious. Some activities, however, can cause elevated levels of these
compounds. For example, mining may cause soils to be contaminated by
metals, and the dumping of solid wastes on land will invariably introduce
a wide variety of pollutants to the soil. On the other hand there are
compounds which do not occur naturally, and their presence in soils and
sediments is due entirely to human activities. These substances include
pesticides, (particularly the organochlorine compounds such as DDT,
toxaphene, aldrin, dieldrin) and artificial radionuclides (e.g. 137Cs,
239
Pu).

As with the other types of media discussed above it is important that
the background levels of pollutants be established in soils, sediments,
and vegetation. One example of a large-scale investigation of this type is
the geochemical survey of stream-bed sediments carried out in England
and Wales.15 Stream sediments are considered to represent a close
approximation to a composite sample of the weathering and erosion
products of rock and soil upstream of the sampling point and in the
absence of pollution provide information on the regional distribution of
the elements.
A second type of monitoring programme is required to establish actual
levels of contamination in land or sediments known or believed to be
affected by pollutants. In this case much more specific and localized
monitoring may be required in order to quantify the degree of contamination. The contamination of sites often arises from their previous uses,
particularly as coal gas manufacturing plants, sewage works, smelters,
waste disposal sites, chemical plants and scrapyards. A typical contaminated site may be found to contain variable concentrations of toxic
elements and organic compounds, phenols, coal tars and oils, combustible material from undecomposed refuse, acidic or alkaline waste
sludges, and sometimes methane accumulations. In addition, contaminated land is often formed by waste tipping and so may be poorly
compacted and very inhomogeneous.
Some sites may contain underground pipework and structures from
their previous uses and so present formidable sampling problems. Site
investigation of this type is very expensive, involving bore holes or trial
pits and, in cases where a detailed site history is unavailable, determination of a large number of pollutants. However, it is most important to be
sure that the investigation is sufficiently rigorous, as remedial measures
are extremely costly and must be based upon adequate data. Common
problems which have been identified are:
— an inadequate number of samples,
— an inadequate range of determinands,
— bulking of samples when individual samples from specific locations
are preferable,
— inappropriate analytical methods,
— inadequate referencing of sample locations,
— inadequate descriptions of samples,
— inadequate descriptions of trial pit strata,
— an ignorance of the nature of the required information.
15

J. S. Webb, I. Thornton, M. Thompson, R. J. Howarth, and P. L. Lowenstein, 'The Wolfson
Geochemical Atlas of England and Wales', Oxford University Press, Oxford, 1978.

Monitoring should be carried out following the application of sewage
sludge or waste waters to agricultural land. Samples of surface water,
groundwater, site soil, vegetation, and sludge applied would normally be
tested for faecal coliform, nutrients, heavy metals, and pH. Details of the
necessary chemical and physical methods may be found elsewhere.9 The
results from the monitoring exercise may be compared to predicted levels
derived from the application rates of sludges to land, soil type, nitrogen,
phosphorus, and heavy metal contents of the waste and the nutrient
uptake characteristics of the cover crop.
When monitoring background levels and more specific pollution on
land or in the sediments of a water body, measurements will often be
made of levels in the plants or organisms that the soil or sediments
support. In many cases flora or fauna provide excellent indicators of the
degree of pollution, as they may act as bioconcentrators (for example of
heavy metals from suspended material in shellfish). Furthermore it is
obviously important to monitor pollution levels in food and through the
food chain. The simultaneous measurements of pollutant levels in soils
and plants as well as in water, sediments, and aquatic biota are therefore
often carried out. However, the relationship between levels in these
various media are often not simple, and sampling and analysis of one of
these is no substitute for a comprehensive monitoring programme. For
example, laboratory experiments indicate that metals are readily
absorbed by some plants but measurements of metals in oil, rainwater,
and plants often reveal a lack of correlation between the corresponding
sets of concentrations.
An example of the large-scale use of aquatic biota as a medium for the
monitoring of pollutants in coastal waters is the US Mussel Watch
Program. This programme began in 1976 with the overall aim of
providing strategies for pollutant monitoring in coastal waters using
mussels and oysters collected on the west, east, and Gulf coasts of the
USA. Analyses for trace metals, chlorinated hydrocarbons, petroleum
hydrocarbons, and radionuclides (including 238Pu, 239 + 2 4 0 p u anc [ 241Am)
were carried out at three laboratories with extensive intercalibration
studies used to ensure compatibility of data. One conclusion of the initial
report of this programme was that for metals the frequency of monitoring need not be yearly, but some small multiple of years for a given site.
Thus for the same resources, human and financial, greater geographical
coverage can be gained by increasing the sampling interval. On the basis
of the data obtained in 1977-1978, national baseline concentrations for
bivalves in unpolluted waters were suggested. The programme also
established the importance of systematic repetitive sampling at the same
sites over periods of years, which allowed elevated cadmium and
plutonium concentrations in some areas to be attributed to coastal

upwelling rather than as a consequence of localized anthropogenic
sources.
Monitoring of concentrations of trace metals in crop plants grown on
sewage sludge amended soils indicates that the levels found will vary with
crop species and properties of the soil substrate on which they have been
grown. More recent studies have concentrated on the physico-chemical
speciation of the metals in the applied material and the receiving soil and
have demonstrated the importance of organic complexes in reducing free
metal activity.
Case Study 5: Metals in dusts. A great number of studies have been
conducted since the 1970s on dust and soil contamination with heavy
metal, in particular lead, due to its predominant source as an anti-knock
agent added to petrol. These studies have established the magnitude of
different source of lead in the environment. However, it is still unclear
whether or not the exposure of young children to relatively small
amounts of lead has a detrimental effect on intelligence, behaviour and
educational and social attainment.
Although various studies have attempted to assess the significance
of different pathways of lead to a child's total lead intake, the study
described here16 was a comprehensive environmental, biological,
behavioural, and dietary investigation of two-year-old children in
Birmingham. From a set of 183 randomly selected children, 97
completed the study. Various blood samples were taken at the
beginning of the study and 56 children also provided a second sample
about five months later.
Environmental monitoring was conducted at each of the children's
homes. Indoor and outdoor air samples were obtained, over seven
consecutive days, using aerosol monitors at a height of 1.5-2 m. House
dust samples were collected from the householder's own vacuum cleaner
and from the child's playroom and bedroom, using a modified vacuum
cleaner. Pavement and road dusts were sampled during dry conditions. A
composite soil sample was also collected. On each of the seven days of
the study, both hands were wiped thoroughly using a total of three wet
wipes. A duplicate of the child's diet over the seven days was provided,
along with a dietary record. Finally, mouthing behavioural patterns were
analysed using a 30 item questionnaire, or interview of the parents and
by recording the child on videotape in set situations. These situations
included a period of free play, lunch, being read a 10 minute story and
watching a 15 minute video film. Each film was analysed to determine
how much time a child touched the floor, objects, and the mouth.
The frequency distributions of lead concentrations in all the samples
16

D. J. A. Davies, I. Thornton, J. M. Watt, E. B. Culbard, P. G. Harvey, H. T. Delves, J. C.
Sherlock, G. A. Smart, J. F. A. Thomas, and M. J. Quinn, Sci. Tot. Environ., 1990, 90, 13-29.

Table 1

Lead in blood, environmental

samples, handwipes, and water
Per cent iles

Sample

Units

Blood

/zg/lOOml

Air
Playroom
Bedroom
External

Mgm~ 3
Mgm" 3
Mgm~ 3

607
599
605

^g g " 1

97
96
94
92
42
97
97

311
464
424
336
615
360
527

105
109
138
97
120
127
195

1030
2040
2093
1440
4300
1340
1170

/*gm~ 2

486

313

1160

Dust
Playroom
Bedroom
Average
Vacuum 3
Doormat
Pavement
Road
Dust'loading'
Soil

fig g "

Handwipes

fig

Geometric
mean

Diet (food & beverages)

/zgweek"

Water

/xgl" 1

704
1

11.7
0.27
0.26
0.43

5.7

5th

95th

0.08
0.09
0.12

1.9

0.88
0.81
1.53

15.1

161

389

100

After Ref. 16.
HousehoIder's own cleaner.

were found to be approximately log normal, consequently geometrical
means were calculated (Table 1). These mean lead concentrations (e.g.
mean indoor dust lead 424/ig g" 1 ) were similar to those found in
previous studies in both Birmingham and elsewhere in the UK, although
exposure to dietary lead was believed to be slightly elevated due to higher
water lead concentrations. The indoor and outdoor air lead concentrations were strongly correlated, as observed in previous studies, with a
mean indoor/outdoor air lead ratio of 0.61. An exception to this
occurred in a house in which old paint was being removed by a machine
sander. On these occasions the indoor lead concentration exceeded that
outdoors. In addition, the dust lead concentration for the householder's
own vacuum cleaner was 23 000 fig g~l.
Relationships between blood lead, environmental lead, and behavioural measures were assessed using multiple regression analysis. This
indicated the importance of the amount of lead in the house dust
combined with a child's rate of hand touching activity to blood lead
levels. Water lead and smoking habits of parents are also significant
factors where blood lead is concerned. These results were used to predict

a mean lead intake of 1 /ig day l via inhalation and 35 jig day l via
ingestion for the Birmingham children.
Case Study 6: Chernobyl-derived radiocaesium in Cumbria, UK. Following the Chernobyl reactor accident on 26 April 1986, a large number
of monitoring studies were conducted in the UK and elsewhere in
Europe. Radiocaesium contamination of sheep reared on upland fells in
Cumbria and North Wales was of particular concern, due to the high
levels of fallout received by these areas (20 kBq m~ 2 137Cs). Restrictions
to movement and slaughter of those sheep with a 137Cs muscle specific
activity concentrations greater than 1000 Bq kg" 1 (fresh weight) were
imposed by the UK government.
One study17 conducted at a farm in West Cumbria investigated
radiocaesium specific activities in local soils, vegetation, and sheep
muscle. Three upland soils and one valley bottom soil were sampled
within the same 1 km2 area, using a 5 cm deep corer. The sampling
programme continued from October 1986 to November 1989 with a
sampling frequency of every two weeks for the first 5 months. Samples
were analysed for total and ammonium exchangeable 134Cs and 137Cs
by y-spectrometry. Vegetation samples were collected from improved
pasture at the study farm and also from the nearby open fell, from
August 1986 until June 1988 with a frequency of once every two weeks.
Live monitoring of the sheep for 137Cs was carried out using a portable
y-spectrometry system which could be taken into the field. A similar
monitoring programme to that for the soil and vegetation was
adopted.
As a result of this survey it was discovered that a high percentage of
the Chernobyl derived 137Cs in upland soils was ammonium exchangeable (86-100%) and thus probably available for uptake by vegetation,
compared to 0% for the valley-bottom soil. This was confirmed by the
generally higher 137Cs content in vegetation from the open fell (September 1987, 140-2080Bq kg" 1 dry weight) compared to the improved
pasture (September 1987, < 100-710 Bq kg" 1 dry weight).
As a consequence of this radiocaesium behaviour, the 137Cs activity
concentrations of ewe muscle (Figure 6) declined when the sheep were
brought on to enclosed pastures, but rose when they were returned to the
open fell. However, despite an overall fall in the 137Cs activity concentration in vegetation with time, the 137Cs levels in sheep were higher in the
summer/autumn of 1987 and 1988 than in the autumn of 1986. This
appeared to be due, at least in part, to an increase in the availability of
radiocaesium originating from Chernobyl fallout, as it had been incorporated into plant material, rather than when it was present as a direct
17

B. J. Howard, N. A. Beresford, and F. R. Livers, 'Transfer of Radionuclides in Natural and SemiNatural Environments', ed. G. Desmet, P. Nassimbeni, and M. Belli, Elsevier, New York, 1990.

Cs-137 Bq/kg Fresh weight

Month (1986-88)
Figure 6

Changes in the 137Cs activity concentrations of ewe muscle as they are moved
between the fell and improved pastures (shading = fell)

(Reproduced with permission from Transfer of Radionuclides in Natural and
Semi-Natural Environments', ed. G. Desmet, P. Nassimbeni, and M. Belli,
Elsevier Science Publishers Ltd, 1990)

deposit on vegetation surfaces as in 1986. Following the accident the UK
Government set up a National Response Plan for dealing with overseas
nuclear accidents, which included establishment of a national radiation
monitoring network and nuclear emergency response system (RIMNET:
Radioactive Incident Monitoring Network).
3 SAMPLING METHODS
3.1 Air Sampling Methods
Sampling systems for airborne pollutants usually consist of four component parts: the intake component, the collection or sensing component,
the flow measuring component, and the air moving device. All these
(other than the collector or sensor itself) must be constructed of materials
which are chemically and physically inert to the sampled air.
3.1.1 Intake Design. The nature of the intake is determined by the
type and objective of the sampling technique, and may vary from a
vertical opening for the passive collection of dustfall in a deposit gauge to
a thin-walled probe used for source sampling of aerosols. Common
problems which may require consideration are the non-reproducible
collection of the sample portion from the air mass due to poor inlet
design, adhesion of aerosols to the tube walls, loss or change of analyte
by chemical reaction with inlet materials, adsorption of gaseous compo-

Cs-137 Bq/kg Fresh weight

Month (1986-88)
Figure 6

Changes in the 137Cs activity concentrations of ewe muscle as they are moved
between the fell and improved pastures (shading = fell)

(Reproduced with permission from Transfer of Radionuclides in Natural and
Semi-Natural Environments', ed. G. Desmet, P. Nassimbeni, and M. Belli,
Elsevier Science Publishers Ltd, 1990)

(a)

(b)

(C)

Figure 7

Schematic diagram of (a.) under-sampling of suspended particles, (b) isokinetic
sampling and (c) over-sampling of suspended particles

nents on inlet materials, and condensation of volatile components within
the transfer lines.
The sampling of aerosols presents particular difficulties in inlet design.
A basic requirement is that the velocity of the sample entering the system
intake should be the same as the velocity of the gas being sampled. This is
necessary because, if the streamlines of the sampled gas are disturbed by
the intake probe, particles travelling in the gas flow and possessing
inertia directed along the streamlines will continue into the probe while
the 'carrier gas' will be diverted away from (if the probe intake velocity is
too low) or into (if the intake velocity is too high) the inlet (Figure 7).
Thus either a greater number of particles per unit volume of gas than
exists in the actual gas flow, or a smaller number, will be collected. Only
when the intake velocity at the face of the probe is equal to the approach
velocity of the gas stream will the streamline pattern remain unaltered
and the correct number of particles per unit volume of gas enter the
probe. This is known as isokinetic sampling.
Sampling of ambient air masses is seldom made isokinetically as
sophisticated equipment is required to maintain the inlet facing into the
wind and to adjust the sampling velocity to match changes in wind speed.
This is feasible, but what is difficult is then to interpret the analysis of the

collected sample as the flow rate is temporally variable. However,
isokinetic sampling is usually practicable, and indeed necessary, when
sampling flue gases.
3.1.2 Sample Collection. The methods most commonly used for the
collection of atmospheric particulate samples are:
—
—
—
—

filtration;
impingement: wet or dry impingers, cascade impactors;
sedimentation: by gravity in stagnant air, thermal precipitators;
centrifugal force, cyclones;

and for gaseous samples:
—
—
—
—

adsorption;
absorption;
condensation;
grab sampling.

Filtration. This is by far the most common technique. The type of filter
medium chosen will depend upon a number of factors. These include the
collection efficiency for a given particle size,18 pressure drop and flow
characteristics of the filter type,19 background concentrations of trace
constituents within the filter medium, and the chemical and physical
suitability of the filter with regard to the sampling environment.
Impingement. Impingers consist of a small jet through which the air
stream is forced, so increasing the velocity and momentum of suspended
particles, followed by an obstructing surface on which the particles will
tend to collect. Wet impingers operate with the jet and collection surface
under liquid and require high flow rates for optimum collection
efficiency.
Cascade impactors use the aerodynamic impaction properties of
particles to separate the sample into different size fractions by use of
sequential jets and collection surfaces. Increasing jet velocity and/or
decreasing gaps between the jet and collection plate fractionates the
sample. Figure 8 shows the principle of a commonly used cascade
impactor. This consists of up to 15 stages backed by a membrane filter,
each stage containing accurately drilled holes which align over a solid
portion of the adjacent plates. The holes in each successive stage are
18

M. Katz, 'Measurement of Air Pollutants, Guide to the Selection of Methods', World Health
Organization, Geneva, 1969.
B. Y. H. Liu, D. Y. H. Pui, and K. L. Rubow, 'Aerosol in the Mining and Industrial Work
Environments', ed. V. A. Marple and B. Y. H. Liu, Ann Arbor Science, Ann Arbor, MI, 1983,
Vol. 3, p. 898.

Air flow

Impactor jet
Aerosol deposit
Impaction surface
1st stage

Backing filter
To pump
Figure 8

Schematic representation of a cascade impact or

smaller than those in the preceding plate, and since air is drawn through
the instrument at a constant flow rate the effective velocity at each stage
increases. The largest particles are impacted on the first stage and the
smallest are collected on the backing filter. The range of particle
diameters collected on each stage may be determined by laboratory
calibration or by theoretical calculations. However impaction sampling
at normal flow rates (about 0.01-0.04 m3 min" 1 or 0.6-1.0 m3 min" 1
for Hi-VoI cascade impactors) is only efficient for particles with
aerodynamic diameters > 0.3 mm. Also the collection efficiency of
each stage will vary according to particle type, some being very
'sticky', others liable to bounce off. Other problems of cascade impactor
sampling include wall losses, the aggregation of particles, and the
mechanical breaking of agglomerates which result in inaccurate size
distribution measurements.
Sedimentation. The collection of particulate material by allowing it to
deposit into a collection vessel is the simplest of all air pollution
measurement techniques. However, the presence of the bowl or cylinder
in the path of the falling particles will change their flow pattern and it is
not clear whether the collected material is truly representative of actual
conditions. The methods are not as widely used now as previously. The
British Standard deposit gauge consists of a collection bowl connected to
a bottle and supported by a galvanized steel stand. During wet weather
dust is washed down from the bowl, but during dry weather high winds
may blow dust out of or into the bowl, so producing erroneous dust
loadings. At the end of the sampling period (usually one month) a
measured volume of water is used to wash any dust in the bowl into the
collection bottle, and the pH, total particulate mass, and water volume
determined.

0.75 m
approx.

1.5 m
approx.
Removable
bottles

Figure 9

Directional deposition gauge

A major disadvantage of the standard deposit gauge is that no
directional resolution of the source of particulate material is possible.
The directional deposit gauge20 consists of four cylinders mounted on a
common post, with open slots facing the four quadrants of the compass
(Figure 9). Each cylinder has a removable collection bottle at its base.
After collection a suspension of the dust may be placed in a glass cell and
a measure of the dust loading made by the amount of obscuration of a
beam of light passing through the cell. Density fractionation of the
collected material is also possible by using a mixture of diiodomethane
and acetone which allows a density gradient of 0.8-3.3 g cm~~3 to be
achieved. By comparison of the density gradient fractionation of
20

'British Standard Method for the Measurement of Air Pollution, Part 5—Directional Deposit
Gauges BS 1747 part 5', British Standards Institution, London, 1972.

material collected in a directional deposit gauge with material collected
from likely sources (e.g. pulverized fly ash storage heaps, slag heaps,
cement works, etc) simple source apportionment studies may be
attempted. Other techniques, including dispersion staining and microscopic examination may also be used in this way, whilst the compilation
of a reference library of dusts from known sources in a given area will
greatly ease the practical difficulties of dust identification.
The siting of deposit gauges often presents problems. When attempting to monitor emissions from one major source the use of several
directional gauges around the source may be successful in confirming
that emissions occur from that source. However in areas of multiple
sources or where there is significant atmospheric turbulence (e.g. in builtup areas) inconclusive data may be obtained. The basic requirement that
gauges should be sited at a distance from any object of at least twice the
height of the object is often insufficient to avoid significant distortion of
the deposition pattern. Siting is further complicated by the need to have
tamper-proof sites at or near ground level. It is worth pointing out that
the standard deposit gauge is most effective for collection of large
particles which readily settle under gravity, whilst the vertical slots of
the directional gauge collect smaller particles impacted by the wind more
effectively. Source apportionment studies using deposit gauges have now
largely been superseded by the specific chemical analysis ('finger printing') of actively sampled particulate material.
Adsorption. The adsorption of gases is a surface phenomenon. Gas
molecules become bound by intermolecular attraction to the surface of a
collection phase and so become concentrated. Under equilibrium conditions at constant temperature the volume of gas adsorbed on the
collection phase is proportional to a positive power of the partial
pressure of the gas, and is also dependent upon the relative surface area
of the adsorbent. Materials commonly used as adsorbents include
activated carbon, silica gel, alumina, and various porous polymers.
When selecting a suitable adsorbent the relative affinity for polar or
non-polar compounds must be considered. For example, activated
carbon is non-polar and therefore will adsorb non-polar organic gases,
but exclude polar compounds such as water vapour. The wide range of
gas chromatographic supports available vary in their degree of polarity
and so allow selection of the appropriate type.
The adsorbent used must not react chemically with the collected
sample unless chemisorption is used intentionally. Also the analytes
must not react with other constituents of the sampled air, either during
collection or storage. It has been found for example that some hydrocarbons may decompose on polymers by reaction with atmospheric

ozone. This may be prevented by the use of a selective prefilter which
removes the oxidant from the air stream but allows the analytes to pass
through.
It is important to determine the retention volume of the adsorbent (i.e.
the volume of air which may be passed without breakthrough of the
analyte) with respect to the species being collected. This should be high
enough to allow sufficient of the analytes to be collected for analysis. The
desorption properties of the material are also important to ensure
quantitative recovery of the sample, preferably with regeneration of the
adsorbent for subsequent use. Activated carbon is a very efficient
adsorber, so making quantitative desorption difficult. Steam stripping
may result in hydrolysis reactions with the analytes and vacuum distillation and solvent extraction are not without their problems. Compounds
on support bonded porous polymers may conveniently be thermally
desorbed by flushing with an inert carrier gas. In the case of reactive
hydrocarbons on polymer a two-stage thermal desorption system utilizing an intermediate cryogenic trap cooled with liquid N 2 followed by
flash heating allows quantitative recovery as well as direct injection of the
sample in a very small volume of gas onto the gas chromatograph
column.
Since adsorption is temperature dependent, collection efficiency and
an increase in retention volume may be achieved by cooling the
adsorbent. However, problems with blockages by ice may then occur.
With the increase in sophistication of detection systems in recent years,
particularly by the interfacing of chromatographic separation techniques
with mass selective detectors, the use of adsorbents as preconcentrators
is also increasing. However, care must always be exercised to avoid nonquantitative collection, breakthrough effects due to exceeding the retention volume of the system, decomposition, and non-quantitative recovery of the sample.
Absorption. Gases may be collected by being dissolved in a liquid
collection phase or by chemical reaction with the absorbent. The simple
Dreschel bottle may be used or may be modified by the inclusion of a
fitted diffuser to create small bubbles and so enhance the collection
efficiency.
An example of a simple absorption technique which allows an estimate
of NO2 concentration to be made with relatively little capital outlay is
the use of triethanolamine diffusion tubes.21 A small acrylic tube is fitted
with a fixed cap at one end. A fine wire mesh coated in triethanolamine is
placed in the closed end of the tube and absorbs NO2 as it diffuses from
the open end. The NO2 is determined spectrophotometrically at the end
21

M. R. Heal and J. N. Cape, Atmos. Environ., 1997, 31, 1911-1923.

of the sampling period. These passive samplers are very cheap to
construct and analyse and have been used as an effective primary survey
technique before embarking upon a more expensive monitoring exercise
based upon the standard chemiluminescent technique.
Condensation. By cooling an air stream to temperatures below the
boiling point of the substance of interest it is possible to condense gases
from the air and so concentrate them. However, a limitation of the
method is that water vapour present in the air will also freeze and so
progressively block the trap. This may be overcome by using a first trap
of large volume designed to collect water and a second trap at a
sufficiently low temperature to collect the analytes. Coolants of temperatures — 183 0C or lower {e.g. liquid N2) should not be used for this
purpose as they will condense atmospheric oxygen and result in a serious
combustion hazard.
Grab sampling. Rather than utilizing a concentration technique in the
field, samples may be collected in an impermeable container and
returned to the laboratory for analysis. Grab samples of this type have
been collected in FEP-Teflon bags or specially treated stainless steel cans
for hydrocarbon determination by GC, for example. In the bag technique the deflated container is housed within a rigid box which is slowly
evacuated. Air is thus drawn into the flexible bag which may be sealed
when inflated. Samples can then be drawn at a later stage from the bag by
hypodermic gastight syringe. Pumps constructed of inert material may
be used to fill rigid cans to a high pressure, allowing a large volume of air
to be sampled.
Specific techniques. Two air sampling techniques are in such widespread use that they require separate consideration. These are:
(a) the UK National Survey of Air Pollution smoke and SO2
sampling apparatus,
(b) the US National Air Sampling Network Hi-VoI method for total
suspended particles (TSP) measurement.
(a) NSAP smoke and SO2 method. The equipment used in the
National Survey of Air Pollution in the UK consists of a pump which
draws about 2 m3 of air per day through a filter paper held in a brass
clamp. This removes particulate material. The air then passes through
dilute H2O2 which removes SO2, converting it to H2SO4, which is
determined by titration. An eight-port sample changer allows eight 24 h
samples to be collected sequentially in an array of eight filter papers and
bottles, and so requires operator attendance only once a week. Although

the H2O2 method will actually measure the net gaseous acidity of the air
it is considered that for ordinary urban situations it will give a good
estimate of the SO2 concentration and hence results are expressed in
terms of fig SO2 m~3. The particulate loading of the exposed filter paper
is estimated by measuring the reflectance of the paper, so obtaining a
measure of the staining property of the air. A calibration curve is then
used to convert the darkness of the stain to concentrations of equivalent
standard smoke.
Inaccuracies in these methods may become pronounced when there
are acidic or basic gaseous components other than SO2 present in the air
(HCl giving a positive interference and ammonia a negative interference)
or when particulates of 'non-standard' staining properties (e.g. light
coloured ammonium compounds) are present. When the filter paper is
extremely heavily loaded the reflectometer method may severely underestimate the actual smoke level. Particulate losses may also occur at the
inlet to the apparatus and in the inlet tube. The smoke stain method gives
concentrations of Standard Smoke which are not directly comparable
with TSP levels determined by the US Hi-VoI method.
Although still widely used, the titration method has been superseded
by the preferred technique of gas phase pulsed fluorescence, which
allows the determination of SO2 concentrations with one or two
minute averaging times with very high precision and accuracy at low
concentrations. Similarly, the smoke stain method has now been
superseded by the state-of-the-art oscillating microbalance method for
determination of particulate loadings in ambient air. Both these
sophisticated methods require substantial capital investment and
demand skilled operator attendance for calibration and maintenance
purposes.
(b) Hi- VoI method for suspended particulates. This method is the
current US EPA reference method for total suspended particulates
(TSP) and is used in the US National Air Sampling Network,22 although
it is being superseded by the use of size selective inlets which excludes
particles of aerodynamic diameter > 10 fim (the PM10 instrument) or
>2.5 jiva (PM2 5). A high flow rate blower draws the air sample into a
covered housing and through a 20 x 25 cm rectangular glass fibre filter
at 1.1-1.7 m3 min" 1 . The mass of particles collected on the filter is
determined gravimetrically and extraction techniques may then be used
to remove the material for chemical analysis. However, when glass fibre
filters are used reaction with acidic components may result in artifact
formation, for example sulfate from gaseous SO2. Although 24 h
22

Environmental Protection Agency, 'Reference Method for the Determination of Suspended
Particulates in the Atmosphere (High Volume Method)', US Federal Register 36, No. 84, 1971.

sampling periods are commonly used, timer devices are available to
switch on and off the blower at predetermined intervals. However,
passive sampling by the settlement of particulates onto the filter during
periods when the pump is not operating may cause a positive error in the
determination. Recent modifications include size selective inlets which
will exclude particles of greater than a given size. However, the cutoff
efficiency is usually dependent upon wind speed and may not be
sufficiently selective.
3.1.3 Flow Measurement and Air Moving Devices. In order to measure
the concentration of an airborne constituent it is necessary to know the
volume of air sampled. This may be achieved by measuring the rate of
flow with a rate meter or by directly measuring the volume of air passed
with a dry or wet test gas meter, a cycloid gas meter, or by the use of a
mass flow controller. All these devices require calibration and regular
checking for leaks. As air volume is dependent upon both temperature
and pressure the measurement of these two parameters at the meter inlet
is essential so that volumes may be expressed at standard temperature
and pressure.
Many different types of pump are available for air sampling, both
mains and battery operated, but the precise type chosen will depend
upon the required flow rate, the availability of power, and whether
continuous or intermittent flow is required.
3.2

Water Sampling Methods

For many applications no special water sampling system is required as an
appropriate sample container immersed in the water may be adequate.
The main requirement is that a portion of the material under investigation small enough in volume to be transported and handled conveniently
but still accurately representing the bulk material should be collected.
Typically a 0.5-2 dm3 volume is sufficient. When samples are required
from depth, two types of collection vessel may be used.9 The first consists
of a cylinder with hinged lids at both ends. The container is lowered into
the water with both lids opened and at the desired depth a messenger
weight sent down the wire which closes them. This type is not suitable for
trace metal work as contaminated surface waters may result in contamination of the vessel and the messenger may scour metallic particles from
the wire. The second type consists of a sealed container filled with air
which is lowered to the required depth. A messenger is again sent down
to open, and another to close, the lid. Alternatively, pressure sensors may
activate the lid.
Automatic sequential samplers are available which will collect a given

volume of water into an array of bottles. They have been used, for
example, in collecting stormwater runoff from roads and the sampling
sequence may be triggered when the flow in a flume reaches a certain
height.
Adsorption or filtration media may also be used to concentrate the
species of interest in situ. Using a completely self-contained sealed unit of
inert material housing a peristaltic pump and power supply with only the
adsorption tube inlet and outlet open to the water, contamination of the
sample can be completely avoided. Very large volumes of water may be
processed. Another ongoing development is the use of passive permeation devices which may be immersed in water for long periods of time, so
giving time-averaged concentrations.
When considering sampling methods for use on inland waters or in
coastal waters and estuaries the sophisticated techniques developed for
use at sea may be found to be impracticable due to their need for heavy
lifting gear on the sampling vessel. Monitoring work must often be
undertaken on such waters using small boats without such equipment.
One ingenious method of collecting water samples at different depths
using very limited resources on a small boat is to lower a weighted plastic
tube to the desired depth and to use a small peristaltic pump to draw
water up and into a collection bottle. In this way completely uncontaminated samples may easily be obtained from depths of 30 m or more.
Whether the particulate fraction of material present in the water can be
quantitatively collected in this manner is not clear.
Whichever method of collecting samples is used care must be taken to
ensure that neither the sample storage containers nor any collecting
vessels used contaminate or alter the sample. Contamination may occur
by:
(a) Leaching of contaminants from the surface of imperfectly cleaned
containers.
(b) Leaching of organic substances from plastics, or silica and sodium
or other metals from glass.
(c) Adsorption of trace metals onto glass surfaces or organics onto
plastic surfaces. In the case of metals this may be avoided by prior
acidification of the container, but this may in turn exacerbate
problem (a).
(d) Reaction of the sample with the container material, e.g. fluoride
may react with glass.
(e) Change in equilibrium between pollutants in particulate and
solution phases.
If a solvent extraction technique is used to concentrate the analytes prior

to analysis, care must be taken to ensure that the reagents and containers
used are themselves sufficiently clean. Some commonly used materials
and techniques have been shown to cause severely elevated metal levels in
water.
Different determinands require different methods of preservation in
order to prevent significant changes between the time of sampling and of
analysis. Generally acidification to pH 2 and refrigeration to 4 0C will be
adequate although complete stability of every constituent can never be
guaranteed and, at best, chemical, physical, and biological processes
affecting the sample can only be slowed down. Samples may be filtered
directly after collection in the field to separate the particulate and
solution phases. The solution phase may then be acidified to prevent
adsorption of the pollutant to the container wall.
3.3

Soil and Sediment Sampling Methods

Soils and sediments are typically very inhomogeneous media and large
lateral and vertical variations in texture, bulk composition, water
content, and pollutant content may be expected. For this reason large
numbers of samples may be required to characterize a relatively small
area. Although surface scrapings may be taken it is often necessary to
obtain cores so that vertical profiles of the determinands may be
obtained or cumulative deposition estimated. Plastic or chromium
plated steel tubing of ~ 2.5 cm internal diameter is often suitable, and if
the samples are sealed into the tubes and air excluded they may be
satisfactorily stored at low temperatures until required. Otherwise they
may be extruded in the field and stored in plastic bags. Various core
sampling devices are available for obtaining cores of bottom sediments
from lakes etc (e.g. the Jenkin corer).
Grab samples of soils are easily obtained manually and stored in precleaned plastic bags. Sometimes composite samples formed by the
bulking together of a number of individual samples may be sufficient,
but generally analyses of individual samples is to be preferred. In the case
of sediments, grab samplers are available for operation at considerable
depths, examples being the Ponar, Orange-peel and Peterson grabs.
Alternatively a dredge may be used to obtain a composite sample along
a strip of the sediment surface.
Wet soils and sediments that are to be analysed while still wet should
not be collected or stored in bags, but in rigid containers. The vessel
should be filled as completely as possible leaving no airspace at the top
and a bung inserted so as to displace excess water without admitting air.
Some determinands in soils and sediments are liable to change during
storage and require the use of preservation techniques. For example

Retain opposite
quadrants

Figure 10

Schematic diagram of the sub-sampling of dried soil or sediment using the
technique of coning and quartering

nitrate in soil can be extracted into potassium chloride solution and
preserved with toluene. Usually, however, air-dried soils and sediments
may be disaggregated, sub-sampled by coning and quartering (see Figure
10) and stored in suitable containers, but as always sample contamination must be avoided at each stage.
There are important effects associated with grain size which should be
considered in the analysis of soils or sediments. First, many pollutants
are associated with particle surfaces and therefore occur in highest
concentrations in the smaller grain sized material. Secondly sub-sampling from a bulk sample may be very difficult due to size segregation
effects and it may be necessary to grind the sample to a very fine powder
to ensure homogeneity prior to division of the sample.
4 MODELLING OF ENVIRONMENTAL DISPERSION
A characteristic feature of environmental monitoring studies is that
substances may be found over very large ranges of concentrations, and
therefore the analytical techniques employed must be extremely flexible.
Some typical concentrations of substances in polluted environmental
media are given in Table 2. Not only will large differences of concentrations be found from area to area but even small temporal and lateral
changes can result in large changes in pollutant concentrations. The
temporal and spatial variability of lead in air and dust samples is
illustrative of the way in which pollutant concentrations vary and have
been discussed in relation to the design of monitoring programmes.23 In
this study it was shown by collating data collected at several sites in
London, UK, that short term concentrations of atmospheric lead (of
23

M . J. Duggan, Sci. Total Emir on., 1984, 33, 37-48.

Table 2

Typical concentrations of substances in polluted environmental media

Pollutant

Medium

Typical ranges

Cadmium
Lead
Lead
Lead
Sulfur dioxide
Sulfate
Benzo(#)pyrene
Carbon monoxide

air
seawater
soil
air
air
air
freshwater
air

0.1-10 ngm~ 3
2-200 pmol dm ~3
5-5000 mg kg ~ l
1-100 ngm" 3
0.5-200 ppb (IO"9 v/v)
0.1-20/igm~3
0.1-10 ng dm" 3
0.01-50 ppm (10~6 v/v)

sampling period less than one week) can vary considerably at the same
site, with a factor of up to 4 between the highest and lowest weekly
averages at any one site. Traffic flow differences cannot entirely account
for these variations, and local weather conditions are probably the
dominant factor. However, when longer averaging periods are used the
temporal variability at any one site is much reduced, but a second effect
becomes apparent. Thus monitoring periods of at least a month or two
are required in order to obtain representative results with possibly some
adjustment made for the season. The spatial variability of urban lead-inair concentrations was found in this study to be rather slight and
although lead concentrations decreased away from the carriageway
small changes in sampling position (e.g. at the ground floor as opposed
to the third floor windows of a building facing the street) made up to
30% difference to lead concentration. Thus the precise choice of
sampling position at a site may not be critical unless the concentrations
are very close to some pre-determined guideline or limit value, in which
case compliance may be found in one position but not at another close
by.
Dust samples collected weekly at the same pavement sites were found
to have lead concentrations which varied by about the same relative
amount as air-lead concentrations (i.e. weekly samples had a coefficient
of variation of ca. 30%). However, spatial variability of lead-in-dust
concentrations was found to be high over short distances (i.e. a few
metres). It was suggested therefore that this can be overcome by taking
dust samples over a large area (e.g. 5 m2) and that in this way
representative dust-lead values may be obtained.
In order to appreciate the variability of pollutant levels, and hence
appreciate the complexity of designing an adequate monitoring programme, it is necessary to have some understanding of environmental
dispersal, mixing, and sink processes and of the time-scales on which
these processes act. Rather than discussing in detail the physical,
chemical, and biological processes responsible for changes in pollutant

concentrations after discharge to the environment, which have been
extensively presented elsewhere, the salient features of some of the
techniques by which these changes can be anticipated will be shown. A
summary of some of the computer codes which are currently available
for modelling environmental dispersion is provided in Table 3.
4.1

Atmospheric Dispersal

Material discharged into the atmosphere is carried along by the wind and
mixed into the surrounding air by turbulent diffusion. In the vertical
plane the dispersion continues until the turbulent boundary layer is
uniformly filled whilst in the horizontal plane dispersion is theoretically
unlimited and usually proceeds more rapidly than in the vertical plane.
In the simplest models of plume dispersion the degree of turbulence, and
hence of mixing, is described by an atmospheric stability classification
that is dependent upon the amount of incoming solar radiation, wind
speed, and cloud cover, but surface roughness is also important in
producing turbulence, especially in the case of large buildings or topographic features.
The most commonly used model of plume dispersion24 is that
described by a Gaussian distribution characterized by standard deviations Oy and az in the vertical and horizontal directions respectively.
The basic equation for Gaussian plume dispersion is:

where x is the pollutant concentration at point (x9y,z) (/xg m~3); Q is the
pollutant emission rate (fig s" 1 ); t/(z) is the wind speed (m s"1); at the
effective emission height, He; oz is the standard deviation of the plume
concentration in the vertical at distance x; ay is the standard deviation of
the plume concentration in the horizontal at distance x; x9y9z are the
lateral, transverse, and vertical directions (m), downwind with the base
of the stack as the co-ordinate origin; HQ is the effective height of the
plume (m).
Assumptions made are that:
— no deposition occurs from the plume at the ground surface;
— pollutant levels are not altered by chemical processes in the plume;
24

F . Pasquill, 'Atmospheric Diffusion', Ellis Horwood, Chichester, 1974.

Table 3

Examples of environmental dispersion modelling tools

Environmental
media
Code

Overview

Availability

Air

ADMS

ADMS is a PC-based model
for atmospheric dispersion of
passive, buoyant or slightly
dense releases from single or
multiple sources

Cambridge Environmental
Research Consultants, 3,
Kings's Parade, Cambridge,
CB2 ISJ, UK

Freshwater

HSPF

HSPF is a comprehensive
modelling package for the
simulation of the quantity
and quality of runoff from
multiple-use catchments, and
of processes occurring in
streams of fully-mixed lakes
receiving catchment runoff.

US Environmental
Protection Agency, Centre
for Exposure Assessment
Modeling, Athens, Georgia,
United States

OTTER

OTTER is a model for the
AEA Technology, Thomson
transport of heavy metals and House, Risley, Warrington
radionuclides through the
WA3 6AT, United Kingdom
freshwater environment,
typically for simulating the
after effects of deposition
from the atmosphere.

PRAIRIE

PRAIRIE is a risk assessment AEA Technology, Thomson
software tool for predicting
House, Risley, Warrington
the risks associated with
WA3 6AT, United Kingdom
accidental releases of
hazardous materials into
rivers or estuaries.

QUAL2E

QUAL2E is a stream water
quality model designed
primarily to simulate
conventional constituents
{e.g. nutrients, algae,
dissolved oxygen) under
steady-state conditions, both
with respect to flow and input
waste loads.

US Environmental
Protection Agency, Centre
for Exposure Assessment
Modeling, Athens, Georgia,
United States

Multi

MMSOILS

The MMSOILS model is a
methodology for estimating
the human exposure and
health risk associated with
releases of contamination
from hazardous waste sites.

US Environmental
Protection Agency, Centre
for Exposure Assessment
Modeling, Athens, Georgia,
United States

Multi

MULTIMED The Multimedia Exposure
Assessment Model
(MULTIMED) simulates the
movement of contaminants
leaching from a waste
disposal facility or
contaminated soils

US Environmental
Protection Agency, Centre
for Exposure Assessment
Modeling, Athens, Georgia,
United States

— there is no effect from surface obstructions (e.g. buildings);
— dispersion by diffusion in the downwind direction is negligible
compared with bulk transport by the wind;
— the constituents are normally distributed vertically and horizontally across the plume.
The effective height of emission He (which is greater than the stack
height due to the momentum and buoyancy of the plume) may be
calculated from Holland's equation:25

where Vs = stack gas exit velocity (m s"1)
d = internal stack diameter (m)
u = wind speed (m s~l)
p = atmospheric pressure (mb)
T8 = stack gas temperature (K)
Ta = air temperature (K)
H = stack height (m)
and 2.68 x 10~3 is a constant with units of mb" 1 m" 1 .
The wind speed at height z (uz) may be obtained from the power law

where W10 is the speed at the reference height of 10 m, and n varies
according to surface roughness, but may be taken to equal 0.2.
The next step is to determine the appropriate Pasquill stability class
from Table 4 and then evaluate the estimate of oy and oz as a function of
downwind distance.24 In order to obtain the ground level concentration
below the plume centre line (i.e. y = z = 0) the general equation reduces
to:

or for a ground level source with no effective plume rise (He = 0)

J. Z. Holland, 'A Meteorological Survey of the Oak Ridge Area, Report ORO-99', Atomic Energy
Commission, Washington, DC, 1953, p. 554.

Table 4

Pasquill's Stability Categories
Day

Night

Surface wind speed
(at 10 m)
(ms" 1 )

Incoming solar radiation
strong

moderate

slight

Thinly overcast
or ^4/8
low cloud

<3/8
cloud

<2
2-3
3-5
5-6
>6

A
A-B
B
C
C

A-B
B
B-C
C-D
D

B
C
C
D
D

E
D
D
D

F
E
D
D

After Ref. 24.
The neutral class D should be assumed for overcast conditions during day or night.

Full treatments of these equations and their applications are available
elsewhere, including modifications to include line and area sources.26'27
A frequently required type of calculation is to determine the minimum
height of a chimney which will give adequate dispersion to ensure that
critical ground level concentrations are not exceeded. A simple algorithm
has been developed which allows this estimate to be easily made.28
All attempts to model atmospheric diffusion and mixing processes are
liable to be, at best, only good estimates of the real situation and care
must be taken to:
(a) understand the physical, chemical, and mathematical limitations
of the model, and
(b) avoid treating the output from models as providing definite
answers.
Although accurate measurements may be made of wind speed it is in the
determination of the turbulence characteristics of the atmosphere that
uncertainties arise, which in turn lead to uncertainties in the model. At
short distances downwind, with steady winds, uniform terrain, and no
local obstructions the Gaussian plume model may be expected to give
good estimates, but in any but ideal conditions they soon become orderof-magnitude estimates only.
26

D . B. Turner, ' W o r k b o o k of Atmospheric Dispersion Estimates', U S Environmental Protection
Agency, Research Triangle Park, N C , 1970.
27
R. H . Clarke, ' A Model for Short a n d M e d i u m Range Dispersion of Radionuclides Related to the
Atmosphere', Report N R P B - R 9 1 , National Radiological Protection Board, Harwell, 1979.
28
H M Inspectorate of Pollution, 'Technical Guidance N o t e (Dispersion) D l . Guidelines o n
Discharge Stack Heights for Polluting Emissions', H M S O , L o n d o n , 1993.

4.2

Aquatic Mixing

The physical transfer and transport of pollutants in the aquatic environment is determined by the same two processes that determine the mixing
of pollutants in the atmosphere. These are:
(a) advection, caused by the large-scale movement of water, and
(b) mixing or diffusion, due to small-scale random movements which
give rise to a local exchange of the pollutant without causing any
net transport of water.
The combined effect of advection and diffusion is known as dispersion.
These two processes occur over a very wide range of scales, both of
spatial extent and frequency, which necessitates the use of averaging
procedures when defining their role in pollutant transfer. As in the lower
atmosphere, there is usually a constraint on the vertical dispersion
component, induced either by water depth, or in deeper waters by
thermal or density stratification. On a large scale in the oceans there
will be a vertical circulation driven by density differences but, on a small
scale, vertical motion due to turbulence or eddy diffusion will tend to be
suppressed by stratification.29 Similar restrictions on vertical mixing
occur in rivers due to limited depth, but here horizontal mixing in the
cross-channel direction is also constrained.
A large number of models have been developed to describe the
movement of pollutants in the aquatic environment, many being analogous to those used in air pollution studies. The fact that such models are
necessary is due to limitations in our detailed knowledge of the velocity
field, and this in turn may lead to uncertainties in the prediction of
dispersion patterns.
The dispersion of pollutants in rivers has attracted a great deal of
study, particularly in the context of effluent discharges and the ability of
rivers to dilute them. Unlike the oceans, which have traditionally, but
unreasonably, been considered to have an infinite capacity for dilution,
the deterioration in water quality due to pollutant discharges is often
manifestly apparent in rivers.
Various models have been applied to the problem of modelling river
dispersion with varying levels of complexity. An extremely simple model
of change in water quality downstream from a discharge into a river
would be to assume that an exponential decay in concentration occurs,
i.e.
29

G. Kullenberg, 'Physical Processes in Pollutant Transfer and Transport in the Sea', ed. G.
Kullenberg, CRC Press, Boca Raton, 1982, Vol. 1.

Concentration
ADZ Model
Concentration (micro curies/cubic ft.)

Fischer Model

Time (2 min. intervals)
Figure 11 Modelling of Copper Creek dispersion data: comparison of ADZ model fit with
the results obtained by Fischer using the conventional Fichian diffusion model
(Reproduced with permission from 'Pollution Causes, Effects and Control',
ed. R. M. Harrison, Royal Society of Chemistry, Cambridge, 2nd Edn., 1990)

where Cx is the concentration at point x, Co is the concentration at the
point of discharge, k is the decay rate and / is the time taken for flow from
the point of discharge to point x.
An alternative approach to river modelling is to consider the river to
be divided into a number of reaches, with each reach being considered as
a continuously stirred tank reactor, and to place an appropriate time
delay between each reach. It is assumed that the major dispersive
mechanism can be explained by the aggregated effect of all 'dead zone'
phenomena in the river between the two sampling points. This 'aggregated dead zone' model is thus a combination of the continuously stirred
tank reactor and a factor to account for the advection component of
dispersion. The two approaches to modelling river dispersion represented by the Fickian Diffusion analysis and the Aggregated Dead Zone
(ADZ) model have been compared with data obtained from a tracer
experiment.30 Figure 11 compares the monitored concentrations with
those predicted by the two models, and in this example the ADZ model is
able to explain the data better than the conventional diffusion model.
30

P. C. Young, 'Pollution: Causes, Effects and Control', ed. R. M. Harrison, Royal Society of
Chemistry, Cambridge, 2nd Edn., 1990.

More recent models31 take account of the various physico-chemical
processes a pollutant may be subjected to following a spill into a river
(e.g. sorption, volatilization, hydrolysis, photolysis, oxidation, and
biological degradation). The effects of weirs on pollutant removal may
also be modelled.
4.3

Variability in Soil and Sediment Pollutant Levels

Obviously, the same mechanisms of dispersion as operate in fluid or
gaseous media do no occur in soils and sediments. Physical mixing may
occur, such as during agricultural practices, dredging of estuaries, or
bioturbation by burrowing organisms, but usually only on a fairly
limited scale. The level of contamination in a soil or sediment will
depend upon the deposition rate of the pollutant and its subsequent
rate of movement through the soil or sediment column. The rate of
movement of a contaminant through these solid media is dictated by the
degree of adsorption to or leaching from the particles and the flux rate of
pore water, transferring the pollutant to deeper horizons. The physicochemical properties of the water, soil or sediment particles and the
pollutant all influence the rate of adsorption or leaching. For example,
conditions that favour the adsorption of radiocaesium and lead in soils
include low rainfall and high clay (particularly illite) content, whereas
mercury and copper tend to accumulate in soils with a high organic
content. Contaminant concentrations are generally always higher in soils
or sediments with finer grain sizes, due to the increased total surface area
available for adsorption. These effects are illustrated in Figure 12 for
trace metal concentrations in sewage sludge in treated and control soil
plots. Copper concentrations in sludge-treated plots show a strong
correlation with organic carbon content, while zinc concentrations in
both treated and control plots increase with decreasing particle size, and
increase with increasing organic carbon content. All these factors lead to
great variability in pollutant concentrations, which simply emphasises
the need for carefully designed monitoring programmes.
5

DURATION AND EXTENT OF SURVEY

5.1

Duration of Survey and Frequency of Sampling

The duration of a pollution monitoring programme is entirely dependent
upon the purpose of the study, and can vary from the time taken to
collect a limited number of, for example, street dust samples in an urban
31

A. Watson, L. S. Fryer, and J. W. Clark, 'Aqueous Pollution Modelling—Approaches used by
PRAIRIE™, AEAT-0843', AEA Technology, Harwell, 1990.

Sludge Treated Plots

Copper
Zinc
Organic carbon

Organic carbon concentration (g/kg)

Trace metal concentration (mg/kg)

(a)

Particle size fraction (^m)
Control Plots
Copper
Zinc
Organic carbon

Particle size fraction (^m)
Figure 12 The distribution of copper and zinc concentrations in (a.) sludge-treated soil
plots and (b) control soil plots
(After J. Ducaroir and I. Lamy, Analyst, 1995, 120, 741-745)

Organic carbon concentration (g/kg)

Trace metal concentration (mg/kg)

(b)

area to tens of years for long term surveillance projects such as the US
EPA Ambient Air Monitoring Program. The choice of the frequency of
sampling, i.e. the duration of each sample period and the interval
between successive measurements, is also dependent upon the objectives
of the study. Pollutant concentrations in air and water fluctuate with
varying degrees of rapidity and in order to characterize their behaviour it
is necessary to measure these changing levels; long term mean data may
be sufficient for some purposes but will not be adequate where information of short term high level episodes is required. Generally, if random
sampling techniques are used, the number of samples required will
increase as the standard geometric deviation of the pollutant concentrations increases, i.e. the greater the fluctuation of the pollutant level, the
more numerous the samples that must be taken to assess the variations
accurately. If the variations in levels during the period of interest are
essentially random, independent, and normally distributed then the
number of samples which must be taken in order to estimate the period
mean within certain limits and prescribed confidence limits may be
calculated. However, in order to do this a reasonable estimate of the
standard deviation of the data is required: it is assumed that the random
variations follow a normal distribution and that the results of successive
samples are not serially correlated but are independent. These criteria are
rarely met.
As a more general guideline it may be assumed to be necessary to have
a sampling interval at least ten times shorter than the fluctuation cycle
time. For example, the entire variance structure of a diurnally fluctuating
pollutant concentration profile may be obtained from roughly 12
samples, each with a two-hour averaging period. If the annual trend of
levels is required then probably 12 monthly samples each year would be
adequate. Thus, although a continuous and instantaneous record of
pollutant level may be required for some purposes (e.g. to monitor very
short-term changes in air quality due to the oscillating passage of a
plume over the sampling station), it is not always necessary. Further, if a
short sampling period is chosen it rapidly becomes necessary to record
and store efficiently the large volumes of data generated, which will
themselves often be averaged over a longer period for analysis. A case in
point is the UK automatic air pollution monitoring network where one
minute average readings are themselves averaged to give one-hourly
values.
Fast-response continuous monitors are now available for the more
common gaseous pollutants and for many determinations of water
quality. The term 'fast-response' implies a response time, measured as a
90% rise, of less than about 2 min, but in most cases it is of the order of
seconds. Thus very rapid temporal variations are measurable and with

Table 5

Summary of commonly employed methods for measurement of gaseous
air pollutants

Pollutant

Measurement technique

Response
Sample time*
collection (continuous Minimum
techniques) concentrations
period
5s

Total
NDIR
hydrocarbons
Flame ionization analyser

0.5 s

b
GC-FID
Specific
hydrocarbons
NDIR
Carbon
monoxide
Catalytic methanation/FID
Absorption in hydrogen
Sulfur dioxide
24 h
peroxide/titration
Fluorescence analyser

0.5 ppm
lOppb
2ppb

2 min

0.5 ppb
5ppb

0.5 ppb

10 sec

1 ppb
lppb

Oxides of
nitrogen

Ozone
Peroxyacetyl
nitrate

Conversion to nitrite/azo
days
dye formation
Chemiluminescent reaction
with ozone
UV absorption
GC/electron capture
detection
C

1 ppm
(as hexane)
lOppb
(as methane)
lppb

Time taken for a 90 per cent response to an instantaneous concentration change.
Grab samples of air collected in an inert container and concentrated prior to analysis.
c
Instantaneous concentrations measured on a cyclic basis by flushing the contents of a sample loop
into the instrument.
b

the advent of on-line microprocessor data handling and reduction
systems the large amounts of data produced are more easily handled.
Fast-response continuous monitors do not generally have a predictable
response so they cannot be calibrated solely in terms of chemical
stoichiometry and hence need calibration with a standard atmosphere
or solution. Some of the commonly employed fast-response methods of
gaseous air pollutant analysis are shown in Table 5, and some of the
methods used for water analysis are shown in Table 6.
One further consideration when deciding on sampling frequency is
that if the measurements are being made in order to assess whether a
given environmental quality standard is being satisfied, then the data
resolution must be sufficient for that purpose. As an example the US
Ambient Air Quality Standard for sulfur dioxide specifies that 3 hourly
and 24 hourly values are required. It would be pointless to collect SO2
data with a continuous fluorescent analyser with a response time of two

Table 6

Summary of some methods of analysis of water

Pollutant or determinand

Measurement technique

pH
Biochemical oxygen demand
Chemical oxygen demand
Metals
Organometallics
Nitrate
Nitrate
Formaldehyde
Phenols

Electrometric
Dilution/incubation
Dichromate oxidation
AAS
GC-AAS
Colourimetric
UV spectrophotometric
Photometric
GC

Response time
10 s
5 days
2h
5 min
1 min
6 min
30 min

minutes, if the only purpose of the exercise is to see whether this standard
is being met.
5.2

Methods of Reducing Sampling Frequency

Once the desired sampling frequency has been selected it may be found to
be impracticable with the resources available and some means of reducing
the number of samples will then be required. This may be done by:
(a) reducing the number of sampling locations,
(b) reducing the sampling frequency, or
(c) reducing the number of determinands.
It has already been shown above that there are several methods available
for the rationalization of an existing monitoring network where the
quality at one location is correlated sufficiently well with that at another
location, and these methods of analysis may be applied to all types of
monitoring networks. Similarly, statistical analysis of past data may
show that one determinand is sufficiently well correlated with one or
more others, such that it may be used as an indicator of quality.
When sampling water, soils, sediment, and flora and fauna the use of
composite samples may be of value in reducing sample numbers.
Composite samples are formed by mixing together individual samples
to give an indication of the average quality over an area or during a
sampling period. They may also be formed by continuous or intermittent
collection of samples into one container over a given period, as, for
example, the collection of atmospheric particulate material on a filter.
Individual samples collected at different locations may be mixed together
in proportion to the volumes of the sampled bodies, as in the case of a
non-homogeneous water body, and so give a better indication of average
quality.

In the case of atmospheric pollutants it is often desirable to estimate
the likely daily maximum as well as the daily mean concentration, and
several methods have been proposed that allow this to be done on the
basis of a few discrete samples of short duration. For example, the same
statistical concentration frequency distribution of SO2 levels as is
provided by continuously recorded data can be obtained from a limited
number of randomly collected short term samples.
5.3

Number of Sampling Sites

The choice of the number of sampling sites to be used in a particular
survey is very dependent, as are so many other design parameters, on the
objectives of the study. In the simplest case of source-orientated
monitoring of atmospheric emissions from a single stack or site then 46 sampling locations might be considered sufficient. This may be thought
to be too few but operational constraints may prevent this number being
increased. In the case of the UK National Survey of Air Pollution it was
previously thought necessary to obtain daily data from about 1200 sites,
although this number has now been greatly reduced.
When too few sampling sites are used in a source-orientated monitoring programme it is possible that atmospheric (or aqueous) emissions
may pass between them without being detected at all. The probability of
a fixed number of sample stations detecting a release is a function of the
quantity released, the number of samplers, the distance of the samplers
from the source, the plume dimensions, and the height and duration of
the release. This type of analysis has been applied to the environmental
monitoring of atmospheric releases.
A recent innovation is the use of hand-held Global Positioning
Systems for accurately determining the longitude and latitude of a
sampling point on the Earth's surface. This is particularly useful when it
is necessary to return to the same point for further sampling.
6 PREREQUISITES FOR MONITORING
Before monitoring begins, certain information, techniques, and methodologies must be available in order for the survey to be successfully
carried out. As has already been stressed, a prime requirement is that the
objectives of the study should be defined, but the following also need
consideration:
— definition of a monitoring protocol;
— availability of meteorological or hydrological data;
— availability of emission data;

—
—
—
—

6.1

likely pollutant concentrations;
availability of suitable monitoring equipment;
availability of sensitive and specific analytical techniques;
definition of suitable environmental quality standards.

Monitoring Protocol

Monitoring is a complex task and carefully planned and documented
procedures are necessary to ensure that reliable and comparable results
are obtained.32 The documented procedure by which this is achieved is
known as a Monitoring Protocol. The main components of a Monitoring
Protocol are as follows:
(a) A reference measurement method—standard methods may not
always be available so some reference methods may actually be
non-standard.
(b) A standard methodology which sets out for each process:
(i) frequency of sampling;
(ii) duration of sampling and the number of samples to be taken;
(iii) pollutants to be measured;
(iv) accuracy required (to some extent this is dictated by the
pollutant and the sampling method used);
(v) variability of the pollutants under investigation;
(vi) analytical methods;
(vii) analytical resolution;
(viii) general principles of sampling (e.g. position, isokinetic);
(ix) plant operating conditions (for source monitoring);
(x) health and safety considerations of sampling staff (and plant
operators for source monitoring);
(xi) procedures to be adopted when meaningful measurements
by standard procedures are not possible.
(c) Quality control procedures which define requirements for:
(i) calibration of measurement devices;
(ii) maintenance of instruments;
(iii) sample storage and transport to ensure that the sample is
identifiable throughout the sampling, sample preparation
and analysis, and to ensure the sample integrity is maintained;
(iv) data handling and reporting.
32

HM Inspectorate of Pollution, 'Technical Guidance Note (Monitoring) M2: Monitoring Emissions at Source', HMSO, London, 1993.

(d) A quality assurance programme to ensure that:
(i) measurements are made in accordance with a standard
methodology;
(ii) the correct quality control procedures are in place;
(iii) the quality control procedures are being adhered to;
(iv) sample identification and routing procedure is well documented;
(v) the correct reporting procedures are used.
It is most important to have a documented Monitoring Protocol when
source monitoring is being carried out for comparison against emission
limits or Environmental Quality Standards. Regulatory Authorities may
wish to agree such a protocol prior to monitoring being carried out.
Since each monitoring programme is different both in the way in which it
is operated and in the pollutants which are monitored, separate Protocols will be needed for each programme. Measurement teams, operators,
and contractors should be instructed in the use of the Protocols.
6.2

Meteorological Data

When carrying out air pollution measurements it is desirable, and often
essential, to have access to meteorological data. Care must be taken to
ensure that the data used are meaningful and representative of the area
of study, wind data in particular being very susceptible to local
interference. Light, robust anemometers (e.g. hog wire) and wind vanes
are now generally available and when mounted on a dismountable 10 m
mast may be used in the field, thus obviating the need to rely on data
obtained from another, possibly less representative, source. Less easily
obtained parameters that may be required are the lapse rate and the
height of any atmospheric temperature inversions. These are rather
difficult to measure, requiring accurate measurements of temperature at
increasing heights or acoustic radar observations, but such data are
usually obtainable from large meteorological stations.
Measurements of low-level atmospheric turbulence are made using a
bivane and anemometer on a 10 m mast. This instrument measures the
horizontal and vertical fluctuations of the wind; siting of the mast is
obviously critical and a data logging system is required to cope with the
large amounts of data generated.
Meteorological data may be required for several reasons. For instance,
when fast response measurements are unavailable it may be desirable to
construct time-weighted pollution roses, which show how pollutant
concentrations vary with surface wind directions. For this the hourly
wind direction at 10 m height is determined, the record is divided into

sixteen 22.5° sectors, and the duration in each sector is tabulated. The
rose may then be calculated using:

where (TWMC),, is the time-weighted mean concentration of the
pollutant for the nth sector, tin is the number of hours for which the
wind was in sector n during the /th sampling period, ct is the concentration of the pollutant for the /th period, m is the number of sampling
periods and n takes values from 1 to 16 (sector n = 1 being 0-22.5°,
sector n = 2 being 22.5^5° etc.).
Alternatively the trajectory of a parcel of air over synoptic-scale
distances may be required for source-apportionment or dispersion
studies. For this the surface pressure field over a very large area is
needed and the geostrophic wind vector estimated from the isobar
spacing and alignment.33 The analogous hydrographic data may be
required for dispersion studies in the sea, and flow data for rivers and
lakes may be needed although these may not be so readily available as
meteorological data.
6.3

Source Inventory

One, often very cost-effective, method of identifying the likely occurrence
of pollutants prior to actual monitoring is by collecting and collating
detailed information on the pollution emissions in a given area or to a
particular river. Such an emission or source inventory should contain as
much information as possible on the types of source as well as the
composition of emissions and the rates of discharge of individual
pollutants. Supplementary information describing the raw materials,
processes, and control techniques used should also be collected. Detailed
discussion on the test procedures to adopt in compiling an emission
inventory is available,34 and the application of the completed inventory
were identified by these authors to be:
(a) guiding emission-reduction efforts;
(b) helping to locate monitoring stations and alerting networks;
33
34

R. I. Sykes and X. X. Hatton, Atmos. Environ., 1976, 10, 925-934.
A. T. Rossano and T. A. Rolander, 'Manual on Urban Air Quality Management', World Health
Organization, Copenhagen, 1976.

(c) indicating the seasonal and geographic distribution of the pollution burden;
(d) assisting in the development of implementation strategies;
(e) pointing out the priority of air or water quality problems;
(f) aiding regional planning and zoning;
(g) air and water quality diffusion modelling;
(h) predicting future air and water quality trends;
(i) determining cost-benefit ratios for air and water pollution control;
(j) community education and information programmes.
Although a very useful tool, the emission inventory is no substitute for
actual measurements of pollutant levels and should be considered as a
complementary technique to monitoring, not as an alternative.
6.4

Suitability of Analytical Techniques

As was shown in Table 2, pollutants may be found in environmental
media over very wide concentration ranges. Often there are several
procedures available by which a pollutant may be analysed, but they
may have widely differing sensitivity and specificity. As an example,
sulfur dioxide in air may be determined by absorption in hydrogen
peroxide and determination of the resultant acid by acid-base titration or by conductivity measurement. Although acidic or basic
compounds interfere, these techniques can yield useful results in
urban areas where sulfur dioxide concentrations are high and the
levels of interfering compounds are low. In rural areas this is not the
case due to natural production of ammonia, and misleading results
are obtained.
Very fast-response analysis of sulfur dioxide is possible using the flame
photometric sulfur analyser in which gaseous sulfur compounds are
burned in a reducing hydrogen-air flame and the emission of the S2
species at 394 nm is measured. The method is very sensitive but unless
used with gas chromatographic separation is not specific to sulfur
dioxide, but rather gives a measure of the total volatile sulfur content.
Alternatively SO2 may be analysed with a fast response time by excitation at 214 nm and the resultant fluorescence measured. This is now the
standard and recommended method for measuring SO2 concentrations
in air.
Some of the considerations that will affect the choice of an analytical
method may thus be summarized:
— sensitivity (depends on detection limit and pollutant levels);
— specificity (to allow unequivocal determinations);

Table 7

Instrumental analytical methods

Method

Sample*

Specificity

Detection limith

Gravimetric
Titrimetric
Visible spectroscopy
Ultraviolet spectroscopy
Flame emission spectroscopy
Atomic absorption spectroscopy
Gas chromatography
Liquid chromatography
Polar ography
Anodic stripping voltammetry
Spectrofluorimetry
Emission spectroscopy
X-ray
fluorescence
Neutron activation
Mass spectrometry

SLG
SLG
SL
SLG
SL
SL
GL
SL
L
L
SL
SL
SL
SL
SLG

good
good
fair
fair
good
excellent
excellent
good
good
good
good
excellent
good
excellent
good

> 1 yug
> 10~ 7 M in solution
> 0.005 ppm in solution
> 0.005 ppm in solution
> 0.001 ppm in solution
> 0.001 ppm in solution
> 10 ppm
> 0.001 ppm
>0.1 ppm
> 0.001 ppm
> 0.001 ppm
> 0.1 ppm
> 10 ppm
> 0.001 ppm
> 0.003 ppm

3
b

S = solid, L = liquid, G = gas.
Approximate only, depending upon the particular element being analysed.

—
—
—
—
—
—

response time;
response range (particularly linearity of continuous monitors);
ease of operation;
ease of calibration;
cost and reliability;
precision and accuracy.

Some of the more commonly used instrumental methods are shown in
Table 7 and described comprehensively elsewhere.35
Figure 13 highlights the difference between precision and accuracy for
an analytical technique. Precision may be defined as the reproducibility
of analyses, whereas accuracy is a true measure of the determinant
present. In order to assess the precision of an analytical method it is
necessary to analyse separate representative sub-samples of the sample
under investigation. These sub-samples should be included at intervals
during the analysis. In this way the drift in the precision of a technique
may also be detected. It is not possible to know how accurate a series of
determinations have been, even if they are precise. However, a best
estimate of the accuracy of an analytical technique may be found from
the regular analysis of national and international standard reference
materials. These have undergone inter-laboratory analyses, using different techniques and calibration standards, for a variety of species.
35

'Instrumental Analysis of Pollutants', ed. C. N. Hewitt, Elsevier, London, 1991.

Not accurate
Not precise

Accurate mean
Not precise

Not accurate
Precise

Accurate
Precise

Figure 13 Schematic illustration of precision and accuracy

6.5

Environmental Quality Standards

Of the six possible reasons for carrying out a monitoring programme
outlined in Section 1, four rely upon the prior formulation of a standard
of environmental quality. Only in the case of source apportionment and
pollutant interaction and pathway studies or when monitoring is carried
out with the intention of obtaining a historical record of environmental
quality is this not a prior requirement. There is little point in monitoring
in order to pinpoint pollutant health effects or to assess the need for
legislative controls on emissions, for example, unless a certain pollutant
level has been defined as being undesirable or likely to cause damage.
Environmental quality standards have been devised and adopted for
many atmospheric and water pollutants, some of which are shown in
Tables 8 and 9.
Having adopted a standard for environmental quality there may be
great difficulty in ensuring compliance. In the case of, for example, lead
in drinking water, various reduction strategies are possible, culminating
in the wholesale removal of lead pipes (although this may not entirely
solve the problem when lead-based solder is used with copper pipe) and
in the case of primary air pollutants similar 'simple' remedies are
possible. The difficulties arise in the case of secondary pollutants {i.e.
those formed within the atmosphere itself) or for pollutants with both
primary and secondary origins. In the case of atmospheric suspended
particles (smoke) both primary and secondary sources are important.
Primary emissions have in recent years been greatly reduced by the use of

Table 8

US ambient air quality standards

Pollutant

Measurement period

Standard

PM 10

24 h average
Annual arithmetic mean

150/igm~ 3
50 fig m ~ 3

PM2.5

24 h average
Annual arithmetic mean

65 /ig m~ 3
15 fig m ~ 3

Sulfur dioxide

Annual arithmetic mean
24 h average
3 h average

80 /igm~ 3
365 fig m ~ 3
1300/igm~ 3

Carbon monoxide

8 h average

10 mg m~ 3

1 h average

40 mg m~ 3

Ozone

8 h average

157 fig m~ 3

Nitrogen dioxide

Annual arithmetic mean

100 jug m~ 3

Lead

Quarterly average

Table 9

1.5 fig m ~ 3

EC water quality standards for inland waters (e.g. rivers, lakes)

Pollutant

Measurement
Country period

Concentration

Aldrin, dieldrin, endrin, and isodrin

0.03 /ig/litre (total)

Annual mean

0.005 /ig/litre endrin
Cadmium and its compounds

Annual mean

5 fig/litre (total)

Carbon tetrachloride

Annual mean

12 /ig/litre

Chloroform

Annual mean

12/ig/litre

DDT (all isomers)

Annual mean

0.025 /ig/litre

p,p'-DDT

Annual mean

0.01 /ig/litre

Hexachlorobenzene

Annual mean

0.03 /ig/litre

Hexachlorobutadiene

Annual mean

0.1 fig/litre

Hexachlorocyclohexane (all isomers)

Annual mean

0.1 /ig/litre

Mercury and its compounds

Annual mean

1 /ig/litre (total)

Pentachlorophenol and its compounds

Annual mean

2 /ig/litre

efficient control techniques on industrial sources and substantial change
in domestic fuel usage from coal towards cleaner fuels. However
secondary particles, produced in the atmosphere by formation of the
ammonium salts of strong acids from industrial emissions of SO2, NO x
and HCl, together comprise a substantial proportion of the atmospheric
aerosol. Thus reduction of primary emissions of particles does not
necessarily ensure a reduction in atmospheric concentrations or compli-

ance with a standard. One secondary air pollutant that is likely to prove
difficult to control adequately in the next decade or so is ozone. This is
formed in the lower atmosphere by reactions involving several primary
pollutants, and our present understanding of the chemistry of the
atmosphere is probably insufficient to accurately predict the effect of
control strategies.
7 REMOTE SENSING OF POLLUTANT
Many highly sophisticated techniques are now available for the remote
sensing of atmospheric and water pollutants. However, their use is
almost exclusively restricted to specialized monitoring exercises due to
the very considerable capital cost of the instrumentation. Probably the
cheapest and most widely used methods are those of aerial photography,
including infrared sensing and optical correlation spectrometry. Uses of
aerial photography include the monitoring of liquid effluent dispersion
using dye tracers and conventional colour film. Infrared photography
has been used for monitoring the condition of crops and forests.
Airborne heat-sensing infrared linescanning equipment has been routinely used to monitor thermal plumes in waters receiving industrial
effluents and also for the detection and mapping of oil spills at sea using
thermal infrared data from satellites.
The most common use of the correlation spectrometer in air pollution
analysis is for the determination of sulfur dioxide and nitrogen dioxide
concentrations in plumes from tall stacks, and this provides a good
technique for studying the transport and dispersion of a plume. The
instrumentation may be ground based in a mobile laboratory or airborne, in which case the plume is viewed from above.
The use of tunable lasers allows long-path absorption measurements
of a range of gaseous pollutants such as SO, NO2, SO2, CO and O3 and
minor reactive species such as OH*. Reliable measurements of this latter
species are of great importance because of its dominant role in the
chemistry of the troposphere. In long-path laser adsorption methods a
detector is used to monitor adsorption of specific wavelengths in the light
path. In lidar techniques, however, the back-scattered radiation from a
laser is monitored. By using a pulsed system the time taken for receipt of
back scatter can be related to the distance of travel, allowing spatial
resolution of pollutant concentration data within the light path. By
monitoring back-scatter intensity at two close wavelengths, one strongly
absorbed by the species of interest and one unabsorbed, the species'
concentration may be inferred as well as its spatial distribution. Care is
required to avoid spectral interferences but this method has been
successfully used for measurement of sulfur dioxide up to a range of ca.

Altitude (km)

16 Oct

28 Aug

Pressure (mb)

McMurdo 1986

Ozone partial pressure (nb)
Figure 14

Ozone profiles from balloon soundings over McMurdo (South Pole) during
1986

(After Hoffman et al., Nature, 1987,326, 59-62) (reproduced with permission
from 'Stratospheric Ozone', Department of the Environment, HMSO,
London, 1987)

2 km. Further significant developments of laser methods using the
Raman back scatter, which is highly characteristic of the scattering
molecule, are likely.
Case study 7: Measurement of stratospheric ozone.36 There are several
techniques by which stratospheric ozone may be measured. From the
ground, the total ozone in a column extending vertically into the atmosphere can be measured passively with a Dobson spectrophotometer.
This spectrophotometer is also used to estimate the vertical distribution
of ozone, by observing the zenith clear blue sky, while the sun traverses a
range of solar zenith angles. Scattering from different altitudes in the
atmosphere allows the height profile of ozone to be deduced. Greatly
improved vertical resolution of ozone concentrations has been achieved
using the differential absorption laser (DIAL) technique. Ozone profiles
have been measured from spectrophotometers carried aloft by balloon or
rocket. The balloon-sondes provided evidence of a reduction in stratospheric ozone over Antarctica during October 1986 (Figure 14).
A major achievement in the remote sensing of ozone has been the
verified measurements of ozone profiles obtained from satellites.
36

UK Stratospheric Ozone Review Group, 'Stratospheric Ozone', HMSO, London, 1987.

Measurements have been made in the ultraviolet (SBUV/TOMS), visible
(SAGE) and infrared (LIMS) spectral regions on board the Nimbus 7,
Applications Explorer II, and Solar Mesospheric Explorer satellites
respectively. An improved version of the visible spectrophotometer
(SAGE-2) is operational on the Earth Radiation Budget Experiment
(ERBE) satellite and uses an on-board calibration lamp to check
instrumental drift.
8 PRESENTATION OF DATA
A monitoring programme, particularly one incorporating automatic or
fast-response systems, can generate a very large amount of information
in a short time. In order for the data to be assimilated and understood,
some means of organizing the information and summarizing its most
essential characteristics is required so that changes, trends, or patterns in
behaviour over time and space may be apparent. For this the methods of
descriptive statistics are required. Another group of statistical methods,
those of inferential statistics, are used when information of the relationships and processes operating between measurements is required. Details
of these methods are available from standard texts and from books
devoted to the environmental sciences (e.g. References 37-38).
Many statistical tests depend upon having data that are normally
distributed, but often environmental analytical data do not satisfy this
criterion. In a normal distribution the arithmetic mean and median are
the same, but in log normally distributed data the geometric mean and
median are the same. This is the situation that applies to many
environmental data sets and comes about from a few results having
high values whilst the majority of results are closely grouped together. If
such data are treated as belonging to a normal distribution too much
weight will be applied to the outlying values and the wrong deductions
may be made. In this case some method of transforming the data is
required before statistical analysis is carried out. For example, it might
be appropriate to use the logarithms of the data, or the square or cube
roots. Similarly it is often better to quote the 95 percentile range of
values, excluding the extreme 5%, again in order to avoid giving
prominence to a few outliers.
Monitoring data may be incorporated in Geographical Information
Systems (GIS) which are specialized database management systems for
handling geographical data—i.e. data whose key characteristic is 'place'.
These manage data on land use, industry, roads, habitations, and so on.
37
38

C. Chatfield, 'Statistics for Technology', Chapman & Hall, London, 3rd Edn., 1983.
R. M. Haynes, 'Environmental Science Methods', ed. R. Haynes, Chapman & Hall, London,
1982.

(a)

Concentration

(b)

Jan

Feb

Mar
(d)

Absorbance

Concentration

(C)

Distance

Concentration

(e)
Smectite

litite

Chlorite

R ad ion ucl ide activity

(f)

Kaotinite
(g)

site 1

site 2

site 3

(h)

site

increasing concentration

Figure 15 Examples of different approaches to presenting data

Data includes both their positions and their various related properties
(for example, information about a road may include its traffic capacity,
traffic flows, and traffic speeds). In addition to simply storing such data,
GIS can manipulate and analyse the data—for example, calculating
contours of NO x emissions around road networks—and include visualization tools.
Many graphical methods of representing data are available to illustrate the changes or emphasise differences or similarities in the results
(e.g. Figure 15). These may take the form of bar charts/histograms
(Figure 15a), line graphs (15b), X Y plots (15c and d), pie charts (15e),
stacked bar charts (15f), geographical presentations (15g and h), or
indeed combinations of these. Figure 15c illustrates the inclusion of error
bars (or standard deviations) in a graphical presentation and also a
regression line. Data exhibiting an exponential decay will plot as a
straight line on a log normal graph (Figure 15d) and regression analysis
may be performed with the inclusion of 95% confidence limits. Where
separate components of a parameter are measured, the relative magnitude of those components may be represented on a pie chart (Figure 15c)
or stacked bar chart (Figure 15f). The overall magnitude of the
parameter can be proportional to the diameter of the pie chart or height
of the bar chart. These types of graphs are useful for displaying
speciation data. In cases where the geographical distribution of data is
important then contour plots may be appropriate (Figure 15g). If data
are limited or specific to certain types of location (e.g. road or river) they
may be displayed as in Figure 15h where the diameter of a circle, width of
bar, or length of line is proportional to the magnitude of the parameter
of interest.
It should always be borne in mind that the rigour applied to the design,
operation, and execution of a monitoring programme must also be
applied to the treatment given to the resultant information and to the
deductions made from it.
Questions
1. Discuss the ways in which the resources required for a monitoring
programme could be reduced without compromising the validity of
the results.
2. Prepare an outline specification for a monitoring programme to
establish whether dioxin emissions from a municipal waste incinerator contribute significantly to the ambient dioxin concentrations
in the vicinity of the incinerator.
3. Compare the advantages and disadvantages of remote sensing with
conventional in-situ environmental monitoring.

4. Describe how the results of regulatory compliance monitoring for
emissions from a sulfuric acid plant should be presented to the
Environment Agency and the general public. Reasons for the
differences in the approach should be highlighted and discussed.
5. Describe, with examples, the differences between planned, fugitive,
and accidental emissions of pollutants. Using a hypothetical
industrial process as an example, outline how a monitoring
network could be used to detect all three types of emissions.
6. Describe how mathematical modelling techniques can be used to
determine the optimum siting of a limited number of monitoring
devices in the vicinity of an industrial plant emitting pollutants to
the atmosphere.
7. A cement kiln company has been asked by the US EPA to provide
reassurance that the short term emissions of sulfur dioxide during
kiln start up do not breach the 3 hour air quality standard for sulfur
dioxide (1300 /zg m~ 3 ). As the company Environmental Manager
you decide to estimate the likely concentration of sulfur dioxide at
the point where the plume grounds.
Given that the height of the stack is 90 m, the stack gas velocity is
17ms" 1 , the internal stack diameter is 2.5 m, the mean windspeed
at a height of 10 m above ground-level (u\o) is 1 m s~l and the stack
gas temperature is 90 0C calculate the effective stack height.
The concentration of sulfur dioxide in the stack gas is 260 mg
m~ 3 during kiln start-up. The plume grounds most quickly in
Pasquill stability category A at a distance of 650 m from the base of
the stack and the concentration is greatest at this point on the
plume centre line. Given that the standard deviation of the plume
concentration in the vertical direction (az) is 97 m and the standard
deviation of the plume concentration in the vertical direction (oy) is
240 m (both at a distance of 650 m from the release point, for a 3 h
release and in Pasquill stability category A), assess whether it is
likely for the 3 hour air quality standard to be breached.
Discuss what action you should take to address the concerns of
the US EPA.

CHAPTER 8

Ecological and Health Effects of
Chemical Pollution
S. SMITH

1 INTRODUCTION
This chapter is concerned with the hazards of chemical pollution to
human health and ecological systems, and an evaluation of these requires
knowledge of the effects of pollutants, the levels which cause such effects,
and the likely occurrence of hazardous levels in the environment. Clearly
in tackling or preventing pollution hazards the aim is to ensure that
environmental levels are below those at which pollutants have known
effects.
There are, on the one hand, substances which we recognize as
pollutants and, on the other, organisms on which pollutants exert their
effect. A basic maxim of toxicology that can be extended to pollution is
that all substances administered at sufficiently high doses are harmful to
biota. Conceivably, biota may be able to cope with a small amount of a
very toxic substance, whereas an otherwise essential one may reach
overwhelming proportions and adversely affect many organisms.
The impact of pollution on biota depends on the amount entering the
environment and its fate in the environment. In the process of dispersion
in air, water, or soil, pollutants may undergo transformations to more
innocuous forms or to ones that are more hazardous. In the latter case
the impact is due to secondary forms of pollution. Dispersion in the
environment leads to dilution of pollution; however, certain environmental processes can concentrate pollutants and where this occurs the
impact may be greater. On the other hand they may become bound up or
immobilized in some way and in this form they are less available to biota
and may be considered less hazardous. The fate of pollutants in the
environment is covered in some detail in the preceding chapters.

Pollutants encompass a broad range of chemical and physical properties which strongly influence both their fate in the environment and their
effects on biological systems. Particularly during this century vast
quantities of many different chemical substances have been released
into the environment; the majority of these substances are waste
products generated by industry and society consuming manufactured
goods. Synthetic fabrics and fibres, pharmaceuticals, fertilizers, pesticides, paints, and building materials, as well as chemicals for industrial
processes, are just some of the products of the chemical industry that are
integral to almost every aspect of modern living. Many such substances
are natural constituents of the environment, others are synthetic chemicals. Inevitably wastes generated during the manufacturing process, the
substances themselves, and perhaps their degradation products, are
released into the environment. Gases (e.g. sulfur dioxide, carbon
dioxide, and oxides of nitrogen) and particulates from combustion
processes are discharged into the atmosphere. Liquids and solids
containing inorganic and organic substances are discharged into water
and hazardous chemical wastes are buried on land, incinerated, or
dumped at sea. Fires, explosions, and tanker accidents can result in
sudden unintentional but devastating pulses of pollutants into the
environment, and pesticides are intentionally released at certain times
of the growing season. It has to be said that there are now a number of
controls regulating such discharges and that these have arisen from
actual and perceived chemical pollution problems in the past.
Pollutants exert their effect on individuals, but recognition of an effect
is only seen when large numbers and even whole communities are
affected. The rate of supply or the amount of pollutant reaching a
receptor (organism, population, or community) is referred to as the
exposure and an effect is a biological change caused by an exposure.
Directly toxic substances must gain access to the exterior membranes or
interior components of tissues and cells in order to exert their effect. The
amount that is taken into an organism is referred to as the dose;
essentially it is a function of concentration and the period of exposure,
although dose is perhaps more correctly defined as the amount of
substance received at the site of effect. Dose can be a difficult parameter
to measure for pollutants, and exposure has become the standard means
of assessing how much is received by a receptor. Therefore, a fundamental goal of pollution studies is to establish the relationship between
exposure and effect, so that a level at which no effect occurs can be
identified and this can be set as an objective to work towards in
controlling pollution.
A pollutant may be supplied in one large pulse (acute exposure) or an
equivalent amount may be supplied at a lower concentration over an

extended period (chronic exposure). Similarly, the effects caused by a
pollutant are categorized as being acute or chronic. Acute effects are
observed immediately following an exposure; they are rarely reversible
and are very often fatal. Chronic effects or damage follow a period of
prolonged exposure which results in biochemical or physiological disturbances. Outwardly they may be manifested as visible or clinical
symptoms of damage, e.g. chlorotic regions on plant leaves, and inability
to maintain homeostatic balance, co-ordinate activities or breed. These
symptoms are often quite diffuse and not specific to a particular
pollutant. Following cessation of exposure, chronic damage is frequently
reversible, although continued exposure may prove fatal. However, it
may be argued that any severe disturbance will impair the efficiency of an
individual and so shorten its life.
As well as the direct impact of pollutants on living species, other
secondary effects require careful consideration. These relate to community and habitat changes that may follow initial pollution damage.
Obvious examples include disturbance of predator-prey relationships
following a dramatic decline of one or more species in a food web and
reduced turnover in biogeochemical cycling of nutrient elements if, say,
decomposer organisms are affected. These secondary impacts are a
feature of the ecosystem level of organization.
Pollution studies are very wide ranging and so extend over a multitude
of situations involving interactions of pollutants with living organisms.
Damage to certain groups of organisms arouses more concern than
others. Clearly humans and their resource species—livestock, crops, and
fisheries—for obvious social and economic reasons are the two groups
which engender most concern. Surveying the range of effects of pollutants, it is evident that these also can be arranged on a scale of increasing
concern, with cellular disruption arousing less concern than an acutely
toxic effect.1
2 TOXICITY: EXPOSURE-RESPONSE RELATIONSHIPS
The toxicity of a substance or industrial discharge is estimated by
determining the concentration or dose which is lethal to a particular
organism, and the most commonly used yardstick is the LC50 or LD50,
the concentration or dose which results in 50% mortality of test
organisms.
Because chemicals discharged into the environment largely end up in
aquatic systems, toxicity tests are commonly based on aquatic organisms. There is an extensive database on the acute toxicities of a variety of
1

M . W. Holdgate, 'A Perspective of Environmental Pollution', Cambridge University Press,
Cambridge, 1979.

chemicals, although by no means all potential contaminants. The
organisms selected as test organisms are those that are easy to obtain
and maintain in the laboratory. In the UK toxicity tests are commonly
undertaken with the water flea {Daphnia sp.) and trout; in the US the
fathead minnow is the fish most frequently used.
Mortality is only one of many end-points of toxicity; others relate to
effects on the general wellbeing and survival of an organism, e.g. growth,
reproduction, and feeding rate. The parameter used in these cases is the
effective concentration or EC50, the concentration that results in 50%
reduction in growth or some physiological function. At the biochemical
level the emphasis is on determining the mode of toxic action. Chronic
toxicity tests are based on the same principle. Test organisms are
subjected to a range of concentrations of pollutants over extended
periods and their effects on growth and reproduction are determined.
Toxicity tests are examples of dose or exposure-response relationships,
in that the proportion of test organisms showing the damaging effect is
measured. These relationships are fundamental to all forms of pollution;
they provide a framework for estimating exposures associated with the
onset of effects and in turn those that have no effect. The no-effect
exposure is the one which is striven for in the environment. To get to this
stage a large number of tests on a variety of organisms are required to
ascertain the most sensitive species and the most sensitive end-point.
Many thousands of chemicals have been and currently are discharged
into the environment, and it is difficult to provide full toxicological data
on all of them. Many belong to chemical groups or homologous series
which have related physico-chemical properties and it is possible to use
these to predict their toxicological characteristics and fate in the
environment. Examples of homologous series are the chlorophenols
and chlorobenzenes. These are based on phenol and benzene respectively
and in each case chlorine can be added to form mono- or polychlorinated
compounds. The end member of the chlorophenols is pentachlorophenol
and that of the chlorobenzenes is hexachlorobenzene. Pentachlorophenol has been used as a wood preservative and hexachlorobenzene has
pesticidal properties. Toxicity increases progressively through each of
these series such that phenol and benzene are the least toxic, while the
most toxic ones are the end members of each series, pentachlorophenol
and hexachlorobenzene.
This progression can be related to the accumulative capacity of the
compounds and to one particular parameter, the octanol-water partition
coefficient (ATQW)- This parameter is a measurement of the tendency of
organic compounds to partition between water and an organic solvent.
Taking equal volumes of water and octanol, a water soluble compound
will reside mainly in the water, whilst a non-polar organic compound

LoglO (1/LC50) of Selected
Phenolic Cpds in Sole

Partition Coefficient (Kow) for
Selected Phenolic Cpds
Figure 1 Relationships between octanol-water partition coefficient ¥^ow and toxicity of
phenol and five chlorophenols to sole, Solea solea

(Reproduced with permission from Furay, PhD, University of London, 1994)

with low water solubility will be mainly in the octanol. The ratio of the
amount of a compound in octanol to that in water is the octanol-water
partition coefficient (Kow) and clearly hydrophobic substances will have
large values for these coefficients. This physico-chemical system is
analogous to the situation where a pollutant in water is exposed to an
organism, the lipid in the cell membranes being equivalent to octanol.
Therefore substances with a high KOw (hydrophobic and low water
solubility) will partition from the water to the lipid component and enter
the cells of the organism and such substances will bioconcentrate to a
significant extent.
Figure 1 shows the relationship between the toxicity and the A^Ow of
chlorophenols, essentially the most toxic members are the ones with the
highest KOw. These relationships are known as 'Quantitative Structure
Activity Relationships' or QSAR. They are useful for predicting the
toxicity of chemicals with related properties. They can also give some
insight into different modes of toxic action. The gradient of a similar
relationship with chlorobenzenes tends towards unity and it is typical of
a large number of organic hydrophobic chemicals that have a general
narcotic effect. Chlorophenols on the other hand form a different
gradient and these chemicals are known to be potent electron uncouplers
in aerobic respiration.

3 EXPOSURE
Exposure is the amount of pollutant reaching a receptor, and is
commonly expressed as the concentration of pollutant in the media (e.g.
air, water, soil) and in food that is available to the receptor (individual,
population, community). It is important to take into account the period
of exposure to estimate the total or rate of supply to a receptor and in
turn the intake (or dose). In the case of the more persistent pollutants,
the amount or concentration in an organism, or some part of an
organism, is a function of exposure. Consequently measurements of
such concentrations can be used to indicate exposure.
Experimental studies aim to establish exposure-response relationships
and measurements of exposure in the environment indicate whether
levels are hazardous. Whilst this is a simple concept, in practice it is not
quite so easy to translate experimentally derived data into the ambient
environment. Exposure can arise from a single source of supply, as in the
transfer of gaseous air pollutants such as sulfur dioxide (SO2), nitrogen
dioxide (NO2), or ozone (O3) from the atmosphere to the leaf surfaces of
plants. Alternatively more than one pathway may contribute to the total
exposure, for example human exposure to toxic metals such as lead and
mercury and many pesticides is from water, air, and food.
General measurements of substances in air, water or soil may not be
representative of the actual amount that reaches a receptor. This may be
a consequence of how, when, and where measurements are taken, or
because a pollutant exists in more than one form, all of which may not be
equally available to biota. Exposure can vary in space and time; in rivers
the concentrations of pollutants vary with season, time of day, magnitude of freshwater runoff, depth of sampling, intermittent flow of
industrial effluent, and hydrological factors such as tides and currents.
Human exposure to atmospheric pollutants such as lead (Pb), SO2, and
airborne particulate matter can be equally if not more variable and
therefore difficult to estimate accurately. Many samples are required to
iron out variation due to methods of collection and analysis as well as
actual spatial and time differences. In practice, however, exposure
estimates are based on data from a small number of monitoring stations
(perhaps only one) representing a relatively large area.
There are further difficulties in taking account of the movement and
different activities of people. Many adults spend a part of their day at
work and children at school, and consequently such groups may
experience different exposure regimes during these times. We all spend
extended periods indoors removed from the outside air, and certain
susceptible groups such as the aged spend most of their time indoors.
Furthermore, such habits as tobacco smoking, which is a very direct

form of air pollution, quite obviously increases a person's exposure to an
array of substances and it is well known that cigarette smoking provides
the major supply of cadmium (Cd) to habitual smokers; smoking 20
cigarettes can result in a daily intake of 2-4 fig Cd.2'3
Much of the research on the effects of gaseous air pollutants on plants
has been done in the controlled environment of growth chambers. The
concentration of the gas in question in the inlet or outlet of the chambers
has been used as an estimate of plant exposure. These values, however,
can be quite different from those at the leaf surface. Design and size of
growth chamber, density of plants, and air velocity are just three of many
plant and environmental factors that can influence the concentration
gradient between ambient air around the plants and that in contact with
the leaf surfaces. These and other factors have a marked effect on the
thickness of the boundary layer of air that surrounds the leaves. Gases
must pass through this almost laminar flow of air by the relatively slow
process of molecular diffusion to gain access to the leaf surfaces. In
situations where varied responses have been reported for seemingly
equivalent exposures, it is apparent that at least a part of the discrepancy
can be explained by the experimental conditions and the way in which
exposure was quantified. Open top chambers permit plants to be exposed
to natural precipitation regimes and their design is such that the
microclimate inside is similar to that in the surrounding environment.
Nevertheless, the forced ventilation within these chambers can have an
effect on pollutant deposition and over a growing season fluxes of
pollutants may be enhanced over that in ambient air, which means that
these chambers may overestimate the sensitivity of vegetation to air
pollutants.4
The nature of vegetation can affect the flux of air pollutants. The
canopy of a forest creates a larger resistance than say cereals or grassland
with the result that there is more turbulence and larger inputs or fluxes of
pollutants to forests. The deposition of O3 to forests is influenced by the
opening and closure of stomata; O3 uptake is known to be less in tree
species that close their stomata during the night-time.4
The physico-chemical properties of pollutants and the nature of the
environment into which they are discharged dictate their fate and
therefore the exposure levels encountered by organisms. This can be
illustrated by a simple example in which equilibrium partitioning of
organic chemicals between air, water, and biota and organic matter, can
2

'Biological Monitoring of Toxic Metals', ed. T. W. Clarkson, L. Friberg, G. F. Nordberg, and P. R.
Sager, Plenum, New York and London, 1988.
'Handbook on the Toxicology of Metals', ed. L. Friberg, G. F. Nordberg, and V. B. Vouk,
Elsevier, 2nd Edn., 1986, Vol. 2.
4
United Kingdom Critical Loads Advisory Group, 'Critical Levels of Air Pollutants for the United
Kingdom', Department of the Environment, London, 1996.
3

Air

Water
V=1E+4
C = %/1E+4

Octanol
C = %/1

Figure 2 Locations of phenol, alkylphenols, and chlorophenols on a triangular diagram of
volume ratios VA: Vw: V0 ofl(f:104:l. The arrow depicts the effect of increasing
substitution of chlorine. A, phenol; B, o-cresol; C, m-cresol; D, p-cresol; E, 2,4dimethylphenol; F, 2,6-dimethylphenol; G, 3,4-dimethylphenol; H, 3,5dimethylphenol; I, 2,4,6-trimethylphenol; J, 2-chlorophenol; K, 3-chlorophenol;
L, 4-chlorophenol; M, 2,4-dichlorophenol; N, 2,6-dichlorophenol; O, 2,3,4trichlorophenol; P, 2,3,5-trichlorophenol; Q, 2,4,5-trichlorophenol; R, 2,4,6trichlorophenol; S, 2,3,4,5-tetrachlorophenol; T, 2,3,4,6-tetrachlorophenol; U,
2,3,5,6-tetrachlorophenol; V, pentachlorophenol
(Reproduced with permission from Environ. Toxicol. Chem., 1995, 14, 1839)

be represented by three properties; water solubility, vapour pressure,
and the octanol-water partition coefficient. Water soluble substances
will tend to remain in water, substances with a high vapour pressure
will evaporate, and hydrophobic or lipophilic ones will partition to
lipids in living and dead organic matter. A^Ow values can be used to
assess this last tendency. Figure 2 shows the relative mass distribution
of phenol, alkylphenols, and chlorophenols between water, air, and
'octanol' compartments. Phenol and some of the alkylphenols tend to
remain in the water compartment, mono-, di-, and tri-chlorophenols are
relatively volatile, and the more highly chlorinated members, with
higher X0W values, accumulate in organic matter. Benzene and the
lower chlorobenzenes have high vapour pressures and therefore escape

to the air; the higher members of the series are lipophilic and
bioconcentrate in biota.
Clearly the actual environment is more complex. Organic matter in
soil and sediment is an important substrate for the adsorption and
accumulation of lipophilic substances. Once bound to organic matter in
particles such lipophilic substances are less available to organisms,
including bacteria. There is evidence that the bioavailability of these
bound forms can decline with time as the compounds become more
tightly sequestered or less accessible within the organic matter.5 A
further level of complexity is that organic chemicals are susceptible to
chemical or biological degradation. Some are broken down quite quickly
whilst others, such as the larger polycyclic aromatic hydrocarbons
(PAHs) and PCBs, are resistant to breakdown and are persistent in soil
and sediment for many years.
Metals exist in different forms. In soils and aquatic systems, metals are
partitioned between solid and liquid phases and within each, further
partitioning or speciation occurs among specific ligands, determined by
ligand concentration and the strength of each metal-ligand association.
Consequently, at any one time the amount of free ion available for
uptake by organisms is less, and in many instances much less, than the
total concentration. As a rule the free ion (e.g. Cd2 + , Cu2 + , Pb2 + ) is the
most bioavailable and therefore the most toxic, although organometallic
compounds such as methylmercury and tributyltin are more available
and therefore more completely absorbed than their free ion counterparts.
Organisms with different feeding habits can have very different
exposures in the same ecosystem. Those foraging in sediments may be
exposed to elevated levels of metals or lipophilic organic compounds
whereas those swimming in the water may have a much lower exposure.
The amount accumulated depends on the bioavailability of the substances in different parts of the system and the integrated level in biota is
one way of assessing this. Increasingly, exposure assessment is carried
out using computer models which simulate environmental processes and
pollutant inputs to predict levels of pollutants reaching different types of
receptors.
4 ABSORPTION
Absorption is the process whereby a substance traverses the body
membranes. By and large, lipid soluble substances are absorbed more
efficiently than polar water soluble substances. Lipid soluble substances
diffuse across the lipid bilayer of membranes and those that are more
5

M . Alexander, Environ. ScL TechnoL, 1995, 29, 2713-2717.

lipid soluble (i.e. those with a larger octanol-water partition coefficient)
tend to be absorbed more efficiently. Complex carrier systems involving
the membrane proteins serve to transport polar substances such as
metals, although certain forms which are more lipid soluble or of
reduced polarity traverse membranes by passive diffusion, and this is
the most likely explanation for the greater accumulation of such
compounds as methylmercury and tributyltin.
The characteristics of the interface between the environment and an
organism strongly influences the absorption of a substance. In mammals
the main routes of entry for environmental chemicals are through the
lungs, the skin, and the gastrointestinal tract. The physico-chemical
properties of a pollutant and the nature of exposure strongly influence
the amount presented at each portal of entry. In terrestrial mammals,
including humans, penetration through the skin is an unimportant route
of entry for environmental pollutants. The skin is relatively impermeable
to aqueous solutions and most ions, although many synthetic organic
chemicals, because of their lipid solubility, can penetrate the dermal
barrier. Such situations tend to be confined to occupational exposures.
Atmospheric pollutants occur as gases or as particulate matter. The
site and extent of absorption of inhaled gases are, for the most part,
determined by their water solubility. For instance, sulfur dioxide (SO2) is
a soluble gas, which is mainly absorbed in the upper respiratory tract
whereas a less soluble gas such as nitrogen dioxide (NO2) reaches the
lower airways. Inhalation is the most important route of uptake for
elemental mercury vapour, and the major site of absorption is alveolar
tissue where about 80% of inhaled mercury vapour is absorbed.2'3'6
The fraction of inhaled particulate material deposited in the various
parts of the respiratory tract is a function of particle size, and the fraction
absorbed from the tract is dependent on the chemical nature of the
aerosol. The upper limit for respirable particles is of the order of 10 /mi
diameter and modern air monitoring equipment has specially engineered
intakes to sample particles of less than this diameter as a means of
estimating the respirable fraction. This is now generally referred to as
PMlO.7 Only particles less than 2 /im diameter penetrate as far as the
alveolar region. Larger particles are trapped in the upper tracheobronchial and nasopharyngeal regions where they may be either absorbed or
transported up the pharynx entrained in mucus propelled by ciliary
action. Subsequently, these particles are swallowed and become available
for absorption in the gastrointestinal tract. Particles taken up from the
alveoli (the most important site of absorption) may pass directly into the
6
'Environmental
7

Health Criteria, 1, Mercury', World Health Organization, Geneva, 1976.
Quality of the Urban Air Review Group, '3rd Report, Airborne Particulate Matter in the United
Kingdom', Department of the Environment, London, 1996.

bloodstream or be retained in lung tissue. For respirable particles of lead,
at a particle size of 0.05 /im about 40% may be retained but for larger
sizes, e.g. 0.5 /mi, only about 20% is deposited and probably only about
50% of the deposited lead is absorbed into the blood, although this will
depend on the solubility of lead compounds in the particles. Deposition
in the alveolar region of inhaled cadmium (Cd) aerosols up to 2 /mi
diameter is 20-35% of which less than 50% is absorbed; the rest is
exhaled or swallowed.2'3
In humans, absorption through the gastrointestinal tract is an important route of entry for many environmental chemicals. The circulatory
system is closely associated with the intestinal tract, and once toxicants
have crossed the epithelium, entry into capillaries is rapidly effected.
Venous blood flowing from the stomach and intestine introduces
absorbed materials to the hepatic portal vein, resulting in transport to
the liver, the main site of metabolism of foreign compounds. Exposure to
metals in the general environment is usually greater via food and drink
than via air. However, the absorption of ingested lead and cadmium in
adults is normally relatively low, being about 10% for lead and 5% for
cadmium. Absorption is influenced by various dietary factors, and
increased absorption of lead has been found in cases of low dietary
calcium. Absorption of dietary lead is much higher in young children. In
the case of inorganic mercury compounds absorption from foods is
about 7% of the ingested dose; in contrast, gastrointestinal absorption of
methylmercury is practically complete.2'3'6
The gills of aquatic animals present another type of interface between
the external medium and an organism and are an important route of
entry for water-dispersed pollutants. They present a large surface area of
diffusion, and at the same time continual circulation of water across the
gill filaments ensures maximum exposure.
One other type of external surface membrane worthy of note is that of
the leaves of plants. Leaf surfaces are covered by a waxy cuticular layer,
interspersed with stomatal pores which may occur on both leaf surfaces
or on a single surface (usually the lower). Stomata, as well as regulating
uptake of carbon dioxide and water loss, are the major route of entry for
gaseous pollutants. Rate of uptake of gaseous pollutants into a leaf is a
function of several physical factors; one such factor is the resistance to
diffusion caused by the boundary layer, which in turn depends on the
velocity and turbulence of airflow over the surface, and another factor is
the stomatal resistance. Before a pollutant can gain access and cause
injury within a plant cell it must first enter into solution in the
extracellular water enveloping the cell wall. In solution, SO2 is active in
the form of either HSO^ or SO2~. Ozone (O3) is less soluble, but since it
is a highly reactive molecule, it is thought that decomposition products

such as hydroxyl radicals and other free radicals are important reactive
species produced from reactions involving organic compounds.
Changes in stomatal aperture induced by air pollutants has attracted
considerable attention. This is not surprising since any change in gaseous
flux to and from metabolic sites in mesophyll tissues may eventually
affect the overall growth and yield of plants. The effect of SO2 on
stomatal aperture is complex; initial studies found enhanced stomatal
opening in Vicia faba plants exposed to SO2. It was also shown that
relative humidity strongly influenced the direction of the response;
humidities above 40% enhanced stomatal apertures, whereas at humidities below 40% apertures decreased. A host of different species have now
been investigated and it would appear that there is no uniformity of
response between species; both concentration of pollutant gas and
duration of exposure influence the outcome.8

5 INTERNAL PATHWAYS
Once inside an organism, a pollutant may follow a number of different
pathways. In general terms four main routes can be identified (see Figure
3): some molecules are metabolized (converted) into other compounds
which are frequently less toxic to the organism than the parent
compound, although some are more toxic. A second pathway involves
storage in certain tissues, e.g. lead in bone, cadmium in kidney, and DDT
in fatty tissues. Thirdly, a pollutant and its metabolites may be excreted
from an organism. Because metabolites are, as a rule, more water soluble
they are more easily excreted than the original pollutant, and metabolism
can be seen as an essential preliminary to excretion. On this rather
simplistic view, the remaining fraction can be regarded as being available
to exert an effect at a site of action.
The fundamental effect occurs at the molecular level; for example, lead
inhibits the action of ALA-D, an enzyme in the haem synthesis pathway,
although the significance of this to the overall toxicity of lead is not clear.
Organophosphorus insecticides specifically inhibit the action of acetylcholinesterase. Perhaps surprisingly, the primary toxic lesion of many
pollutants is unknown. A large group of non-polar organic substances
are thought to have a general narcotic action, whereby metabolism is
seen to slow down in proportion to exposure and eventually it ceases. In
recent years there has been considerable attention directed at finding
fundamental modes of toxic action. The realization that several pollutants can mimic natural hormones, and as a result disrupt the endocrine
8

J . Wolfenden and T. A. Mansfield, 'Acid Deposition—Its Nature and Impacts', ed. F. T. Last and
B. Watling, Proc. Roy. Soc. Edinburgh, 1996, Section B, 97, 117.

EXPOSURE
ABSORPTION
FREE FORM

BOUND FORM

TRANSLOCATION

SAIT
OF
CETION
STORAGE

METABOLS
IM

EXCRETION
CHEMICAL+ METABOLITES
Figure 3 Diagrammatic representation of the possible internal pathways followed by
pollutants

(Adapted with permission from T. A. Loomis, 'Essentials of Toxicology', Lea
and Febiger, Philadelphia, 2nd Edn., 1974)

system of animals, has raised the possibility that this is a basic event that
could explain widespread reproductive problems caused by many pollutants in wildlife and possibly humans (this is discussed more fully below).
Another area of growing importance is the possible role of highly
reactive free radicles in causing molecular disruption which may be
linked to short term damaging effects such as inflammation of lung
surfaces caused by air pollutants or to longer term damage to DNA and
the induction of cancers. Free radicals are atoms or molecules which are
capable of independent existence and which contain one or more
unpaired electrons, examples include H#, O*2~ and OH\ They are
produced in the body all the time and there are several scavenging
mechanisms for removing them and so preventing damaging effects.
Several pollutants have free radical properties or may be converted into
species with such properties, for example primary metabolites of PAHs.
Put simply, if the production of free radicals exceeds the body's capacity
to scavenge them, damage to molecular systems will occur and, in
particular, this may include damage to DNA. There is also a growing

number of reports providing evidence of metals such as Cd, PAHs and
PCBs having effects on the immune system of animals. The nervous,
immune, and endocrine systems are all intricately linked with one
another and it is not easy to establish in which system a fundamental
lesion takes place, in that an observed effect in one system may actually
be a response to an effect in another. This is clearly a rapidly developing
area and in the next five or six years considerable advances will be made
in identifying and understanding the mechanisms involved in the toxicity
of pollutants.
At low intakes the balance between accumulation in tissues, metabolism, and excretion is such that few if any adverse effects are observed. As
exposure is increased it is envisaged that these controlling processes are
progressively overwhelmed, primary lesions appear which in turn lead to
major physiological damage and ultimately death of an organism.
Internal pathways of the major phytotoxic gases are short, i.e. they
exert their effect near the point of entry. For mesophyll tissues of plant
leaves, for example, much of the damage due to sulfur dioxide involves
disruption of chloroplasts and depression of photosynthesis along with
changes in stomatal aperture. It is widely believed that much of the effect
of ozone involves permeability changes in the membranes of palisade
cells. Pollutants like ozone are so transitory that it is virtually impossible
to detect them within tissues; their presence is detected from the effects
induced. Exposure to sulfur dioxide can increase plant sulfur content
three- to four-fold; it is rapidly incorporated into organic molecules,
notably the amino acids, glutathione and cysteine. In fact the balance
between incoming sulfur dioxide and its destruction is probably a critical
factor in resistance to acute injury in different varieties and species of
plants. Nitrogen dioxide is similarly incorporated into amino acids such
as glutamine and asparagine. The incorporation of SO2 and NO2 into
metabolic pathways in this way explains why these pollutants can
stimulate growth of plants where S and N are limiting nutrients.
For many metals, storage or accumulation within a tissue or an organ
is essentially a detoxication step. Inorganic lead is transported in the
bloodstream attached to red blood cells (erythrocytes) but a major
proportion, about 90% in adults and 70% in children, accumulates in
bone. The biological half-life in this tissue is about 10 years but turnover
of lead in the bloodstream and soft tissues is rapid and responds quite
quickly to changes in lead intake and exposure.
In normal healthy humans about 50% of the cadmium in the body is in
the kidneys and liver with about one-third in the kidneys alone. The
biological half-life in these tissues is probably more than 10 years and as
a consequence cadmium accumulates with age. A significant proportion
occurs bound to metallothioneins, which are low molecular weight

proteins rich in sulfhydryl groups. In this bound form it is believed that
cadmium is less available to exert its toxic action. A number of other
trace metals, including zinc, copper, mercury, and silver also induce and
bind to metallothioneins. The precise role of metallothioneins is unclear
and the regulation of intracellular concentrations of metals also involves
other mechanisms; for example, metals may be incorporated into
extracellular structures such as carbonate granules, or bound to intracellular components such as nuclei, mitochondria, lysosomes, and phosphate granules.
DDT, PCBs, and other chlorinated organic compounds because of
their lipophilic properties, are concentrated in the fatty tissues of animals
and those with the largest amount of fat accumulate the highest
concentrations. In normal circumstances there is a slow turnover of
these compounds in the body's fat reserves. However, during periods of
stress and starvation mobilization of the fat deposits can release the
stored organochlorines into the bloodstream which in turn may induce
toxic effects and ultimately death. Laboratory studies have shown that
starved animals exposed to organochlorines such as DDT are critically
affected at lower exposures than well fed ones. A large number of
guillemot deaths were recorded in September 1969 from the coasts
around the Irish Sea and it is believed that they were due to a
combination of stress and mobilization of organochlorine residues,
perhaps PCBs, into the blood stream.1
A feature of most, if not all, vertebrate and invertebrate animals is a
built-in capacity to metabolize a wide range of foreign compounds to less
toxic entities that are more easily eliminated from the body. In higher
organisms metabolism mainly takes place in hepatic tissues and it
essentially consists of two phases; phase 1 consists of the monooxygenase
system which is responsible for metabolizing foreign compounds by such
mechanisms as hydroxylation, dealkylation and epoxidation; phase two
involves conjugation reactions in which glucuronide and sulfate derivatives are formed from the oxidized products of phase one. The net effect
is to convert lipophilic foreign compounds to water soluble metabolites
which facilitates their elimination from the animal. Numerous hydrocarbon compounds, including various phenolic and benzene compounds,
are known to undergo biotransformation to water soluble conjugates.
The influence of metabolism in regulating tissue concentrations and
hence on the toxicity of a substance can be deduced from studies that
have used inhibitors of the biotransformation pathways. Salicylamide is
a potent inhibitor of the phase II glucuronide conjugation reactions and
it has been shown that pretreatment of fish with salicylamide significantly increases the toxicity of phenols as a consequence of preventing
the production of the conjugated metabolite, phenylglucuronide.

Organophosphorus insecticides on the other hand need to be metabolized to become active, so inhibiting this step results in the pesticide being
less toxic. In animal tissues DDT is rapidly converted to the more
persistent DDE and this explains why DDE, rather than DDT, is the
most widespread and abundant form in wildlife.9
A fundamental relationship in pollution and ecotoxicology studies is
that which relates exposure, body burden or accumulation, and toxicity
of a substance. Residues in an organism (or some part of it) form an
important link between exposure and biological response or damage.
For the more persistent pollutants, such as metals and organochlorine
compounds, enhanced exposure from food or the surrounding media
generally results in a greater concentration in an organism. When uptake
(absorption) of a substance exceeds elimination (including metabolism)
accumulation occurs in the whole organism or some part of it. Eventually a steady state may be attained when uptake and elimination are
equal.
For many organic chemicals, the bioconcentration factor (i.e. the ratio
of the concentration of the chemical in the organism to the exposure) is
proportional to the lipid solubility and inversely related to the water
solubility of the compounds; it can therefore be estimated from the noctanol-water partition coefficient.
Many studies have examined the relationship between blood lead
concentration and concentration of lead in air, water and diet. A
curvilinear plot is obtained for the total intake versus blood lead
concentration, indicating that successive increments in intake or exposure result in progressively smaller contributions to blood lead concentrations. This relationship can form the basis of predicting one
parameter from the other. The accumulation and elimination rates of
methylmercury in humans conform to a single exponential first order
function (Figure 4). The interesting feature of this model is that,
assuming a constant hair-to-blood ratio, mercury analysis of segments
of hair can be used to predict both the body burden and blood
concentration of mercury at the time the hair segment was laid down.
Organisms in aquatic systems can acquire residues from both food and
the surrounding medium. From laboratory based studies comparing
equivalent exposures from both routes, uptake directly from the water
medium would seem to be more important. However, in reality lipophilic
substances partition to living and dead organic matter in the system and
the concentration in water would be very small, therefore exposure to
such substances would be mainly from ingesting contaminated food. In
9

F. Moriarty, 'Ecotoxicology: The Study of Pollutants in Ecosystems', Academic Press, London,
1998.

Body burden (ug/70kg) and blood (ng/100ml)

Exposure period

Figure 4

Body burden & blood

Hair

(mg/kg)

Hair

Number of half-times
The changes in the body burden and hair and blood concentration of mercury
during constant daily exposure (shaded area) and after exposure. This calculation was based on a daily intake of 10 fig of methylmercury during the exposure
period, an elimination half time of 96 days, and a hair-to-blood concentration of
250
(Reproduced with permission from WHO, Environmental Health Criteria,
'Mercury', 1976,1)

terrestrial ecosystems, most if not all lipophilic substances, including
pesticides, are transferred via the food chain. Seed corn dressed with
dieldrin undoubtedly killed many pigeons in England in the 1950s and
1960s; the poisoned pigeons were responsible for killing predators such
as foxes and perhaps peregrine falcons. It is common to find that residues
of organochlorine compounds are highest in carnivorous birds, followed
by insectivorous and then herbivorous birds. Pollution of the St.
Lawrence-Great Lakes basin with persistent organochlorine substances
first became apparent in the 1960s and it has been estimated that there is
a 100 million-fold bioaccumulation of PCBs between water, through the
food chain and into bald eagle's eggs.10

6 ECOLOGICAL RISK ASSESSMENT
Ecological risk assessment is concerned with estimating whether current
or predicted environmental levels of contaminants have the potential to
cause harmful effects on populations and communities. Such assessments
combine estimates of hazard and exposure. Hazard is defined in terms of
toxicity and is based mainly on single species tests that quantify
10

M. Gilbertson, Environ. Toxicol. Chem., 1997,16, 1771-1778.

responses to acute and chronic exposures. The general effects measured
are related to survival, growth, and reproductive performance and they
are expressed as LC50 and EC50 values (as discussed in Section 2). From
such tests the lowest observable effect concentration (LOEC) can be
estimated and at some concentration below this is the No Observed
Effect Concentration (NOEC).
Extrapolation of NOECs based on laboratory test organisms to
wildlife populations is a contentious issue. Tests are based on a few
species that are used as surrogates for wildlife, and they are carried out
under highly standardized conditions. Uncertainty factors are usually
built into estimates of NOEC and the magnitude of these depend on how
well effects have been characterized on species representative of the
ecosystem in question, ranging from as large as 1000 where tests are
restricted to a few species and with limited end-points, to as low as 10
where chronic toxicity is well established for test species representative of
the different levels of organization in the system and especially where
data are available for the most sensitive species. Hence NOEC values are
a synthesis of many toxicity tests with different organisms and a range of
end-points.
Exposure is measured or predicted from models as discussed in Section
3. The ratio of NOEC to the measured or predicted environmental
concentration (PEC) is taken to indicate the degree of potential hazard.
If the ratio is less than one, then environmental concentrations are
expected to equal or exceed the NOEC value and the chances of harmful
effects occurring in wildlife are high, but if it is less than one, or less than
the built-in uncertainty factor, then the likelihood of adverse effects is
low. Figure 5 highlights the concept, and also shows that whilst from
initial screening tests the ratio may be less than unity there is uncertainty
because the estimates are based on very few measurements and so more
tests with different species and additional end-points are required, and
perhaps more sophisticated experiments that closely simulate the
ambient environment. At the same time more concrete information is
required on the behaviour or fate of the toxicant in the actual environment. A stage is reached where a no-effect level can be defined with some
degree of confidence and it can be used as an objective for environmental
levels, either as an upper level for new chemical inputs or as a target to
aim for in reducing the levels of existing chemicals.
A related concept of critical concentrations and critical loads has been
developed for controlling the impacts of air pollutants on the environment. Critical loads refer to the total deposition of sulfur, nitrogen, or
hydrogen ions, and they are primarily aimed at preventing acidification
or eutrophication of sensitive aquatic or terrestrial ecosystems. Critical
levels refer to the direct effects of atmospheric pollutants on vegeta-

Confidence intervals

Concentration of chemical

No Observed Effect
Concentration (NOEC)

Margin
of safety

Expected Environmental
Concentration (EEC)

Time and increasing complexity of tests
Figure 5 Qualitative representation of the relationship between No Observed Effect
Concentration and Exposure with progressively increasing complexity of tests

tion.4'11 As shown in Figure 6 critical levels are the concentrations in the
atmosphere of pollutants above which direct adverse effects on receptors,
such as plants, ecosystems, or materials, may occur according to present
knowledge. Across a region such as Europe the number of receptors is
very large and therefore the emphasis is on determining exposureresponse relationships for the most sensitive receptor at a particular
location.
7 INDIVIDUALS, POPULATIONS, AND COMMUNITIES AND
THE ROLE OF BIOMARKERS
As a general rule, increasing pollution exposure results in progressively
more severe and damaging effects. Figure 7 provides a useful model
summarizing the interaction of pollution with receptors.1 The so-called
'cascading effect' suggests that some degree of homeostatic control or
carrying capacity operates at each level of organization. As exposure
increases in magnitude, the capacity of cells, tissues, whole organisms,
populations, and communities are successively overwhelmed and
11

K. R. Bull, Environ. Pollut., 1991, 69, 105-123.

Effects %

Critical
load

Deposition (load)
of concentration (level)

Figure 6

Theoretical dose-response curve showing a threshold, or 'critical load' at which
effects are observed
(Reproduced with permission from Environ. Pollut., 1991, 69, 105-123)

damaged. Disruption at the biochemical level may be quite specific but at
higher exposures, as physiological systems are affected, more general
symptoms of injury become apparent; reproductive capacity and recruitment may be affected and behavioural responses may be evident. At still
higher exposures fatalities start to occur and as these become as
significant as losses due to natural causes, noticeable reductions in
population size result. The cascade model is generally accepted;
however, in practice the relationships between the different levels of
organization have yet to be firmly established. Moreover, the model is
useful for explaining the role of biomarkers and bioindicators.
A biomarker can be defined as 'a biological response that can be
related to an exposure to, or toxic effect of, an environmental chemical or
chemicals'.12 The biological response can range from one at the biomolecular level to one that indicates changes in ecosystems. Most examples
come from the lower levels of organization. Their potential lies in their
ability to signal the onset of adverse exposures or effects at a higher level
of organization. Biomarkers can be categorized according to their
specificity. Only one or two biomarkers are sufficiently well developed
or understood to be used for quantifying exposure and or toxicity. The
inhibition of acetylcholinesterase by organophosphate and carbamate
insecticides is specific and reproducible. The inhibition leads to disrup12

D. B. Peakall, Toxicol. EcotoxicoL News, 1994, 1, 55-60.

Biochemical effect

Detoxification
biochemical repair
excretion
Physiological effect

Compensation

Population and
species
decline

Recruitment

Ecological
adjustment
Replacement by
other species

Figure 7

Ecological
disruption

Crude model of 'cascading' effect of pollution, as a series of reservoirs. Effects do
not spill over from one level to another unless inputs exceed capacity of relieving
systems depicted on left of each reservoir. Variation between individuals can be
thought of in terms of variation in capacity of these systems, which may also be
increased by evolution, producing more resistant organisms
(Reproduced with permission from M. W. Holdgate, 4A Perspective of
Environmental Pollution', Cambridge University Press, Cambridge, 1979)

tion of the nervous system; 20% inhibition has been used as a criterion of
exposure and 60% inhibition as a criterion of death caused by these
compounds. The monooxygenase system represents a broader category
and the induction of these oxidases has been related to exposures to
dioxins, PCBs, and PAHs in both field and laboratory studies (an
example is given below).
Case Study 1. Scope for growth in mussels, an example of the practical use
of a biomarker in the assessment of biological effects. The common

mussel, Mytilus edulis, has been used since the mid-1970s as a 'sentinel'
organism in 'Mussel Watch' monitoring programmes to assess spatial
and temporal trends of chemical contamination in estuarine and coastal
environments.13 The mussel can also be used as a means of demonstrating whether contaminant levels measured in their tissues are high enough
to have adverse effects. This is done by measuring the physiological
response termed 'scope for growth' (SFG) and it is one of the most
sensitive measures of pollution-induced stress. It is an integrated physiological response which estimates the energy available for activity,
growth, and reproduction from assimilation of food, after the energy
expended in respiration and excretion have been taken into account.
Exposure to a pollutant can cause a decrease in the energy available and
it may decline to negative values at high exposures as the animal draws
on the body's reserves. Feeding rate or clearance rate, defined as the
volume of water cleared of algal food particles per hour, is generally the
component of SFG most sensitive to pollutant stress.
One of the largest monitoring programmes to have been undertaken in
which chemical contamination and biological effects were measured at
the same time involved combined measurements of SFG and chemical
contaminants in the tissues of mussels at locations covering 1000 km in
the North Sea.14 In all, 26 coastal locations from the Shetland Islands to
the Thames estuary and 8 offshore light vessels were monitored. SFG
values were higher at the northern locations than at the southern ones,
which reflected the clean water condition of the inflow of the North
Atlantic into the North Sea and the increased contamination of water
associated with the urban and industrial locations further south. Coastal
regions of the Humber-Wash and the Thames estuary recorded some of
the lowest values for SFG (Figure 8). For several of the contaminants
detected in the mussel, experimentally derived tissue concentrationresponse relationships were available and these were used as a basis for
estimating the additive contribution of the different contaminants to the
overall decline in SFG. It was found that toxic hydrocarbons (mainly
PAHs) accounted for most of the reduced SFG at the majority of sites,
whilst tributyltin made a significant contribution to the toxic effect at a
number of locations. Metal concentrations were generally below those
associated with a reduction in SFG. This example demonstrates the
practical use of biomarkers in the assessment of the biological effects of
pollution and in the detection of the toxic agents concerned with these
effects. Other examples of biomarkers such as eggshell thinning in birds
caused by exposure to organochlorine pesticides and vitellogenin induc13
14

J. Widdows and P. Donkin, 'The Mussel Mytilus\ Elsevier, Amsterdam, 1992, Ch. 8.
J. Widdows et ai, Mar. Ecol. Prog. Ser., 1995, 127, 131-148.

Sites (N to S)
Qlust Voe
Voxter Voe
Orkney
Ythan
Montrose
Lucky Beacon
Musselburgh
Berwick upon Tweed
HoIv Island
Coquet Estuary
Cress well
Blyth
Trow Rocks
Tees mouth
Whitby Harbour
Fi ley Brigg
Humber Bull Fort
Cleethorpes
Wash
Hunatanton
Walberawick
Crabknowe Spit
Creekaea
Southend

Scope for growth (J g

h* f )

Mean i 95% CJ.

Swale

Whitatable
July 1990

Ranking
Figure 8

High stress

Mod. stress

Low stress

Mytilus edulis. Scope for growth of mussels collected from sites along the UK
North Sea coastline (mean ±95% CI, n = 16)
(Reproduced with permission from Mar. Ecol Prog. Ser., 1995,127, 131-148)

tion in relation to effects on the endocrine system in fish are discussed
below.
An important aspect of pollution studies is the diverse response of
organisms to equivalent exposures. Variation in sensitivity is apparent
between:
(1) Individuals;
(2) Groups or populations differentiated with regard to sex, stage of
development, nutritional well-being, etc;
(3) Varieties of strains;
(4) Species.
In general, differences between individuals are much less evident than
those between different species. Differences in sensitivity to pollution
may be viewed broadly in terms of differences in homeostatic capacities
between various groups of organisms i.e. differences in rate of absorption, metabolism, storage and excretion.
(1) In a set of individuals a measured response will reveal a group
which is sensitive, another which is moderately sensitive and one

that is relatively resilient; a spread or statistical distribution about
a median value is usually apparent and this is the basis for the
response of experimental populations in toxicity tests. The rate of
elimination of methylmercury from the human body is subject to
individual variation and a person with a slow elimination rate
would accumulate a higher body burden than another with a more
rapid rate.
(2) Very often the early stages of life are particularly vulnerable to
pollution stress. For example, children are known to be more
susceptible than adults to lead poisoning, and depletion of fish
stocks in acidified lakes is primarily due to high mortality among
eggs, larvae, and fry. At the other end of the scale, elderly people,
and especially those with a history of heart and lung disease, are
the most vulnerable during air pollution episodes. In humans
receiving similar lead exposure, higher blood lead concentrations
are generally found in males than females.
(3) Certain varieties of tobacco {Nicotiana tabacum) show different
degrees of sensitivity to photochemical oxidants. Three in particular, which are described as supersensitive, resistant, and intermediate, have proved useful biological indicators of elevated
ozone levels.
Prolonged exposure to certain pollutants can generate resistant or
tolerant varieties. This is especially true of fast breeding forms of life.
Examples include:
(a) Genotypes tolerant of one or more trace metals {e.g. Cu, Zn,
Pb, Ni) have been identified in many plant species growing on
metal-rich substrates. Such substrates include soils of mineralized areas, metal mine wastes, and soils contaminated by
metal-rich fumes originating from smelters, refineries, and
automobiles.15
(b) A genotype of the grass Lolium perenne, known as Helmshore
(named after the district 15 miles north of Manchester where it
was first discovered), has a higher resistance to sulfur dioxide
than the cultivated variety, S23.16
(c) An enormous number of insect pests have developed resistance to various insecticides and in the late 1970s resistant
15
16

M.H. Martin and P. J. Coughtrey, 'Biological Monitoring of Heavy Metal Pollution - Land and
Air', Applied Science, London and New York, 1982.
J. N. B. Bell, M. R. Ashmore, and G. B. Wilson, 'Ecological Genetics and Air Pollution', ed. G. E.
Taylor et at., Springer-Verlag, New York, 1991.

varieties had been recorded in over 400 species of insects and
mites.9
(d) In the lead-, copper- and zinc-rich sediments of old metal mine
workings in estuaries of south-west England, tolerant varieties
of the polychaete Nereis diversicolor and asellus among
others, can be found.17
(e) Darker, melanic forms of many species of the larger moths,
e.g. Biston betularia, occurred with increasing frequency in
areas within and adjacent to the industrial regions of Britain.
The distinctive feature of these darker varieties is the greater
preponderance of melanic granules in the cuticle and wing
scales. The significance of the adaptation would seem to lie in
concealment from bird predation on darker surfaces which
are typical of areas affected by air pollution.9
(4) Species, even closely related ones, can differ quite markedly in
their response to one form of pollution or another. This is an
important consideration in the assessment of the ecological effects
of pollution since it is common to find that communities that are
subject to pollution stress undergo a reduction of species diversity
as the most susceptible ones disappear first, to leave behind fewer
and more resilient types. Some notable examples include forest
ecosystems affected by air pollution and freshwater systems
subject to biodegradable wastes, artificial eutrophication, or
acidification. In certain situations changes in community structure
can be used as a sentinel of the degree of pollution stress, e.g.
biotic indices that are used to indicate water quality.
Bryophytes (mosses and liverworts) and lichens (fungal-algal symbioses)
are among the most sensitive receptor organisms to several atmospheric
pollutants. Corticolous lichens (i.e. those growing on tree trunks) along a
gradient of sulfur dioxide as say from a rural area to an industrial centre,
reveals that the lichens become progressively impoverished as higher
sulfur dioxide concentrations are encountered and commonly within
urban-industrial centres there is a complete absence of lichens. Hawksworth and Rose18 devised a scale composed of 11 zones based on
presence and abundance of corticolous lichens, and a map of England
and Wales depicting these lichen zones shows them to be closely related
to the winter mean sulfur dioxide concentration for the particular year of
the survey (1972-73).
17

G. W. Bryan, 'Pollution and Physiology of Marine Organisms', ed. W. B. Vernberg and F. J.
Vernberg, Academic Press, XXX, 1974, p. 123.
18
D. L. Hawksworth and F. Rose, Nature, 1970, 227, 145.

Frequency distribution of ozone
resistance in 61 native species
grasses
clover

% of species

BeIW
tobacco

High

Figure 9

Ozone resistance R%

Low

Frequency distribution of the ozone resistance of 61 native species, compared
with that of the sensitive bioindicator, Nicotiana tabacum cv. Bel-W3. Ozone
resistance is expressed as growth in ozone as a percentage of that in filtered air

(Data provided by Prof. A. W. Davison)
(Reproduced with permission from United Kingdom Critical Loads Advisory
Group, 'Critical Levels of Air Pollutants for the United Kingdom', Department of the Environment, London, 1996)

It is increasingly recognized that the most important impact of O3 on
semi-natural plant communities is through shifts in species composition,
loss of biodiversity, and changes in genetic composition. Figure 9 shows
the frequency distribution of resistance to O3 of 61 plant species native to
Derbyshire, together with the resistance of the tobacco variety, Bel W3.
All the species were subjected to the same O3 exposure over a period of
two weeks, and whilst between 25 and 45% of them showed no
detectable effect, over 25% of species showed a marked reduction in
growth and they were found to have a similar sensitivity to the tobacco
cultivar.19 A similar experiment in which populations of the broad leaved
plantain {Plantago major) collected from different geographical areas of
the UK were exposed to O3 demonstrated that there was as much
variation in resistance within this one species as there was between the
61 species. The O3 resistance of the populations was significantly
correlated with the O3 concentrations prevailing in the areas from
which they were collected. The most resistant populations were from
southern England, indicating that these populations had evolved resistance to O3.20
In some cases, differences in species sensitivity may be more apparent
than real because certain organisms may experience greater exposure as a
consequence of pollutants becoming more concentrated in specific parts
19
20

K. Reiling and A. W. Davison, New Phytol., 1992, 109, 1-20.
K. Reiling and A. W. Davison, New Phytol., 1992, 122, 699-708.

of an ecosystem such as the upper soil layers or the sediments at the
bottom of lakes and estuaries. Organisms which inhabit these areas and
in particular those that do so by virtue of their feeding habits, e.g.
earthworms or filter feeds such as mussels, will experience greater
exposures from ingesting contaminated particles.
Fatalities due to pollution do not necessarily affect the overall size of a
population. Organisms die from a variety of natural causes, and unless
pollution is on a par with these its influence on population size is
unlikely to be significant. In broad terms a population is controlled by
birth rate, death rate, and the balance between immigration and
emigration. For many wildlife populations, recruitment (i.e. the
number of 'recruits' produced annually) has a strong influence on
population size. This is certainly the case for most commercially
important fish species in which fluctuations of the early life stages
account for much of the fluctuation in overall population size, and
environmental factors, rather than spawning stock, control the number
of early larval and juvenile stages.
Natural populations of animals and plants are subject not only to
short term changes but also long term fluctuations of population size
which are often large-scale and irregular events that may take place
rapidly. These changes are often due to changes in weather patterns such
as periods of lower or higher temperatures. Clearly for disturbances
caused by pollution to be apparent in a population, they must be on a
scale over and above fluctuations due to natural causes. In addition,
other human activities, e.g. overexploitation of a natural resource, can
cause significant perturbations.
Whether or not pollution exerts a permanent effect on the population
size will depends on the persistence of the polluting agent. Once the stress
has been withdrawn, populations have a remarkable ability to recover
from large-scale fluctuations, due to whatever cause. The rapid recolonization of the tidal river Thames is a striking example of recovery of an
ecosystem from pollution. However, to conclude that recovery is
complete assumes a precise knowledge of the situation prior to pollution
onslaught and this is rarely known.
In general terms, the ecological effects of pollution are reductions in
species diversity, productivity, and biomass. These trends have been
described as a retrogression to an earlier successional stage of ecological
development. However, perception and evaluation of such changes at a
community level is difficult, since even the simplest of ecosystems contain
a considerable number of species, population interactions are complicated, and measurement of pollutant exposure in the various niches of
the system can be very problematic.

8 HEALTH EFFECTS OF THE MAJOR AIR POLLUTANTS
Health effects of air pollution in the general population are associated
with an increase in mortality and worsening of lung and heart conditions
during pollution episodes. The potential of smoke laden air from coal
fires to increase incidences of respiratory disease in London was
recognized as long ago as 1661 by Evelyn.21
Case Study 2. The London Smog 1952. The most notorious incident
occurred in London in early December 1952, when 3500-4000 deaths
above the norm were recorded during an exceptional smog episode that
lasted unremittingly for five days.22'23 At the time, much of the air
pollution was due to coal combustion, with numerous near-ground
sources generating considerable quantities of smoke and sulfur dioxide.
The fog developed across the capital as a consequence of a very stable
high pressure zone and an inversion layer. This prevented the dispersal of
pollutants, and concentrations built up to very high levels, the particles
of smoke acting as condensation nuclei which added further to the
density of the fog. The epidemic went largely unnoticed by the population of London and it was only when death certificates for the whole of
the London area were later examined that the sudden upsurge in the
number of deaths became apparent. Deaths were mainly confined to the
elderly and those people with a history of heart and lung diseases. The
central areas of London, where the fog was at its densest and most
persistent, had the greatest increases in mortality, some 200% more than
the average for that time of year. Estimates indicate that sulfur dioxide
and smoke concentrations (48 h mean) during the London episode
attained very high values and were in the region of 3.7 mg m~ 3
(1.3 ppm) for SO2 and above 4.5 mg m~ 3 for smoke. The episode was
also at the time of the Smithfield Agricultural Show at Earls Court and it
is interesting to note that a number of cows also died during the course of
the show. The London smog episode was very important in establishing
the strong link between air pollution and respiratory health.
It is generally accepted that the combined effects of sulfur dioxide and
smoke particles on the respiratory tract were responsible for the excess
deaths and for exacerbating heart and lung disease in susceptible people.
To the healthy majority of Londoners the pollutant laden fog was merely
an inconvenience. Epidemiological studies established that the spatial
21

J. Evelyn, 1661, reprinted in 'The Smoke o f London: T w o Prophecies', ed. J. P. Lodge, Maxwell
Reprint, N e w York, 1969.
Ministry of Health, 'Report on Public Health and Medical Subjects N o 95', H M S O , London,
1954.
23
'Environmental Health Criteria, 8, Sulphur Oxides and Suspended Particulate Matter', World
Health Organization, Geneva, 1979.

and temporal trends of sulfur dioxide and smoke correlated closely with
the distribution of the enhanced mortalities, but other pollutants would
also have been elevated, e.g. carbon monoxide, sulfates, and sulfites.
However there are insufficient data on such substances to allow exposure-response relationships to be established.
Follow-up studies in other large cities concentrated on more moderate
day-to-day variations of mortality and morbidity in relation to pollution
levels. A WHO Task Group provided a summary23 of the salient features
of many of these studies and the collated data have formed the basis for
developing short and long term exposure-response relationships. It is
important to emphasise that concentrations of sulfur dioxide and
suspended particulate matter vary considerably from place to place and
from one time period to the next. Also, most monitoring data relate to
levels prevailing in the outdoor environment and take little account of
indoor exposure, which is usually lower but is where most people spend
much of their time. The elderly and the chronically sick spend most of
their time indoors. Also there is no consideration of occupational
exposure, which can be significant, and smoking which is the most
direct form of air pollution. Notwithstanding, short term exposures of
500 /ig m~ 3 (24 h mean) for both sulfur dioxide and smoke can be
expected to result in excess mortality among the elderly and the
chronically sick. Exposures of 250 /ig m~ 3 (24 h mean) to both pollutants are likely to lead to worsening of the condition of patients with
existing respiratory disease. With regard to long term exposure,
increased prevalence of respiratory symptoms among both adults and
children and increased frequency of acute respiratory illnesses in children
are more likely to occur when annual mean concentrations of sulfur
dioxide and smoke exceed 100 fig m~ 3 . Based on these relationships, and
incorporating a margin of safety, WHO advocate, as a guideline, that
24 h mean values of sulfur dioxide and smoke should remain below 100150 fig m~~3, with an annual mean below 40-60 /ig m~ 3 .
The relationship between photochemical smog episodes and human
health is less clear cut. Early epidemiological studies, based mainly on
Los Angeles, failed to establish direct relationships between increased
mortality rates and the frequency of smog episodes, although eye
irritations (perhaps due to peroxyacyl nitrate compounds) and upper
respiratory tract discomfort are common complaints during smog
episodes. Increased breathing difficulties in heavy smokers and asthmatics and reduced performance in people indulging in physical activity
are associated with periods of high oxidant concentrations.24
Chamber studies with short term but realistic exposures to O3 and
24

'Environmental Health Criteria, 9, Photochemical Oxidants', World Health Organization,
Geneva, 1979.

healthy adult subjects have consistently shown effects such as reductions
in lung function (lung volume and expiratory flow-rate), increases in lung
reactivity to other irritants, and pulmonary inflammation.
Field studies of adults who exercise heavily for short periods have
shown short term reversible decreases in pulmonary function in
association with ozone concentrations at or near the previous US
National Ambient Air Quality Standard of 120 ppb. Much of the field
evidence comes from children attending summer camps in North
America where ambient O3 concentrations, particularly above 120 ppb
have been associated with short term declines in the average lung
function.25'26 However, other studies have failed to reproduce this
relationship. A recent study re-analysed the data from six summer
camp studies on lung function and O3 exposures. They were all based
in North America with two in north-west New Jersey, two in southern
California and two in Ontario. Data were available for forced
expiratory volume in 1 second (FEV1) and peak expiratory flow rate
(PEFR) as well as O3 exposure. Statistically significant positive
relationships were found between FEV1 and O3 exposure but those
with PEFR were inconsistent.27 The 1997 revised US ambient air
quality standard for ozone is 80 ppb daily maximum eight hour
average over three years.
In London, daily hospital admissions for respiratory disease over a
five year period (1987-92) were examined in relation to air pollution
(NO2, SO2, O3, and smoke) and it was found that O3 above all the others
had a small yet significant effect on admissions.28
One of the key issues at the present time is whether the lower
concentrations of respirable particles encountered in towns and cities
today are having an effect on respiratory health, and much has been
made of the possibility that they are responsible for the recent increase in
asthma among children.
The primary source of particulate matter in towns and cities is traffic
and so with the ever increasing volume of traffic respiratory health has to
be assessed in relation to this source. Traffic-related air pollution,
however, is not restricted to particulate matter, the other main pollutants
being nitrogen dioxide, ozone, carbon monoxide, lead and hydrocarbons.
Over the last seven years or so, a number of studies have demonstrated
an association between elevated PM10 concentrations and hospital
25

M . Berry, P. J. Lioy, K. Gelperin, G. Buckler, and J. Klotz, Environ. Res., 1991, 54, 135.
D . M. Spektor, G. D. Thurston, J. Mao, D. He, C. Hayes, and M. Lippmann, Environ. Res., 1991,
55, 107.
27
P. L. Kinney, G. D. Thurston, and M. Raizenne, Environ. Health Perspect., 1996, 104, 170-174.
28
H . R. A. Ponce de Leon, R. Anderson, J. M. Bland, D. P. Strachan, and J. Bower, J. Epidemiol.
Comm. Health, 1996, 50, S63-S70.
26

admissions for respiratory disease.29 In particular, admissions for
chronic obstructive pulmonary disease (COPD) shows the largest
increase during periods of high PM10 concentrations (> 50 /ig m~ 3 ).
In one localized study, centred on a steel mill in Utah, it was shown
that hospital admissions of children with respiratory disease in winter in
the Utah Valley were associated with emissions of particulates from the
nearby steel mill. In one particular winter a strike by the workforce
resulted in much reduced PM10 concentrations and fewer hospital
admissions.
Other time-series and cross-sectional studies have shown similar
associations between airborne particulate matter concentrations and
respiratory disease and mortality. The association is strongly linked to
respirable particles regardless of weather conditions (winter or summer
pollution episodes), confounding influences of other pollutants such as
SO2 and O3, different geographical areas, and different types of sources
of particulate matter (e.g. wood smoke versus traffic emissions). Estimates indicate that for a 100 //g m~ 3 rise in PMi0 the relative risk factor
of increased admission to hospital with respiratory disease is of the order
of 1.1-1.3; although this is a relative small degree of risk, in large cities it
does mean that tens of thousands are likely to be affected.29
There is therefore strong evidence that current levels of air pollution in
the towns and cities of Europe and North America are having an effect
on the respiratory and perhaps the cardiovascular health of the vulnerable and those with a history of these diseases. The balance of evidence
suggests that respirable particulate matter is the major cause, although
some recent studies have suggested that the effects can also be related to
O3.
The possible role of air pollution in the increased incidence and
prevalence of asthma in countries such as Europe, North America,
Australia, and New Zealand is controversial and far from clear. It has to
be stressed that while asthma prevalence and hospital admissions for this
disease have shown a dramatic increase in the last two decades in the USA
and elsewhere, the trends in the major air pollutants have been distinctly
downwards. There is no obvious correlation between air pollution and
asthma hospital admissions in the UK where the prevalence of childhood
asthma has increased by about 50% over the past 20-30 years. During
this period SO2 and particulate matter concentrations have declined
markedly. In Britain and elsewhere there is little correspondence
between the geographical distributions of childhood asthma and air
pollution, and in particular there appears to be no difference in the
prevalence of the disease between children living in rural and urban areas.
29

J. Schwartz, 'Health at the Crossroads: Transport Policy and Urban Health', ed. T. Fletcher and
A. J. McMichael, John Wiley & Sons, Chichester, 1997, Ch. 2.

9 EFFECT OF AIR POLLUTION ON PLANTS
The main pollutants which affect vegetation are SO2, NOx, O3, fluorides,
and ethylene. The effects on vegetation include reduction in yields, visible
leaf damage, loss of sensitive species and therefore reduction in diversity
of plant communities, and changes in sensitivity to other stresses. In the
UK, concentrations of SO2 have decreased in recent decades, NO x
concentrations are elevated but steady, and there would appear to be
increased incidences of photochemical pollution producing elevated O3
concentrations. Fluorides and ethylene are more localized pollutants.
Oxides of nitrogen and sulfur are ultimately metabolized by the endogenous nitrogen and sulfur biochemical pathways and thus can contribute to the normal N and S metabolism in plants. At relatively low
exposures ecosystems absorb or act as a sink to the influx of pollutants
and little or no harm can be detected. Indeed, there are reports that
growth can be stimulated in N and S deficient soils.
In the past the main concern was focused on visible damage and
impaired growth due to SO2 (and acid aerosols) and smoke around point
sources and in cities. One of the earliest works in urban areas was in the
early twentieth century in Leeds and a nearby industrial area where air
pollution was shown to cause reduction in yields of lettuces by up to 70%
and extensive damage to evergreen privet during autumn and winter
months. These changes were correlated with soot and sulfite deposits.
Certain conifer species are now known to be especially sensitive to sulfur
dioxide and other pollutant gases. For a number of years, attempts to
establish conifer plantations in southern parts of the Pennine Hills in the
North of England generally proved unsuccessful. At one stage Scots Pine
(Pinus sylvestris) was either absent or very sparse over a 50 km wide
corridor, downwind of the major conurbations of Greater Manchester
and Mersey side. The frequency of the species inversely corresponded to
mean winter sulfur dioxide concentrations in the area, and no trees
occurred wherever SO2 concentrations exceeded 0.076 ppm. Indigenous
grassland has also been shown to be tolerant of SO2 pollution.16 With the
decline of sulfur dioxide concentrations there has been marked improvements in conifer plantation in the area.30 General improvement in urban
air quality can be seen in a number of areas of the UK from the gradual
recolonization by lichens on the bark of trees, including in parts of
London.31
Growth chamber experiments which subject plants to unfiltered
(ambient) air and filtered (control) air provide strong supportive evi30

J. N. B. Bell, 'Gaseous Air Pollutants and Plant Metabolism', ed. M. J. Koziol and F. R. Whatley,
Butterworth, London, 1984, Ch. 1.
31
C. I. Rose and D. L. Hawksworth, Nature, 1981, 289, 5795, 289-292.

dence of urban air pollution damage to vegetation. One of the earliest
studies (1950-51), examined the effect of pollutants in the air of a
Manchester suburb on the growth of perennial ryegrass, Lolium
perenne. Significant reductions in growth were found in the grass
exposed to the ambient air which had a mean SO2 concentration of
0.07 ppm. Later investigations in the Sheffield area recorded reduced
yields in plants exposed to much lower mean SO2 concentrations. In
other studies using more realistic open-top chamber designs, growth
reductions only occurred at higher SO2 exposures. The differences may
be due to the different design of the chambers, although other factors
may be involved. For instance, chronic SO2 injury can be enhanced if
plants are exposed to SO2 during winter, when they are growing more
slowly. Furthermore in these studies, only SO2 was measured and it is
now evident that other phytotoxic pollutants would have been also
present, e.g. NO* and O3. The Sheffield studies coincided with some of
the highest O3 concentrations yet recorded in Britain.30
Vegetation injury caused by photochemical smog was first reported in
the Los Angeles basin in 1944 and it has continued to be a chronic
problem in southern California. It was established that O3 was the main
phytotoxic agent in the smog complex. Ozone has caused widespread
injury to agronomic and horticultural crops and natural and managed
forest ecosystems, not only in California but also in many other states
where meteorological conditions and primary pollution concentrations
were favourable. Ozone has been the most economically damaging air
pollutant to vegetation in the USA. Large scale injury to tobacco crops
in the eastern USA has been reported over a number of years. Extensive
pine needle damage to Ponderosa and Jeffrey pines in western locations
and white pine in the east were due to atmospheric oxidants. In the mixed
conifer forest ecosystems in the mountains of Southern California, the
dominant tree species, Ponderosa and Jeffrey pines, for many years,
suffered annual mortalities of about 3%, with the result that hundreds of
thousands of trees have died. In the 1970s it was estimated that in the San
Bernardino forest 46 230 acres had suffered severe ozone-type injury,
53 920 acres moderate injury, and 60 800 acres light or no injury.
Evidence that ambient ozone concentrations in Britain during the
summer can inflict deleterious effects on vegetation comes from a series
of experiments, using open-top chambers, at a rural site near Ascot in
Berkshire, some 32 km west of London. At this location SO2 and NO x
concentrations are generally low, but episodes of high ozone concentrations have been regularly recorded. During three such incidents between
1978 and 1983, Pisum sativum, Trifolium repens, and T. pratense, grown
in open-top chambers and receiving unfiltered air, developed visible leaf
necrosis typical of O3 damage. Concentrations of O3 during the course of

Relative Grain Yield [%]

AOT40 [ppm.h]
Figure 10 The relationship between relative grain yield of spring wheat and ozone
exposure expressed as AOT40for 3 months, based on data from eight European
open-top chamber experiments
(From Fuhrer, J. (1994) in: Fuhrer, J. and Ackermann, B., Critical levels for
ozone; a UN-ECE workshop report. FAC Report No. 16, Swiss Federal
Research Institute for Agricultural Chemistry and Environmental Hygiene,
Liebefeld-Bern)

these experiments exceeded 0.1 ppm, which is considered to be the
threshold above which visible leaf necrosis appears in these sensitive
plants. Field observations in the surrounding areas revealed symptoms
typical ofO 3 injury in a variety of crop plants. It has since been realized
that Pisum sativum crops grown in the UK for a number of years,
frequently developed necrotic lesions typical of ozone injury without the
cause being known.32 Field observations have established that many
commercially grown crops in Europe develop visible injury, such as
chlorosis or bronzing, due to ambient levels of O3 and that these
symptoms commonly develop in response to short term episodes rather
than longer term average concentrations.4
Studies have shown that increasing the concentration of O3 has a
greater impact than increasing the period of exposure and that levels
over and above a threshold concentration give a better fit to reductions in
growth. The European open-top chamber experiment (EOTC), conducted in 9 countries over a five year period, was a major international
research study on the effects of O3 over the complete live-cycle of several
crop species. Figure 10 shows the linear relationship between O3
exposure, expressed as the accumulated exposure over a threshold of
40 ppb (AOT40), and the relative yield of spring wheat. Similar relation32

M. R. Ashmore, Proceedings of an International Workshop on 'The Evaluation and Assessments
of the Effects of Photochemical Oxidants on Human Health, Agricultural Crops, Forestry,
Materials and Visibility', Swedish Environmental Research Institute, (IVL) Goteborg, 1984, p. 92.

ships have recently been established for tree species such as beech and
Norway spruce. Beech is more sensitive to O3 with an AOT40 value of
10000 ppb h associated with a 10% reduction in growth. Recent work
with semi-natural vegetation suggests that changes in species composition are more important than either growth reduction or visible injury. In
grassland communities a marked shift from forb to grass species has been
observed at comparable O3 exposures to those which produce yield
reduction in spring wheat.4
Other effects of air pollutants include damage to the epicuticular wax
layer of leaves and needles which can lead to increased water loss and
there are reports that air pollutants can delay the onset of winter
hardening. There is also evidence that air pollutants acting singly and in
combination (SO2, SO2/NO2 and O3) have an effect on the relative
distribution of growth above and below ground, with root growth being
more severely affected than shoot growth. This may be a consequence of
interference with phloem translocation and assimilation in the plant.
There are also reports that plants in polluted areas are more susceptible
to attack by insect pests and exposures of plants to small or medium
doses of SO2 and/or NO2 result in increases of the population growth of
aphids feeding on the plants. This in turn can result in significant
increases in pest damage to plants.8'33
The past two decades has seen a change in emphasis from the impact
of local sources of air pollution to effects across national boundaries.
Thus extensive areas of Europe and North America are subject to
elevated levels of gaseous air pollution and in such instances a polluted
air mass rarely contains a single phytotoxic agent but a complex gasaerosol mixture that includes varying proportions of SO2, nitrogen
oxides (NOx), ozone (O3), and acid aerosols. This has led to the
application of the critical level concept whereby each air pollutant for a
defined set of environmental conditions is given a limiting value, above
which available evidence predicts that adverse effects are likely to
become apparent. The receptors are classified according to the type of
vegetation—crops, forests and woodlands, and semi-natural vegetation.
Modifying factors such as winter stress, or nutrient deficient soils can be
taken into account. Taking semi-natural vegetation as one category, the
critical levels for SO2 and NO x are 20 and 30 /ig m~ 3 respectively and
that for O3 is expressed as 3000 ppb h (ppb O3 above a threshold of
40 ppb accumulated during daylight hours). In the UK it has been
estimated that whilst the land area of semi-natural vegetation where the
critical levels of SO2 and NO x are exceeded is in the region of 1 %, that of
33

'Air Pollution and Forest Ecosystems in the European Community', ed. M. R. Ashmore, J. N. B.
Bell, and I. J. Brown, Air Pollution Research Report 29, Commission of the European
Communities, 1990.

O3 is of the order of 70%, highlighting the current importance of this
pollutant.
These critical levels are based on individual pollutants and do not take
account of pollutant combinations. This is a very complicated subject, as
research has shown that responses of vegetation to pollutant mixtures is
not simply the sum of the responses to individual pollutants. For
example SO2 and NO2 can act synergistically and there is evidence that
NO2 and O3 may behave antagonistically. Thus the exposure-response
relationship for one pollutant may be modified by the presence of
another and the exposure at any one location and time is a complex
function of source strength, meteorology, and atmospheric chemistry.
10 ECOLOGICAL EFFECTS OF ACID DEPOSITION
Over large parts of western Europe and north-eastern North America
streams, rivers, and lakes have become progressively more acidic. It is
now generally accepted that acid deposition with major acidic, or
acidifying ions is the cause of freshwater acidification in geologically
sensitive areas. Such areas are those with slowly weathering granites
and gneiss rocks that support acidic and weakly buffered soils. Freshwater acidification is typified by a loss of acid neutralizing capacity
(essentially the carbonate buffering system), decrease in pH by as much
as 1-2 pH units, increases in sulfate, nitrate, ammonium ions, aqueous
aluminium, and metals such as manganese and zinc. Acidification is
responsible for the loss and depletion of fish populations from numerous freshwater ecosystems in parts of Norway, Sweden, UK, Canada,
and USA.
Freshwater organisms generally maintain their internal salt concentration by active uptake of ions (in particular Na + and Cl~) from water
against a concentration gradient. Substantial experimental evidence and
field data indicate that fish mortality at low environmental pHs is
primarily caused by a failure to regulate internal salt concentrations at
gill surfaces. A characteristic symptom of acid-stressed fish is therefore a
reduction in the salt or electrolyte ion concentration of blood plasma.34
At an early stage it was realized that mortalities were not simply a
function of acidity. Field data indicated that fish mortality occurred in
waters of pH 5, whereas laboratory experiments with purely acid
solutions failed to reproduce these results and significant mortalities
only occurred at pH 4. It was later shown that the dissolved concentration and chemical species of aluminium was the important factor
influencing toxicity in acidified rivers and lakes. This common element
34

H. Leivestad and I. P. Muniz, Nature, 1976, 259, 391.

in rocks is more soluble in acidic environments, and various surveys of
lakes and rivers have revealed that concentrations of aqueous aluminium
increase in acidified waters. Because of reactions with hydroxide ions,
aluminium toxicity to fish is pH dependent and the most bioavailable
forms of aluminium occur at pH 5. Several reports have shown the effects
of the interaction of acidity and soluble aluminium on fish survival. The
survival of brown trout fry is reduced in concentrations of aluminium of
250 fig l~l and at lower concentrations if calcium is low. Effects on
growth have been found at much lower concentrations, for example the
growth rate of brown trout is reduced at a concentration greater than
20 /Hg I" 1 , a concentration much lower than that found in acid waters.
The most toxic combinations were found with elevated aluminium
concentrations at pH 5. Increasing calcium concentrations from 0.5 to
1.0 mg I" 1 offset the effect of increasing aluminium and acidity on the
survival of brown trout. Sulfates, fluorides, and soluble organic matter in
acidified waters which form complexes with aluminium can also have a
significant ameliorating influence on aluminium toxicity.35
In summary, loss of fish populations can be expected in acidified clear
water lake ecosystems, low in organic matter and low in calcium. Such
environments tend to be characteristic of mountain areas and in southern Norway in particular these are the regions where the loss of fish
populations was first recorded.
The effects of acidification are not restricted to fish; all facets of
freshwater ecosystems are affected. The assemblages of all the major
groups, micro-organisms, plants and animals, alter considerably and
there is a general tendency for a reduction in species diversity at all
trophic levels. For many groups this is accompanied by a reduction in
productivity and a loss or decline of populations, e.g. several species of
Crustacea (snails and crayfish), amphibians and fish.
The following summary of the effects of acidification on the major
trophic levels in freshwater systems is based on a review of the subject by
Muniz.36
Decomposers. Accumulation of undecomposed and partly decomposed organic matter is taken to indicate reduced activity of bacteria,
fungi, and protozoans in acidified waters. Some field and experimental
studies have shown decreased rates of decomposition, but others have
found no evidence of any changes.
Primary Producers. Reductions in the numbers of species of several
35

D . J. A. Brown, 'Acid Deposition, Sources, Effects, and Controls', ed. J. W. S. Longhurst, British
Library Technical Communication, 1989, p. 107.
36
1 . P. Muniz, 'Acid Deposition—Its Nature and Impacts', ed. F. T. Last and B. Watling, Proc. Roy.
Soc. Edinburgh, 1991, Section B, 97, 228.

phytoplanktonic algae, especially the green algae (Chlorophyceae)
have been reported. Dinoflagellates dominate the plankton of many
acid lakes. Fewer species of periphyton occur on submerged surfaces
and these habitats are commonly superseded by mass encroachments
of filamentous algae (e.g. Mougeotia, Zygogonium, and Spirogyra). In
many lakes in Sweden macrophyte populations of Lobelia have
declined and have been invaded by Sphagnum species. Synoptic
surveys generally indicate that biomass and primary productivity are
less in acidified lakes, although experimentally acidified lakes have
shown no change and even increases in biomass.
Zooplankton. Species diversity decreases with increasing acidity,
particularly below pH 5.5-5.0 and this is usually followed by a
reduction in biomass. Acid sensitive species of Daphnia and Cyclops
frequently disappear below pH 5.
Benthic macroinvertebrates. Many species of mayflies, amphipods
(freshwater shrimps), crayfish, snails, and clams are very sensitive to
acid conditions. In Norway, snails were found to be absent in waters
below pH 5.2 while mussels disappeared at pH 4.7. Recruitment
failure and reduced growth are important causes of the elimination of
species. Crayfish decline has been due to the effects of acidity on the
eggs and juvenile stages, and the moulting stages are particularly
sensitive. A similar situation has been found for freshwater shrimps,
mayflies, and snails.
Fish. There are numerous examples for the loss of individual species
and changes in the composition of fish communities in acidified
waters. Minnows seem to be some of most sensitive fish and population declines often start below pH 6. An approximate order of
sensitivity to acid stress is as follows (beginning with most sensitive):
roach, minnow, Arctic char, trout, European cisco, perch, white pike
and eels are consistently the most tolerant species. The younger life
stages, particularly newly hatched or 6swim-up' fry just starting to
feed, are the most sensitive stages to acid stress. As a consequence
many acid-stressed populations are dominated by older fish due to
recruitment failure.
Amphibians. Reproductive failure is the main effect of acid stress on
amphibians. In Sweden, populations of the frog Rana temporaria have
been lost and reproductive failure of the toad Bufo bufo has occurred.
Birds. Studies have shown that the breeding density of dippers in
Wales (songbirds that feed almost exclusively on aquatic invertebrates)
has decreased as acidity has increased. This has been related to the
decrease in abundance of mayflies and caddis fly larvae in acidified
streams. In general, some species of songbirds that breed alongside
acidified lakes and streams produce smaller clutches with thinner

eggshells, and breed later with fewer second clutches. Evidence also
indicates that their young grow less rapidly and suffer more mortality.
Although the loss of decline of populations of animals and plants is
largely attributable to changes in water chemistry, indirect effects may
also cause additional stress. These essentially arise out of changes in
predator-prey relationships caused by a decline of food resources. For
example, instances of increases in phytoplankton biomass may be due to
a decrease in the performance of herbivores. The decline of the dipper is
strongly associated with the disappearance of its prey from acidified
streams.
Case Study 3. Lake Acidification; an example of a large scale field
experiment. Experimental acidification of entire lakes has yielded very
valuable information and increased understanding of the effects of
acidification on communities and ecosystem function. Lake 223, in the
Experimental Lakes Area of Canada, was acidified over an 8 year period
during which time the pH was gradually reduced from 6.8 to 5.0. One of
the surprising features of the study was that primary production and
decomposition showed no overall reduction. However, the composition
of the phytoplankton and zooplankton communities showed distinct
changes. At pH 5.9 several key organisms in the lake's food web were
severely affected; the opossum shrimps (Mysis relicta) declined from
almost 7 billion to only a few animals and fathead minnows (Pimephales
promelas) failed to reproduce. At pH 5.6 the exoskeleton of crayfish
(Oronectes virilis) hardened more slowly after moulting and remained
softer. The animals later became infected with a microsporozoan parasite.
When the acidification of Lake 223 was reversed, several biotic
components recovered quickly. Fish resumed reproduction at pHs
similar to those at which it failed on acidification. The condition of lake
trout improved as the small fish on which they depended for food
returned. Many species of insects and crustaceans returned as the pH
was raised,37'38
Diatom communities in lakes have provided valuable information on the
history of aquatic systems and in particular on their acidity. These
organisms have characteristic siliceous scales which are preserved in
lake sediments. Species of diatom can be classified according to their
37

D . W. Schindler, K. H. Mills, D. F. Malley, D. L. Findley, J. A. Shearer, I. J. Davis, M. A. Turner,
G. A. Linsey, and D. R. Cruikshank, Science, 1985, 228, 1395.
D . W. Schindler, T. M. Frost, K. H. Millerin, P. S. S. Chang, I. J. Davis, D. L. Findley, D. F.
Malley, J. A. Shearer, M. A. Turner, P. J. Garrison, C. J. Watras, K. Webster, J. M. Gunn, P. L.
Brezonik, and W. A. Swenson, 'Acid Deposition—Its Nature and Impacts', ed. F. T. Last and B.
Watling, Proc. Roy. Soc. Edinburgh, 1991, Section B, 97, 193.

relative tolerance of acid or alkaline conditions and assemblages of
different species can be used to indicate the pH of lake water. Palaeolimnological data from sediment cores have been used to construct the
pH histories of lakes and in relation to recent changes they have provided
evidence of the major causes. It is clear that sensitive lakes in the UK
with moderate to high sulfur deposition have been acidified since the
middle of the nineteenth century. Figure 11 shows the diatom assemblages at different depths from Lochnagar in the north-east of Scotland,
the onset of acidification is evident from about 1890 when the acid
neutralizing capacity had been exceeded and the pH of the lake decreased
by 0.5 pH units.39
Evidence for widespread acidification of soils in sensitive areas of
catchments showing surface water acidification was slow to change.
Many such soils are naturally quite acidic and it was difficult to envisage
what further changes would occur. However, the acidifying ions of
sulfate and nitrate are important in bringing about soil chemical
changes. In particular they promote the leaching of base cations,
impoverishing the nutrient status of such soils further, creating stronger
acidic conditions and promoting mobilization of aluminium. In less
acidic soils with a higher base cation status and ones in which cation
exchange processes act as a buffer, the first signs of acidification are
increased leaching of base cations as hydrogen ions replace them on
exchange sites. Ultimately the soil becomes more acidic and deficient in
base cations, aluminium is mobilized and in turn also displaces further
cations from the exchange sites.
Long term changes in the pH of soils of south-west Sweden have been
shown to have occurred by repeating measurements that were done
over 50 years previously on exactly the same soil plots. Figure 12 shows
that in the surface litter layer of forests of equivalent ages, pH levels in
1984 were lower than in 1927; in lower horizon, pH of soil was
independent of stand age and recent soil samples had a lower pH.
Although slightly smaller in scale, similar results have been reported for
soils in northern Sweden. The falls in pH are thought to be caused
mainly by acid deposition although biological processes also could have
been involved.
The critical loads approach is being applied over large areas of Europe
and North America in an attempt to reverse surface water acidification.
The objective is to reduce sulfur dioxide emissions such that the resulting
acidic input to aquatic systems is equal to or less than the production of
base cations from catchment weathering processes. A critical alkalinity
39

V. J. Jones, R. J. Flower, P. J. Appleby, J. Natkanski, N. Richardson, B. Ripley, A. C. Stevenson,
and R. W. Batterbee, J. Ecol., 1993, 81, 3-24.

total diatoms %
Figure 11

Summary diatom total percentage frequency diagram (left) and pH reconstructions (right), based on weighted averaging (
) and
multiple regression (
) methods from Lochnagar. Arrows show 1900 (lower) and 1950 (upper) levels at each site based on l Pb dating
(Reproduced with permission from J. EcoL, 1993, 81, 3-24)

Humus layer
1927
1984

Years after clearcutting or planting
pH

C-Horizon

1927
1984

Years after clearcutting or planting
Figure 12 pH in water suspensions of fresh soil as a function of the logarithmic stand age
of Norway spruce stands at the Tonnersjoheden experimental area in southwest Sweden. Data are shown for the humus layer (O horizon) and the C
horizon. Vertical bars show the standard deviation of the mean for n > 3
(Hallbacken, L. & Tamm, C. O., Scand. J. Forest Res., 1986,1, 139-143)

value is chosen to protect vulnerable aquatic species or as a critical point
to aim towards in reducing the acid inputs. These reductions in acid load
are translated into reduction of SO2 emissions which are needed to meet
the critical load. Recently it has been argued that whilst a fixed alkalinity
value may define the critical load for a particular species, it does not
represent the overall alkalinity of a lake. Palaeolimnological methods
using diatoms to reconstruct historical pH values of lakes show that the
pH has varied by at least one pH unit throughout its history. It has been
suggested that palaeolimnological information obtained from diatom
studies provide a useful means of setting critical loads for a particular
system.40 Critical loads with respect to soil acidification are based on
ratios of base cations (e.g. Ca2 + ) relative to Al concentrations in the soil
solution.41
40
41

R. W. Batterbee, T. E. H. Allott, S. Juggins, A. M. Kreiser, C. Curtis, and R. Harriman, Ambio,
1996,25,366-369.
H. Lokke, J. Bak, U. Falkengren-Grerup, R. D. Finley, H. Llvesniemi, P. H. Nygaard, and M.
Starr, Ambio, 1996,25,510.

11 FOREST DECLINE
Reports of problems in the health of forests in western Europe first
appeared in the late 1970s when white fir stands at high altitude in
Germany were found to be in an unhealthy state and then in the early
1980s Norway spruce (Picea abies), the dominant silviculture tree in
central Europe, began to show increasing signs of defoliation and needle
discolouration. Extensive death of the fine root system of trees has also
been reported. These symptoms have since been observed in the UK,
France, the Benelux countries, Scandinavia, the Alps (Switzerland and
Austria), the Czech and Slovak Republics, Poland, and the north-eastern
USA. The decline in forest health is not confined to Norway spruce;
other conifer and broad leaved species have been affected. It has been
argued that whilst regional declines of individual species can occur, for
example due to disease epidemics, the synchronous decline of several
species across a broad area suggests a common cause such as air
pollution. In 1987 and 1988 a survey of the severity of damage across
the European Community (EC) showed that the forests of southern
Europe were in better health than those of central and northern Europe.
In the major coniferous forests of northern Europe, 15%, 22% and 28%
of Picea abies, Picea sitchensis, and Abies alba trees were unhealthy.
Among the hardwoods of northern and central Europe, 16%, 15% and
12% of Quercus robur, Quercus petrea, and Fagus syhatica showed
symptoms of damage.33
Several reports have been published throughout Europe on the subject
of forest decline. However, definitive cause-effect relationships of forest
decline have yet to be established. In the early stages it was attributed to
acid deposition, through its effect on leaching of nutrients from foliage,
loss of nutrients from soils, and increasing bioavailable forms of toxic
metals such as aluminium in soils. It is believed that the actual explanation probably involves a combination of factors which include natural
and pollution stresses. A number of hypotheses have been put forward.
They include multiple stress, soil acidification and aluminium toxicity,
interaction between ozone and acid mist, magnesium deficiency, and
excess nitrogen deposition. There are different regional types of damage
and decline and in Norway spruce five different types have been
identified. The widely recognized needle yellowing at higher altitudes
appears to be associated with Mg deficiency brought on by a combination
of factors such as acid deposition, management practices, and abiotic
factors. It has been suggested that excess nitrogen deposition could also
play a part by triggering increased growth, resulting in Mg deficiency
and/or rendering trees more susceptible to damage by frosts and pests.42
42

L. W. Blank, T. M. Roberts, and R. A. Skeffington, Nature, 1988, 336, 27-30.

It is difficult to reconcile pollution patterns with the distribution of
forest decline on a regional scale such as Europe. The distribution of air
pollution which is typically a complex mixture of gases (SO2, NO*, O3)
and acidic aerosols and ammonium sulfate varies markedly from region
to region and over short distances within regions. Equally, forest
condition varies enormously across the region and between the major
forest species. Field observations on Norway spruce have shown that
damage was associated with areas of Europe where air pollution is
greatest, in particular deposition patterns of S and N pollutants and
summer concentrations of O3.43
Forest decline is frequently associated with foliar deficiency of certain
nutrients, and in particular magnesium deficiency, e.g. at elevated sites in
western Germany, although in other areas potassium deficiency has been
found. However, the cause of the nutrient deficiency is unknown. There
are reports of increased foliar leaching of mineral nutrients due to acidic
deposition or other pollutants but this is thought to be of secondary
importance and the supply of mineral nutrients from the soil is probably
more critical. There is growing evidence that the deposition of sulfur and
nitrogen pollutants have significantly modified soil chemistry and plant
nutrition. Increased nitrates and sulfates in the soil solution can increase
leaching of cations such as calcium and magnesium from soils and
enhance soil acidification. Soil solution chemistry, notably decreases in
the ratio of calcium and/or magnesium to aluminium, is known to affect
root development and hence water and nutrient uptake.33'43'44 Recent
work has shown that defoliation in Picea abies in Norway is associated
with reduced Mg and Ca in soil and it has occurred in areas with the
highest acidic deposition. It was concluded that acidic pollution in
excess of estimated critical loads was consistent with the pattern of
defoliation.45
12 EFFECTS OF POLLUTANTS ON REPRODUCTION AND
DEVELOPMENT: EVIDENCE OF ENDOCRINE
DISRUPTION
There are a number of instances of pollution-related events in wildlife
populations where changes in reproduction and development have
occurred which have led to population declines and on occasions to
extinctions. There is growing evidence that reproductive and developmental abnormalities observed are a consequence of effects on the
43

E.-D. Schulze and P. H. Freer-Smith, 'Acid Deposition—Its Nature and Impacts', ed. F. T. Last
and B. Watling, Proc. Roy. Soc. Edinburgh, 1991, Section B, 97, 155.
44
E ^ D . Schulze, Science, 1989, 244, 776-783.
45
C. Nellemann and T. Frogner, Ambio, 1994, 23, 255-259.

endocrine system, although the fundamental mechanism has yet to be
established.

12.1

Eggshell Thinning

A wealth of literature exists on the relationship between DDT (and its
metabolite DDE) and the phenomenon of eggshell thinning in various
bird populations. 9 ' 46 It was first discovered in the peregrine falcon (Falco
peregrinus) in the UK. This particular falcon is widespread throughout
Eurasia and North America, feeding almost entirely on live birds caught
in flight. From about 1955 onwards, for no obvious reason, the numbers
of falcons rapidly declined in southern England and subsequently this
decline spread northwards to parts of the Scottish highlands. By 1962 in
the UK, some 51 % of all known pre-war territories had been deserted
and this figure was as high as 93% for southern England. At the time
biologists in several countries were also investigating populations of
birds that had slowly declined to critical levels. It emerged that they all
had residues of DDT, its metabolites, other chlorinated hydrocarbon
insecticides, polychlorinated biphenyls (PCBs), and other chemicals in
their tissues.
Ratcliffe47 examined the eggshells of peregrine falcons and sparrow
hawks in the UK and using an index of eggshell thickness (eggshell
weight/(egg length x egg breadth)) found that during 1947 and 1948 this
index decreased significantly by, on average, 19%. Eggshell thinning has
been found in several American species of birds, notably the bald eagle,
osprey, and peregrine falcon. In a study of over 23 000 eggshells of 25
bird species, shell thinning was found in 22 of the species and a
correlation was found between the degree of thinning and DDE residues
in the eggs. Experimental work with mallards and American kestrels
showed that DDE caused eggshell thinning. Lincer48 compared the
degree of thinning with the concentration of DDE in eggs of the
American kestrel {Falco sparverius) in field populations and in eggs
produced by captive birds fed food dosed with DDE. Shell thickness
decreased with increasing DDE residues in the eggs of both populations
(Figure 13).
It is well accepted that DDE is the primary cause of eggshell thinning
in many kinds of birds and that some bird species are more susceptible
than others to DDE-induced shell thinning. Eggshell thinning was
definitely a major cause of low reproductive success and population
46

T. J. Peterle, 'Wildlife Toxicology', Van Nostrand Reinhold, New York, 1991.
D . A. Ratcliffe, / . Appl. EcoL, 1970, 7, 67.
48
J. L. Lincer, / . Appl. EcoL, 1975, 12, 781.
47

Shell thickness (Rofcliffe's index as % of pre-DDT thickness)

DDE residues in eggs (parts x IO~6)
Figure 13 The relationship between mean clutch shell thickness and DDE residue of
American kestrel eggs fFalco sparverius,) collected from wild populations in
Ithaca, New York, during 1970 (•) and the same relationship experimentally
induced with dietary DDE ( x )
(Reproduced with permission from Appl. EcoL, 1975,12, 781)

decline in some species, but chlorinated hydrocarbons probably contributed to declines in other ways. In general terms, reproductive
trouble tends to increase as shells become thinner and thinning of over
20% is likely to result in reproductive failure and population decline. In
the case of peregrine falcon populations in the UK, the incidence of
broken eggs within clutches rose from the normal 4% to 39% during
the period 1951-66. Although thinner eggshells resulted in an increased
breakage of eggs and, in turn, a reduction in breeding success of a
particular pair of falcons, it was realized that the overall size of the UK
population was unaffected by this effect at the time. The actual
population decline took place at least five years after the onset of
eggshell thinning. This coincided more closely with upsurge in the use
of dieldrin as an anti-fungicidal seed dressing. Although it has never
been substantiated, it is widely believed that the abrupt decline was due
to a combination of factors, involving eggshell thinning and the later
ingestion of toxic doses of dieldrin derived from a diet of contaminated
pigeons.
One of the earliest examples in which PCBs were implicated in bird
mortalities was an incident in the autumn of 1969 when 15 000 dead and
dying seabirds were washed up on the coasts around the Irish Sea. Nearly
all the birds were guillemots (Uria aalge) and almost all the victims were

adults just after the annual moult. The bulk of the birds were washed up
after storms in late September, but even during the previous fine calm
weather some dead and dying birds were spotted. Chemical analyses of
dead guillemots for a variety of pollutants revealed the presence of high
concentrations of DDE and in particularly very high levels of PCBs in
their livers. It has been suggested that moulting, followed by the stress of
withstanding stormy conditions resulted in mobilization of fat reserves
and in turn release of toxic doses of PCBs into the blood stream.1
Experimental studies with mink and ferret have established that PCBs
are highly disruptive of reproductive processes. In the American mink
(Mustela vison) tissue concentrations of over SOmgkg" 1 PCB are
associated with reproductive failure. The otter, which is a closely related
species to the mink, has experienced population declines over wide areas
of Europe since the 1950s and it has been found that animals from
populations showing the greatest decline commonly had PCB concentrations in their tissues of over 50 mg kg~ *.
12.2 GLEMEDS
Organochlorine compounds such as PCBs, DDT, and TCDD have been
associated with physiological, reproductive, developmental, behavioural, and population level problems in fish-eating birds of the Great
Lakes for over 30 years.10 Common features have been mortalities and
deformities of eggs and chicks which led to the condition being called
GLEMEDS (Great Lake Embryo Mortality, Edema, and Deformities
Syndrome). The symptoms showed a number of similarities to chickedema disease in poultry caused by exposure to dioxin contaminated
food which was first recognized in the 1950s. The syndrome first came to
light in common terns and herring gulls in the 1970s with observations of
high incidences of egg and chick mortalities and deformities, particularly
in colonies from Lake Ontario. By the mid-1970s organochlorine
concentrations had declined and this corresponded with the disappearance of the gross effects in Lake Ontario herring gulls. During the same
period there were declines in bald eagle populations. Subsequent studies
reported stronger associations with PCBs and dioxins, e.g. deformities
and egg mortality in colonies of double crested cormorants and Caspian
terns were correlated with dioxin contamination and with specific
congeners of PCBs in Foster's terns from Lake Michigan. Reproductive
and developmental effects have been found in other wildlife populations,
e.g. egg mortalities and deformities in snapping turtles from the Great
Lakes basin, and increased tumour incidence in the Beluga whale in the
St Lawrence have been associated with levels of organochlorines and in
particular PCBs. Recently developmental problems in infants from

regions of Lake Michigan and Lake Ontario have been related to
maternal exposure to organochlorines as a result of consumption of
contaminated fish.
Between 1992 and 1994 immunological responses and related variables
were measured in pre-fledgling herring gulls and Caspian terns at
colonies across a broad range of organochlorine contamination (mainly
PCBs), as measured in the eggs. There was a strong exposure-response
relationship in both species between organochlorines and suppressed Tcell-mediated immunity. Suppression was most severe (30-45%) in
colonies in Lake Ontario (1992) and Saginaw Bay (1992-94) for both
species and in western Lake Erie (1992) for herring gulls.49
Although most of the evidence is through correlation, taken as a whole
there is good reason to believe that the reproductive and developmental
problems highlighted are due to organochlorines with PCBs and dioxins
as the major problem agents. Whilst inputs of organochlorines to the
Great Lakes have largely been controlled and concentrations have
declined there are still measurable levels in wildlife and concentrations
in fish remain a significant hazard. The residual levels are due to the
persistence of these substances, leachates from landfill sites, and inputs
from the atmosphere.
12.3 Marine Mammals

There is a history of reproductive problems in seal populations in the
Baltic and in the Dutch Wadden Sea. Between 1950 and 1970 the
population in the Wadden Sea dropped from more than 3000 to less
than 500 animals. The reduction in numbers is associated with low
breeding success. The populations from both the Baltic and Wadden
were found to contain high levels of organochlorine compounds and in
particular PCBs.
Case Study 4. Epizootic disease and impaired immune response in marine

mammals. There have been a number of major disease outbreaks
among seals and dolphins which have been attributed to infection with
known or newly recognized morbilliviruses. In 1988 the previously
unrecognized phocine distemper virus caused the death of 20000
harbour seals in north-western Europe. Other morbilliviruses have been
shown to be the sources of infections in porpoises and dolphins, and
mass mortalities due to this type of virus occurred among striped
dolphins in the Mediterranean from 1990 to 1992.50 The severity and
49

K. A. Crassman, G. A. Fox, P. F. Scanlon, and J. P. Ludwig, Environ. Health Perspect., 1996,104,
(suppl. 4), 829-842.
R. L. De Swart, T. C. Harder, P. S. Ross, H. W. Vos, and A. D. M. E. Osterhaus, Infect. Agent.
Dis., 1995,4, 125-130.

extent of these virus related diseases has led to speculation that pollution
and in particular the bioaccumulation of organochlorines could have
had damaging effects on the immune systems of the animals and as a
consequence made them more vulnerable to infection.
Semi-field experiments in which groups of harbour seals were fed fish
contaminated with PCBs have shown a strong link with reproductive
failure. Experiments were over two years or more and during this time
one group of harbour seals were fed a diet offish from the contaminated
Baltic Sea and another group were fed relatively uncontaminated fish
from the Atlantic. The reproductive success of the group receiving the
higher dose of PCBs was significantly lower than the other group with
the lower dose of PCBs.51 Follow up studies have shown more conclusively that PCBs are the main group of pollutants associated with these
effects and immunotoxicological studies under the same semi-field
conditions have demonstrated impaired immune responses in animals
with elevated body burdens of PCBs.52

12.4

Imposex in Gastropods

Case Study 5. Tributyltin and imposex in gastropods. Sex abnormalities in neogastropod molluscs were first recorded in 1970. Many females
in dogwhelk populations (Nucella lapillus) from Plymouth Sound, UK
were found to have penis-like structures, this was followed by the
reporting of male characteristics in female American mud snails from
the Connecticut coast, USA. This development of male characters in
females is referred to as imposex. In the 1970s reproductive failure and
major population declines of the English native oyster occurred and the
introduced Pacific oyster, Crassostrea gigas, at certain locations along
the British east coast and the French west coast, started to show poor
growth and unusual thickening of their shells. Populations of both
gastropods and bivalve molluscs showing the severest abnormalities
were invariably in the vicinity of yachting marinas and it was later
shown that their tissues had a high tin content. In the 1980s it was firmly
established that antifouling paints containing tributyltin (TBT) were
responsible for these effects. This was confirmed by laboratory tests
which reproduced imposex in gastropods and shell thickening in the
bivalve molluscs at the same concentrations of TBT found near to
yachting marinas and it was also shown that the tissue concentrations
51
52

P. J. H. Reijnders, Nature, 1986, 324, 456.
R. L. De Swart, P. S. Ross, J. G. Vos, and A. D. M. E. Osterhaus, Environ. Health Perspect., 1996,
104, (suppl 4), 823-827.

of TBT associated with these effects were the same in field surveys and
laboratory experiments.53
Many other species of gastropods have been found to exhibit TBTinduced masculinization of the female and the phenomenon is worldwide
in occurrence with over 100 species showing varying degrees of imposex.
The consequences of imposex vary according to species; in some the
abnormality does not appear to affect reproduction, whilst in others such
as the dogwhelk, abnormality can be so severe that breeding is prevented
and as a result populations have declined drastically and even become
extinct. In the dogwhelk the degree of imposex can be related to the level
of TBT exposure and the onset of symptoms is associated with exposures
of less than 1 ng/1 TBT. Concentrations of 1-3 ng/1 result in more
pronounced masculinization with the penis size approaching that of
males, although at this stage breeding is unaffected. However at 5 ng/1
the development of the male sex organs begin to cause blockages in the
female tissues and the snail in this condition is unable to breed. The noeffect exposure with respect to imposex is very much lower than that
estimated from initial screening studies directed at other end-points and
mortality. The use of TBT as an antifouling agent on small vessels was
banned in the UK in 1987 and affected populations have started to
recover. However it is still permitted for use on large vessels and imposex
has been found in the edible whelk from the North Sea close to busy
shipping routes.
Imposex is associated with elevated titres of testosterone and it has
been shown that administration of testosterone to females produces
masculine characters. One mechanism, for which there is good evidence,
is that TBT inhibits an aromatase enzyme involved in the metabolism of
testosterone to 17/?-oestradiol.54^56
12.5

Endocrine Disruptors

The possible effects of chemicals on endocrine function in humans and
wildlife has become one of the major issues. The functions of all organ
systems are regulated by the endocrine signalling system and disturbances of this system resulting in hormone imbalances can lead to a long
lasting damage especially during the early stages of the life cycle.
Chemicals that can disturb normal endocrine homeostasis are referred
to as endocrine disruptors.
53

Tributyl Tin: Environmental Fate and Effects', ed. M. A. Champ and P. F. Seligman, Elsevier
Applied Science, 1990.
54
J. M. Ruiz, G. Bachelet, P. Caumette, and O. F. X. Donard, Environ. Pollut., 1996, 93, 195-203.
55
G. W. Bryan and P. E. Gibbs, 'Metal Ecotoxicology: Concepts and Applications', ed. M. C.
Newman and A. W. Mclntosh, Lewis, Ann Arbor, 1991, p. 323.
56
P. Matthiessen and P. E. Gibbs, Environ. Toxicol. Chem., 1998,17, 37-43.

Studies of human populations indicate a deterioration in male reproductive health in recent decades. These include reports from Belgium,
Denmark, France, and the UK of declines in semen quality in the 1990s,
although it has to be said that there is criticism of the methods of analysis
and therefore interpretation of the results. Over the same period there
has been an increased incidence in testicular cancer, and certain disorders
of the male reproductive system, hypospadias and cryptorchidism,
appear to be increasing. The synthetic oestrogen, diethylstilbestrol
(DES) has been widely prescribed to women during pregnancy as a
means of combating abortion and other complications of pregnancy.
There is now strong evidence that this has resulted in a higher incidence
of reproductive disorders in the sons of these women. Diethylstilbestrol
has also been found to produce the same reproductive abnormalities in
animal studies.57'58
Several pollutants have been shown to be capable of disrupting
endocrine systems and in particular of mimicking the activity of natural
oestrogens. It is believed that these oestrogenic substances produce
effects through interaction with the oestrogen receptor. A variety of
substances, including the organochlorines, DDT, PCBs, and dioxins
have been shown to have oestrogen disrupting activity, although their
affinity for the oestrogen receptor is much lower than that of natural
oestrogens. There is therefore considerable concern that the widespread
reproductive and developmental problems reported in humans and
wildlife are related and that they are caused by substances which affect
the endocrine system in animals.
As long ago as 1968, DDT was reported to have oestrogenic activity59'60 and recently its metabolite /?//-DDE was reported to have potent
antiandrogen properties (demasculinization).61 Injecting DDT into eggs
of California gulls (Larus californicus) at similar levels to those found in
wild populations in the late 1960s produced feminization in male birds. It
has been observed in large gull populations that the female to male ratio
is skewed to females, suggesting a possible causal link.
Lake Apopka in Florida was polluted by DDT and related compounds following an accidental spill by a chemical company in 1980. In
the following years the alligator population declined whereas at other
nearby locations populations increased. The decline was associated with
reproductive disorders and in particular male alligators showed a
number of reproductive abnormalities. Sex differentiation in the alligator
57

R. J. Kavlock et al., Environ. Health Perspect., 1996, 104, (suppl. 4), 715-740.
Medical Research Council, 'IEH Assessment on Environmental Oestrogens: Consequences to
Human Health and Wildlife', University of Leicester, Leicester, 1995.
59
D . M. Fry and C. K. Toone, Science, 1981, 213, 922.
60
J. C. Bitman et al., Science, 1968, 162, 371-372.
61
W. R. Kelce et al., Nature, 1995, 375, 581-585.
58

is temperature dependent and can be altered by oestrogen treatment, so it
is suggested that the organochlorine compounds were acting as oestrogenic agents impairing normal sexual development of male alligators.62
Metabolites of PCBs have been shown to have strong oestrogenic effects
in turtles.63 These animals exhibit temperature-dependent sex differentiation and eggs incubated at 26 0C normally produce 100% males but
when they were exposed to PCB metabolites under the same conditions
the offspring were female.
Case Study 6. Vitellogenin as a biomarker for oestrogenic disruption in
fish. Vitellogenin is a yolk protein in egglaying vertebrates and invertebrates. In fish it is produced in the liver of females and transported in the
bloodstream to the oocytes. The synthesis of this protein is regulated by
oestrogens and during egg production it is normally elevated in blood
serum of females. However in males and immature females vitellogenin
production can be induced by exposure to oestrogens and oestrogenic
agents. In the UK elevated levels of vitellogenin have been reported in
male freshwater fish exposed to sewage effluents and similar results have
been reported in the USA and in marine fish exposed to sewage effluent.
The oestrogenic agent responsible for these raised levels is unclear and
may vary between locations depending on the composition of the
contaminating effluent. It has been suggested that natural and synthetic
oestrogens could be involved. Also another group of substances, the
alkylphenols, which are commonly present in such effluents, have been
shown to induce vitellogenin and impair reproductive development in
male trout. These substances are degradation products of non-ionic
surfactants produced during sewage treatment.64"68
The number of instances of reproductive and developmental impairment
in wildlife and human populations is clearly of concern and perhaps
indicative of a common cause. However the link between environmental
oestrogens and human disorders has yet to be established and in wildlife
the causal link is known in only a few cases. The endocrine system in all
animals is complex and not fully understood and so the consequences of
inhibiting or enhancing it at one point or another are largely unknown.
The system also interacts with both the nervous and immunological
62

L. J. Guilette, T. S. Gross, G. R. Masson, J. M. Matter, H. F. Percival, and A. R. Woodward,
Environ. Health Perspect., 1994, 102, 680-687.
J. M. Bergeron et al., Environ. Health Perspect., 1994, 102, 780-781.
64
J. P. Sumpter and S. Jobling, Environ. Health Perspect., 1995, 103, (suppl. 7), 173-178.
65
C. M. Lye, C. L. J. Frid, M. E. Gill, and D. McCormick, Mar. Pollut. Bull., 1997, 34, 3 ^ 4 1 .
66
J. E. Harries, D. A. Sheahan, S. Jobling, P. Matthiessen, P. Neall, E. J. Routledge, R. Ryecroft,
J. P. Sumpter, and T. Taylor, Environ. Toxicol. Chem., 1996, 15, 1993-2002.
67
J. E. Harries, D. A. Sheahan, S. Jobling, P. Matthiessen, P. Neall, J. P. Sumpter, T. Taylor, and N.
Zaman, Environ. Toxicol. Chem., 1997, 16, 534-542.
68
S. Jobling and J. P. Sumpter, Aquatic Toxicol., 1993, 27, 361-372.
63

Direct acting
endocrine disruptor
Direct acting target
Xenobiotic
organ toxicant

Endocrine
system

Indirect acting
endocrine disruptor

Target Organs
Reproductive system
Immune system
Neurological system i
Adverse*
effect

Figure 14

Summary of the effect of xenobiotic substances on the endocrine system in
animals, highlighting the possible interactions between the endocrine, nervous
and immunological systems
(Reproduced with permission from Environ. Health Persped., 1996, 104
(suppl.4), 715)

systems and disruption to one of these by pollutants could have
repercussions on the others. Figure 14 provides a summary of the
possible interactions and it is interesting to note that some of the
observations in wildlife, particularly in marine mammals, points
towards possible interactions between these systems.
13 HYDROCARBONS IN THE MARINE ENVIRONMENT
Crude oil is a complex mixture of a great variety of organic substances
with many different physical and chemical properties. The physical and
chemical composition of oils of different origins also vary widely. In
March 1978, the supertanker Amoco Cadiz released most of its 223 000
tonnes cargo of Iranian and Arabian crude oil into the coastal waters of
Brittany. The oil was light, it had a low viscosity, and it contained 3035% of highly toxic aromatic hydrocarbons, 39% saturated hydrocarbons, 24% polar material, and 3% residuals and there were more than
300 compounds in the fresh crude oil. Gullfaks oil which was involved in
the Braer spill in the Shetlands in January 1993 is an unusual one in that
it is a light biodegraded oil with a low concentration of such components
as w-alkanes and asphaltenes and a high aromatic and naphthenic
content.69
Following a spillage, oil is subject to various environmental processes
that affect its chemical and physical composition, including evaporation,
dissolution, and microbial and photochemical degradation. The lower
molecular mass fractions are generally more volatile, more soluble, and
more easily degraded and so an ageing or weathered oil has lost these
components and it is made up of the more viscous higher molecular
69

G. A. Wolff, M. R. Preston, G. Harriman, and S. J. Rowland, Mar. Pollut. Bull., 1993, 26, 567571.

weight compounds. The more persistent and larger polynuclear aromatic
hydrocarbons (PAHs) and their transformation products at this stage
assume greater ecological significance. Despite sensitive and modern
sophisticated instrumentation only a small spectrum of the compounds
in crude oil are routinely determined. Consequently, it is not easy to
identify the specific toxic components in oil. In the immediate aftermath
of a spill, the lighter aromatic fractions such as monocyclic aromatic
hydrocarbons, naphthalenes, and phenanthrenes are present and these
are acutely toxic and partially water soluble. The effects on marine
organisms at this stage are largely due to short term toxic effects and
narcotization. The more persistent PAHs are chronically toxic and 5-6ring PAHs have carcinogenic and mutagenic properties. Because of their
relative insolubility in water but solubility in lipids they accumulate in
organisms and in organic matter of suspended particles and sediments.
They can remain bound to sediments for several years. The bioavailability in this form is therefore an important factor in the toxicity of
PAHs.
A major oil spillage in the marine environment has immediate
catastrophic results. However, such events constitute a relatively small
percentage of the total quantity of petroleum compounds discharged
each year. Chronic inputs from shipping, offshore wells, industry, rivers,
and the atmosphere together make a very significant contribution. The
fate and effects of these discharges are particularly difficult to monitor
and assess, as the more toxic fractions which tend to be more volatile,
soluble, and easily degraded in the environment are continually being
removed from the oil. In consequence, much of our knowledge and
understanding comes from investigations undertaken following a major
accident.70'71
Accidents of this nature in the marine environment have considerable impact on marine birds and the adjoining coastal ecosystems,
bearing in mind that changes in the composition of the oil occur fairly
rapidly, and much of the acutely toxic components disappear in a
matter of days. Deleterious effects are evident at sea, but because of the
added influence of dispersal, they tend to be short lived covering a time
course of days or at most weeks. The effect of oil reaching the coastal
environment depends on the distance of travel and therefore the time
taken for the oil to reach the coast. In this respect it is interesting to
compare the Torrey Canyon incident (a spillage of 100000 tonnes some
200 miles off the coast of Cornwall) with that of the Amoco Cadiz. By
the time that oil was stranded on the Cornish beaches from the Torrey
70

Royal Commission on Environmental Pollution, Eighth Report, C m n d . 8358, H M S O , L o n d o n ,
1981.
71
R. B. Clarke, Phil Trans. Roy. Soc. London, B, 1982, 297.

Canyon it was almost biologically inert and the damage that ensued
was largely caused by dispersants used to clean the beaches. In the case
of the Amoco Cadiz the spillage was only 1.5 nautical miles from the
shore and a major part of the acutely toxic aromatic fraction was still
present when the oil reached the coast causing considerable fatalities
among marine life.
Oil washed onshore interacts with a variety of coastal features,
extending from high energy eroding rocky promontories to low energy
accumulating environments. The processes that degrade oil in the open
water operate in the coastal environment but to differing degrees. In high
energy and medium energy coastal systems the hydraulic action of
breaking waves mechanically disperses and erodes the oil rapidly.
During the Braer spill conditions were particularly stormy, so the oil
was rapidly dispersed and acute fatalities were confined to the immediate
vicinity of the wreck. By contrast, oil may persist for many years in the
low energy environments that include lagoons, estuaries, and marshes,
since in these areas wave energy is low and degradation is dependent on
the slower processes of microbial degradation and dissolution. With
time, attention shifts from the immediate short term acute effects of oil
and focuses on the more persistent long term effects of PAHs in the low
energy coastal systems.
Case Study 7. The impact of oil spills in coastal waters. About 360 km of
the Brittany coastal environment was polluted by oil following the
wrecking of the Amoco Cadiz. This included rocky and sandy shores,
salt marshes, and estuaries. During the first few weeks after the disaster a
very heavy mortality or 'acute mortality crisis' affected the intertidal and
subtidal fauna. Populations of bivalves, periwinkles, limpets, peracarid
crustaceans, heart urchins, and seabirds were most severely affected.
Populations of polychaete worms, large crustaceans, and coastal fishes
were less affected. Highest mortalities were found in a 5 km radius
around the wreck and in locations further afield where the oil was
blown and accumulated ashore by the wind. At one of these locations
(St Efflam) for example, mortalities included 106 heart urchins (Echinocar dium cor datum), 1.5 x 106 cockles (Cardiumedule), and 7 x 106 other
bivalves (Solenidae, Mactridae, Veneridae). Within a radius of 10 km of
the wreck about 104 fish were found dead, mainly wrasses (Labridae),
sand eels (Ammodytes sp.), and pipe fishes (Syngnathidae), as well as
large crustaceans (Cancer crangon, Leander seratus). There was no
evidence of a severe impact of oil pollution on any intertidal or subtidal
species of algae. Delayed effects on mortality, growth, and recruitment
were still observed up to three years after the spill. For example,
estuarine flat fish and mullets exhibited reduced growth, fecundity, and
recruitment and many were affected by varying degrees of fish-rot

disease. Populations of clams and nematodes in the meiofauna declined
one year after the spill. Weathered oil remained for a number of years in
low energy areas although the biological consequences of this were
unknown.72
The longer term biological effects of oil are mainly concerned with
PAH residues. The level of exposure can be assessed by measurement of
residues accumulated in tissues of fish, but because fish are able to
metabolize PAH rapidly other indicators of exposure are used. In fact
activation of monooxygenase enzymes (P4501A) involved in PAH
metabolism is the primary biological response to PAH contamination
in fish and several studies have demonstrated raised levels of this enzyme
system in relation to exposure to oil contamination, e.g. Exxon Valdez in
Alaska, and the Gulf. In the process the metabolites produced, notably
trans-diols, cause DNA damage and there is growing evidence correlating P4501A induction to PAH levels, DNA damage, hepatic carcinogenesis, and other pathological conditions. The life time of these
activated metabolites in fish depends very much on a further group of
enzymes (Phase II enzymes) to conjugate and detoxify them.
The Braer oil spill occurred close to several salmon farms situated
along the coast of Scotland, and caged fish at these farms provided
captive sentinels for determining the impact, spatially and temporally, of
the oil. A study of these fish demonstrated that, during the initial phase
of the spill, fish were exposed to oil in water, resulting in contamination
of tissues with PAH and the induction of both monooxygenase and
conjugating enzymes. The exposure and biochemical responses declined
rapidly with time, such that they had returned to those at control sites
after six months. In common dab populations exposed to oil in the water
and in sediments, pathological analysis of livers showed changes or
lesions in fish from the most contaminated sites symptomatic of the
progression to neoplasia. These lesions were only observed in fish a year
after the spill but it was not clear whether they were a consequence of the
initial short term exposure to oil in the water or to longer term exposure
to oil in sediments.73
Seabird mortality due to oil pollution generates considerable public
outcry and such incidents are mainly associated with major or acute oil
pollution episodes. In addition, various surveys have indicated that large
numbers of seabirds are killed by oil in the North-West Atlantic
throughout the winter, peaking in January-March. These winter deaths
72
73

L. Laubier, Ambio, 1980, 9, 268.
R. M. Stagg, A. M. Mclntosh, C. F. Moffat, C. Robinson, S. Smith, and D. W. Brim, Proceedings
of the Royal Society of Edinburgh Conference on 'The Impact of an Oil Spill in Turbulent Waters:
The Braer\ HMSO, London, XXXX.

are associated more with the ongoing and diffuse chronic oil pollution
problem, although clearly various stresses due to the weather are likely to
influence mortalities. However, the number of birds killed annually by
oil (tens of thousands per annum) is small compared with losses due to
natural causes (hundreds of thousands). There has been considerable
concern that local populations of certain species are particularly at risk
from oil pollution. However, of the many species of seabirds, relatively
few, notably the auks (Razorbills, Guillemots, Puffins, etc.) and the
diving sea ducks (Eiders, Scoters, etc.), have suffered severe mortalities.
These species are particularly susceptible to oil pollution since they spend
almost their entire lives at sea, collect their food by diving and, in most
cases, have a low breeding rate. Furthermore, these birds are highly
gregarious, particularly in their breeding and wintering areas and so a
localized spillage can inflict large numbers of casualties.
The impact of two contrasting refinery effluent discharges on a rocky
shore (Milford Haven) and a saltmarsh/seabed community (Southampton Water) have been followed for a number of years. Monitoring has
continued at both locations more or less since the commissioning of the
refineries. In 1974, in the vicinity of the effluent of the refinery at Milford
Haven, the density of barnacle and limpet (Patella vulgata) populations
were considerably reduced. Laboratory and field experiments showed
that this was due to inhibition of larval settlement. When the refinery at
Milford Haven closed in 1983, the barnacle and limpet populations
showed a steady recovery over the next few years. The saltmarsh
communities in Southampton Water, which were extensively damaged
during the 1950s and 1960s, have shown a gradual recovery over a
number of years, in response to marked improvements in the quality of
the effluent.74
A widely held view is that oil spills are unlikely to cause long lasting
damage to the marine environment and that oil pollution generally does
not constitute a chronic threat to either marine ecosystems or indirectly
to humans. Recolonization after a spill favours those species with a high
fecundity, a short life cycle, and/or a planktonic stage in their life cycle.
However, recruitment can be very unstable and full recovery may take
several generations; species of such organisms as clams and fish with life
expectancies of 5-10 years would not be expected to attain stable
populations for up to 30 years, and seabird populations are likely to
take even longer.
It is still the case that surprisingly little is known about the long term
effects of oil and in particular about the long term bioavailability of
74

'Ecological Impacts of the Oil Industry', Institute of Petroleum, John Wiley & Sons, Chichester,
1989.

PAHs to communities foraging among contaminated sediments. There is
also little known about the vulnerability of polar and tropical ecosystems
to oil pollution.75 Jackson and co-workers investigated the effects of a
spillage of over 8 million litres of crude oil into a complex region of
mangroves, seagrasses, and coral reefs just east of the Caribbean
entrance to the Panama Canal. At the time it was the largest recorded
spillage into coastal habitats in the tropical Americas. The study is
significant because many of the habitats damaged by the oil had been
studied since 1968, following an earlier oil spill in the region, and also
because observations of the effects of the spill began as the oil was
coming ashore. In each of the oiled habitats the cover of the major
groups was greatly reduced. In the oiled habitats most of the roots
sampled for epibiota were dead broken or rotting and so the habitat will
not be restored until new trees grow. Seagrass, intertidal reef flat, and
subtidal reef habitats were similarly damaged. Of particular note was the
extensive mortality of subtidal corals and infauna of seagrasses, as this
contradicts the view that these habitats are not affected by oil spills. Also
sub-lethal effects of the corals including bleaching or swelling of tissues,
conspicuous production of mucus, and dead areas devoid of coral tissue
may affect the long term wellbeing of the habitat.76

14 HEALTH EFFECTS OF METAL POLLUTION
14.1

Mercury

Several major episodes of mercury poisoning in the general population
have been caused by consumption of methyl- and ethylmercury compounds. Methylmercury is a neurotoxin and the main clinical symptoms
of poisoning reflect damage to the nervous system. The sensory, visual,
and auditory functions, together with those of the brain areas, especially
the cerebellum, concerned with co-ordination, are the most commonly
affected. Symptoms of poisoning increase in severity in line with
increased exposure, as follows: (1) initial effects are non-specific symptoms including paraesthesia, malaise, and blurred vision; (2) in more
severe cases, concentric constriction of the visual field, ataxia, dysarthria,
and deafness appear more frequently; and (3) in the worst affected cases,
patients may go into a coma and die. The effects in severe cases are
irreversible due to destruction of neuronal cells. There is a latent period,
75

E. D. Da Silva, M. C. Peso-Aguiar, M. T. Navarro, and C. Chastinet, Environ. Toxicol. Chem.,
1997,16, 112-118.
J. B. C. Jackson, J. D. Cubit, B. D. Keller, V. Batista, K. Burns, H. M. Caffey, R. L. Caldwell,
S. D. Garrity, C. D. Getter, C. Gonzalez, H. M. Guzman, K. W. Kaufman, A. H. Knap, S. C.
Levings, M. J. Marshall, R. Steger, R. C. Thompson, and E. Weil, Science, 1989, 243, 37.

usually of several months between the onset of exposure and the
development of symptoms.6'77
Case Study 8. Epidemics of mercury poisoning in the general population;
exposure-response relationships.
The Iraqi outbreak. This epidemic of methylmercury poisoning
occurred in agricultural communities in Iraq in the winter of 1971-72.
Over 6000 people were admitted to hospitals in provinces throughout the
country and over 400 people died in hospital with methylmercury
poisoning. The poisonings arose from misuse of imported high grade
seed grain treated with alkylmercury fungicide. The imported seed was
intended for sowing but in many areas the grain was ground directly into
flour and used in the daily baking of homemade bread. Depending on the
number of loaves consumed, individual exposure ranged from a low nontoxic intake to a prolonged toxic intake over 1-2 months. Because of the
brief exposure period, epidemiological investigations began 2-3 months
after it had ceased and in most cases after the onset of poisoning. This
made calculation of ingested dose, and body burden of mercury at the
time of exposure to methylmercury difficult to estimate.
Figure 15 shows both dose-effect and dose-response relationships that
have been established between symptoms of poisoning and the estimated
body burden at the time of cessation of ingestion of methylmercury in
bread. An increased body burden of mercury is associated with an
increase in the severity of symptoms experienced by patients. The doseresponse curve for each sign or symptom shows the same characteristic
shape, a horizontal and a sloped line which is referred to as a 'hockey
stick' line. The horizontal line represents the background or general
frequency of each symptom and the sloped line shows that an increased
body burden is associated with an increasing frequency of each
symptom. The intersection of the two lines has been taken as the
'practical threshold' of the mercury-related response and this increases
with increasing severity of the effects; for paraesthesia it is at a body
burden of about 25 mg Hg, for ataxia it is at 50 mg, for dysarthria it is at
about 90 mg, for hearing loss it is at about 180 mg, and for death it is
over 200 mg. These relationships do not demonstrate cause and effect
and the only proof that methylmercury produced the above effects is that
the effects followed a known high exposure to methylmercury, the
frequency and severity of these effects increased with increasing exposure
to methylmercury, the effects are similar to those seen in other outbreaks
of methylmercury poisoning, and the major signs have been reproduced
in animal models.
77

'Environmental Health Criteria, 101, Methylmercury', World Health Organization, Geneva,
1990.

Frequency of response (% cases)

Paraesthesia
Ataxia
Dysarthria
Deafness
Death

Estimated body burden of mercury (mg)
Figure 15 The relationship between frequency of signs and symptoms of methylmercury
poisoning and the estimated body burden of methylmercury (the two scales of
the abscissa result from different methods of calculating body burden of
methylmercury)
(Reproduced with permission from Science, 1973, 181, 230)

The Minamata and Niigata outbreaks. In Japan, two major epidemics
of methylmercury poisoning have occurred, one in the Minamata Bay
area and the other in Niigata. In each case the problem arose as a result
of the local population consuming seafood contaminated with methylmercury. Mercury compounds, including methylmercury were released
from industrial sources into the aquatic environment and this resulted in
the accumulation of methylmercury in seafood. These outbreaks of
poisoning were first discovered during the 1950s and early 1960s and,
by the mid-1970s, about 1000 cases (with 3000 suspects) in the Minamata
area and over 600 in the Niigata area had been recorded. By and large,
symptoms of poisoning followed a very similar pattern to the Iraqi
epidemic, although an important difference was in the nature of the
exposure which was lower but more prolonged. Moreover, the latent
period was longer and in some isolated cases it was as long as 10 years
between initial exposure and the onset of symptoms. In other cases it was
observed that clinical symptoms worsened with time, despite reduced or
discontinued exposure. All this would appear to be related to long term
accumulation of mercury in the brain.
Because mercury even from natural sources in fish is predominantly in
the methylmercury form, there is concern that certain groups of people

who depend largely on a fish diet will be exposed to slightly elevated
intakes throughout their lives and hence possibly accumulate mercury
to toxic levels. One such group is the Canadian Indians but because of
several confounding factors, not least of which are high incidences of
malnutrition and alcoholism, the assessment of health risk due to
methylmercury is difficult. One study, involving 35 000 samples
obtained from 350 communities, found that over two-thirds had
mercury blood concentrations within normal limits (<20 fig P 1 ) but
2.5% (over 900 individuals) had levels in excess of 100 fig I" 1 and which
therefore could be considered as a group 'at risk' and in need of close
surveillance.
Mercury analysis of segments of hair provides a meaningful index of
past exposures and hence body burden (see Figure 3). As a guideline,
blood mercury concentrations of 200-500 fig I" 1 , hair concentrations of
50-125 fig g~\ and a long term intake of 3-7 fig kg" 1 body weight are
likely to be associated with the onset of the initial symptoms of
methylmercury poisoning, such as paraesthesia.
Clinical and epidemiological evidence indicates that prenatal life is
more sensitive than adult life to the toxic effects of methylmercury. The
first indications came from Minamata in the early stages of the
outbreak, where it was found that mothers who were only slightly
poisoned gave birth to infants with severe cerebral palsy. A similar
situation has been reported in the Iraqi outbreak of 1971-72 with
infants which had been prenatally exposed showing severe damage to
the central nervous system. Using the maximum maternal hair concentration during pregnancy as an index, the lowest level at which severe
effects have been observed was 404 fig g~l. More recent follow-up
studies with infants in Iraq have found evidence of psychomotor
retardation (delayed achievement of development milestones, a history
of seizures, abnormal reflexes) at maternal hair levels well below those
associated with severe effects.

14.2

Lead

Lead is a neurotoxin and the overt toxic effects of lead have been known
for many centuries. Probably the first reported cases of lead poisoning
due to environmental sources was a group of children diagnosed as
having lead palsy by clinicians at the Brisbane Hospital in Queensland,
Australia, at the turn of the century. A total of ten cases of lead
poisoning were found by Health Officials and it was later shown that
the source of the lead was lead-based paint which was turning to powder
on the walls of homes and railings.

Lead poisoning due to exposure to lead-based paints has affected a
large number of children over the years. Most of the epidemiological
studies have been centred on the USA. The disease is confined to
children, especially those living in inner city areas in dilapidated buildings with surfaces of flaking and peeling lead-based paint. Children
playing in the vicinity can take in particles and flakes of paint by
inhalation and hand-to-mouth activities. Certain children have a
craving for eating non-food items such as flaking paint. This habit
called 'pica' and the more normal hand-to-mouth activities of children
are capable of introducing excessive amounts of lead into the body: e.g. a
square centimetre of paint may contain over one milligram of lead.
The disease in the USA was neglected for a long time and the full
number of children who have suffered from lead poisoning and the
number at risk from the disease emerged over several decades. The
probable reasons for this time lag include the poor socioeconomic status
of the children in the high risk groups and the difficulties in recognizing
the disease. In 1971, the disease became fully recognized when the US
Government introduced The Lead-Based Paint Poisoning Prevention
Act'.
It is important to recognize that 40 years or so ago, a child was
diagnosed as suffering from lead poisoning only when acute encephalopathy was evident. Typically this was characterized by a progression
to intellectual dullness and reduced consciousness and eventually to
seizures, coma, and, in very severe cases, death. Although unknown at
the time, acute encephalopathy would only develop when lead concentrations had exceeded 80-100 ^g dl" 1 of blood. At about 80 /ig dl" 1
severe but not life-threatening effects of the CNS can be expected. Lead
encephalopathy in children is usually accompanied by peripheral neuropathy, especially foot drop, and general weakness. Acute renal damage is
common in severe cases.
Case Study 9. Effects of lead on neurobehavioural development in chil-

dren. The main focus of attention in recent years has been the effects
associated with blood lead concentrations (PbB) less than 40 jug d P 1 ,
which include effects of prenatal and early childhood exposures on
physical and neurobehavioural development of children. Behavioural
and attentional deficits as rated by teachers {e.g. disordered classroom
activity, restlessness, easily distracted, not persistent, inability to follow
directions, low overall functioning) have been significantly associated
with children's tooth and PbB levels. On the basis of these studies, it has
been concluded that neurobehavioural deficits may occur at PbB levels at
and below 30 /xg dl" 1 .
Less severe symptoms associated with lead poisoning which have been
recognized in recent years are deficits in neurobehavioural development

and effects on the synthesis of haem in the blood. These effects can be
expected to occur at blood lead concentrations of 20-30 /ig dl" 1 and
even less.78'79 It is now well established that levels below 25 jug dl" l have
effects on cognitive (reduction in IQ scores) and behavioural development in children. For blood Pb levels below 10-15 fig dl" 1 range, effects
are difficult to detect because of confounding variables and lack of
precision of analytical and psychological measurements. Animal studies
support a causal relationship between Pb and nervous system effects and
there are reports of intellectual deficits in monkeys and rats with blood
Pb levels in the 11-15 jug dl" 1 range. However, attempts to attribute
subtle deficits in child development to lead exposure is controversial as
many other factors (genetical, nutritional, medical, educational, and
parental and social influences) can strongly influence the development of
a child. The developing nervous system is particularly sensitive to the
toxic effects of Pb and experimental studies have shown that a large
number of the effects in the nervous system are due to interference by Pb
with biochemical functions dependent on calcium ions and impairment
of neuronal connections dependent on dendritic pruning.80
Prenatal exposure to lead has been associated with a reduction in the
mental development of infants. The effect of environmental exposure to
lead on children's abilities at the age of 4 years was studied in a cohort of
537 children born between 1979 and 1982 to women living in the vicinity
of a lead smelter at Port Pirie in Australia. The study indicated that
elevated PbB concentrations in early childhood had deleterious effects on
mental development up to the age of 4 years.81
In the US and Europe measures have been taken to reduce Pb
exposure. These have included reductions of Pb in gasoline, in food due
to reduced use of Pb solder, removal of Pb from paint, and abatement of
housing containing lead-based paint. The US Food and Drug Administration (FDA) Market Basket Surveys indicate that the typical daily
intake of Pb for 2-year-old children has dropped from 30 fig day" 1 in
1982 to about 2 fig day" 1 in 1991. The National Health and Nutrition
Examination Surveys (NHANES) of blood Pb levels in children of 15 years of age showed that in the 10 years between one survey (1976-80)
and the next (1988-91) mean concentrations had decreased to 78% from
15 jug dl" 1 to 3.6 fig dl" 1 . Continuing factors that enhance risk to Pb
78

' L o w Level Lead Exposure: T h e Clinical Implications of C u r r e n t Research', ed. H . L. Needleman,
Raven, N e w Y o r k , 1980.
79
Royal Commission on Environmental Pollution, Ninth Report, Cmnd. 8852, HMSO, London,
1983.
80
R. A. Goyer, Environ. Health Perspect., 1996, 104, 1050-1054.
81
A. J. McMichael, P. A. Baghurst, N. R. Wigg, G. V. Vimpani, E. F. Robertson, and R. J. Roberts,
New Eng. J. Med., 1988, 319, 468.

exposure, particularly during fetal life are low socioeconomic status, old
housing with Pb-containing paint, and poor nutrition, particularly low
dietary intake of Ca, Fe, and Zn. Prenatal exposure may result from
endogenous sources such as Pb in the maternal skeletal system or
maternal exposures from diet and the environment.80

15 CONCLUSION
Pollutants are not of a distinct type and all manner of physical and
chemical properties are represented in this broad group of substances.
Cause-effect or more specifically exposure-effect relationships form an
important basis for the perception and evaluation of risk or damage
from pollution. For many of the important pollutants, it is perhaps
surprising to learn that these basic relationships have yet to be firmly
established.
The problems stem from the complexity of the interaction of
pollution with the environment and with systems within organisms.
Cause-effect relationships cannot be derived from a single approach,
but they are developed from field and experimental investigations and
observations. In this regard it is important to establish a consistent
relationship between the measured effect and the suspected cause. The
observed association should have a reasonable biological explanation.
It should be possible to isolate the causal agent and to reproduce the
effect under controlled conditions. The cause should normally precede
the effect.
In the past, pollution studies were generally initiated in response to an
obvious problem and whilst most of the acute problems have been
identified and tackled, even today new chronic problems such as
endocrine disruption are still being uncovered. This is at a time when
there is increased obligation to prevent such events by implementing
hazard and risk assessment procedures. As preventive pollution control
strategies are implemented for existing forms of pollution and new
chemicals it is important that risk assessment procedures take account
of all possible risks. There is some doubt whether or not these are
sufficiently rigorous to tackle the relatively newly uncovered problem of
endocrine disruption. The concluding remarks of the Tenth Report of
the Royal Commission on Environmental Pollution are still relevant
today, 'an important feature of the type of long-term environmental
protection policy ... is that it should guard against creating situations
which, though they may initially appear innocuous, have the potential
for erupting disastrously—in other words, it should ensure that no new
"time bombs" are set'.

Questions
1. Explain the importance of exposure-response relationships in
pollution studies. Discuss how information derived from such
relationships can be used in the process of risk assessment.
2. Describe the main ecological changes that take place as freshwater
systems undergo acidification.
3. What is understood by the term critical loads or concentrations?
Demonstrate how the concept is used in the management and
control of air pollution.
4. Review the evidence which suggests that endocrine disruption is a
fundamental mode of action of organochlorine pollutants.
5. Using specific examples demonstrate how biomarkers can be used
to assess biological effects in wildlife.
6. Discuss the way in which epidemics of mercury poisoning in the
general population have been important in formulating doseresponse relationships.
7. Critically evaluate the effects of low-level lead exposures on
neurobehavioural development in children. What are the main
sources of lead contributing to lead exposure and how can they be
reduced?
8. Assess the fate of hydrocarbons following an oil spill into the sea
and evaluate the short and long term effects on wildlife.
9. With reference to experimental and field evidence formulate
exposure-response relationships between tributyltin and imposex
in gastropods.
10. Critically examine the effects of air pollution on respiratory health
in the general population.

CHAPTER 9

Managing Environmental Quality
ANDREW SKINNER

1 INTRODUCTION
The objective of this chapter is to link together the various facets of
environmental chemistry and pollution of the environment and relate
these to the legislative, economic, and social practices adopted by
society to improve and maintain the quality of the natural environment.
It will cover the assessment, both subjective and objective, of environmental quality, the setting of standards, legislative controls on pollution, and the way that these are enforced by government agencies. It
will also consider the way in which society is increasingly seeking
methods, beyond pure legislation, to improve environmental protection
by education and incentive. The topic is a large one and the approach
taken will be to discuss principles and issues rather than delving into
specific details.
What do we mean by environmental quality? How do we go about
measuring it and setting standards and criteria by which we may judge
whether our management is effective and meeting the minimum standards perceived necessary? The setting of standards goes well beyond
measures of the chemical quality of land, air, and water. Standards might
include measures based upon the health, size, and diversity of biological
populations or criteria based on aesthetics and collective personal
judgements about what is acceptable. Someone living near to an urban
watercourse might well consider that the number of supermarket trolleys
to be seen in and around the river was a more relevant measure of the
quality of their environment than the chemical quality of the water in the
river. People living near industrial sites may well feel that noise and smell
were the most important characteristics describing the impact of the
process on their environment, rather than the chemical quality of the
emissions. These considerations lead easily to the conclusion that it is

most unlikely that any one method of assessment will satisfactorily
characterize environmental quality.
A good example of an attempt to establish multiple measures is the
General Quality Assessment used by the Environment Agency to
characterize the quality of surface waters.1 For surface fresh waters
the concept of separate 'windows' measuring chemical, biological,
nutrient, and aesthetic quality has been developed and is summarized
in Table 1.
A demonstration of how combining different quality assessments can
improve the insight into the total environmental quality and help focus
remedial action can be seen from a comparison between the chemical and
biological quality of urban rivers. It is commonly observed that urban
rivers are ranked higher in chemical quality than if the same watercourses are classified by biological means. The reason for this is that, as
sewage treatment standards are raised, the prime source of contamination in urban rivers comes from episodic events. A common example is
pollution caused during a period of intense rainfall in an urban catchment when accumulated organic pollutants from drains and storm
sewers is mobilized, causing a sudden and dramatic reduction in
dissolved oxygen and elevation of BOD and ammonia concentrations.
In chemical terms the quality of the river can quickly recover after such
an event but the fingerprint of pollution is retained by the longer
sustained impact on the biota, giving a lower biological score than
would be expected from the average chemical quality.
It is also necessary to remember that issues of quality can be strongly
linked to issues of availability. This is best demonstrated by the example
of water resources. Diminution of flow in watercourses, whether a
natural consequence of drought or whether artificially induced by overabstraction, may often be the most serious impact upon the environmental quality of the watercourse and upon the natural ecosystems that
depend upon it. The link to chemical quality is an indirect one, but may
be readily apparent, for example, where an authorization for an effluent
discharge has been set assuming a particular flow regime which is not
being achieved. The reduced dilution beyond that planned will lead to
poorer downstream chemical quality.
These examples demonstrate the complexity of the problem and the
need for a comprehensive view of environmental media and the links
between them if informed judgements and ones that society will accept,
are to be made about the quality of our environment and the way in
which it is managed.
1

'The State of the Freshwater Environment in England and Wales', Environment Agency, HMSO,
London, May 1998.

Table 1 Environment Agency General Quality Assessment—windows on water
quality
Window Characteristics
Chemical

The chemical scheme is a six-grade assessment against chemical standards
for dissolved oxygen, biochemical oxygen demand (BOD), and ammonia.
Ammonia and BOD are indicators of pollution, which apply to all rivers
because of the ubiquitous nature of the risk of pollution from sewage or
farms. Dissolved oxygen is essential to aquatic life. High BOD and ammonia
concentrations can lead to low DO concentrations so these concentrations
are also important in assessing water quality. Some forms of ammonia are
also toxic to fish

Biological The biological scheme is based on the groups (known as taxa) of macroinvertebrates that are found on the riverbed. Macroinvertebrates are used
because they do not move far, have reasonably long life cycles and respond
to the physical and chemical characteristics of the river. They respond to
pollutants which occur only infrequently and which are not measured by the
spot sampling procedure used in the chemical GQA scheme. For GQA
assessment, species of macroinvertebrates are linked together into 85 taxa.
These are given scores of 1 (for pollution-tolerant taxa) to 10 (for pollutionsensitive taxa). The groups are purely taxonomic and assume that the
members of each group have similar pollution tolerance. By comparing
taxa found in the sample with those you would expect to find if the river were
pristine, rivers are classified into one of six grades
Nutrient

The Environment Agency is testing a pilot scheme for phosphorus based on
average concentrations of orthophosphate, measured as phosphorus. The
six-grade classification is based upon standards related to differing levels of
phosphate concentrations and is subject to review in the light of experience.
The grade boundaries have been set so that significant changes in orthophosphate inputs to rivers will be reflected in a change of grade and to
contribute to the definition of eutrophic waters required to implementing
the EC Directive on Urban Waste Water Treatment (91/271/EEC)

Aesthetic

The aesthetic quality of a river is determined by a mix of perceptions
including the clarity of the water, odour, stagnation, colour, and the
presence of oil, litter, or foam. The Agency scheme makes assessments by
surveying the river and its bank. A standard method has been devised. Sites
are assessed on one or both banks, depending on public access. The number
of items of litter are counted, a visual inspection is made of the cover by oil,
foam, fungus, and ochre, and the colour and odour of the water is noted.
Each type of measurement (litter, odour, and colour) is graded from 1 to 4
and each is given a weighted score for each class according to its acceptability based on the findings of a public perception study. Sewage litter is
weighted as the most unacceptable. The site is then graded from 1 to 4,
described as Aesthetically Good Quality, Fair Quality, Poor Quality, and
Bad Quality respectively

Table 2

Objectives, standards, and limits

Objective

A general environmental objective, often described in terms of more than
one environmental standard and possibly involving staged targets

Standard

A measure of quality to be achieved in the environment, generally measured
as a concentration in one or more environmental media
A measure of the quality of an emission into the environment, generally
expressed as concentrations of pollutant and volume of effluent or as their
load per unit of production

Limit

2 OBJECTIVES, STANDARDS, AND LIMITS
The setting of criteria to manage environment quality can be best
envisaged as a hierarchy of objectives, standards, and limits. The specific
terminology used has varied widely depending upon methods and
traditions of pollution control in different countries and for different
media. Integration of pollution control procedures and the influence of
EU Directives are now driving convergence of terminology, which is
summarized in Table 2.
A good description of how this has been tackled in one particular
situation can be found in the United Kingdom Government's National
Air Quality Strategy.2
2.1

Environmental Objectives

An environmental objective may be very general, for example 'to render
polluting emissions harmless', but to be helpful objectives have to be
more specific and, in general, be set within a time frame for achievement.
The objective will usually contain reference to a numeric standard or set
of standards that provide a benchmark against which the achievements
or objectives can be monitored. Examples of objectives used in the
United Kingdom for air and water objectives are given later in the
chapter.
Environmental objectives are achieved in practice by regulatory
pressures working in combination with changes driven by various types
of economic or social measures, for example tax incentives, social
changes, and corporate commercial policies. Different organizations,
including Government itself, will have different roles to play if objectives
are to be achieved. In order to give these objectives force in the policies
and priorities of the various organizations involved, it is usual for them
to be accorded some kind of statutory or quasi-statutory status. One of
2

'The United Kingdom National Air Quality Strategy', HMSO, London, March 1997.

the prime purposes of the United Kingdom National Air Quality
Strategy2 and the United Kingdom National Waste Strategy3 is to
provide a formal basis for setting, monitoring, and achieving standards.
National and Local Government policies and the activities of the
regulatory agencies are required at all times to support and encourage
the achievement of the objectives set.
In the case of water quality objectives in the United Kingdom, the
Water Resources Act 1991 gave the Government powers to set statutory
objectives independent of a national water strategy. The Surface Waters
(River Ecosystem) (Classification) Regulations 1994 introduced a
scheme whereby river stretches could be assigned one of five classes
relating to the type of ecosystem that should be maintained in that
stretch. The ecosystem approach was used to provide a generic status
which characterizes the 'use' to which a particular stretch of river might
be put and recognized the inherent differences in quality expectations
between, say, an upland stream and a lowland river with an urbanized
catchment. These classes would become Statutory Water Quality Objectives set by the Secretary of State after consultation. Once set, it would be
a duty on the regulator to ensure that the objectives were achieved at all
times (after the date set for compliance), as far as this lay within its water
pollution control powers. The scheme is intended to establish a formal
and open mechanism, involving public consultation, for taking decisions
to protect and improve river water quality. It is the intention that
following the introduction of the scheme to rivers, further schemes will
be developed for other types of controlled waters, such as lakes,
estuaries, groundwaters, and for more specifically defined uses, such as
abstractions for industry and agriculture, water sports, and specialized
and vulnerable ecosystems. The approach has presented both political
and practical difficulties and is yet to be implemented. It seems likely that
the current system of Non-Statutory Water Quality Objectives first
established in 1979, which has provided a reliable basis for planning
water quality improvements, will continue for some time yet.

2.2

Environmental Standards

An environmental standard will normally be specified as the concentration of a pollutant in one or more environmental media. It is set with
regard to scientific and medical evidence in relation to impacts upon
public health or upon natural ecosystems at a level of minimum or zero
risk. Where, as is often possible, a no-effect level of pollution for a
3

'Making Waste Work: A Strategy for Sustainable Waste Management in England and Wales',
HMSO, London, 1995. Command paper 3040.

particular substance can be identified, this will guide the setting of the
environmental standard. A large number of standards in relation to
public health derive directly from standards created by the World Health
Organization, although these may be, and often are, modified when
adopted into, say, European legislation. In the absence of adopted
European standards, as is the current case for some air quality standards
in the United Kingdom, then national standards will be adopted. These
standards, often called Environmental Quality Standards (EQS), are
those concentrations of pollutant considered to be acceptable in the
wider environment.
One way in which Environmental Quality Standards can be met is to
ensure that individual permits control all point source discharges into the
medium. These permits will contain emission limits appropriate to the
achievement of the standard, having regard to the dispersion and
dilution that will take place. The assessment of these emission limits will
demand a high level of information about the level and variability of the
ambient concentrations of the relevant pollutant as well as the level and
variability of the expected discharge if emission limits are to be reliably
set. For example, in the setting of consents to discharge to water in the
United Kingdom, a Monte Carlo simulation technique is used that
combines information on the statistical distributions of upstream river
and discharge flow and quality, and calculates the appropriate emission
limits to meet the downstream Environmental Quality Standard. The
standard is therefore set in terms of the environmental capacity of the
receiving medium rather than on the basis of what may be achieved by
the process producing the emission.
2.3

Emission Limits

An alternative approach is to base the regulatory strategy around
emission limits, using criteria based upon the capability of the technology. This leads to uniform emission standard for a process or a site,
based upon assessments of process and pollution abatement techniques.
This approach has the advantage of setting uniform environmental
standards for any particular industry or process and is therefore
favoured by industry in providing the much sought after 'level playing
field'. It is also an approach that can be used without a wide knowledge
of the quality and variability of the receiving medium. However, where
such information is available, a sophisticated application of the emission
limit approach allows local variation. In Germany, for example, uniform
sectoral (that is, specific industry-wide) standards applying to discharges
to all media have been established in consultation with industrial
representatives and independent technical associations. Uniform emis-

sion standards are derived from assessments of the state of the art in the
process and pollution abatement technology for dangerous substances
and non-biodegradable substances. These have become binding standards that translate directly into process authorizations.
Uniform emission standards are relatively easy to apply. An environmental capacity based system is harder to administer particularly if, as
may be the case for air pollution control or for water pollution control to
large river catchments such as the Rhine, the issue comes under the
jurisdiction of a number of national regulatory agencies. However where
environmental capacity is limited, for example in the case of small rivers
with little dilution capacity, a uniform emission standard approach could
easily lead to problems if account was not taken of the limited
assimilative capacity of the receiving medium.
Technology-based emission standards offer a practical way of comparing alternative options for emissions to different environmental
media. They are a key element of the Integrated Pollution Control
(IPC) regime operated in the United Kingdom. The concept of IPC was
developed by the Royal Commission on Environmental Pollution in its
12th (1988) report.4 They recommended that an IPC inspectorate be
established, empowered to 'impose technology-based controls designed
to achieve a Best Practicable Environmental Option (BPEO) for the
specific process or waste' having consideration of all environmental
media. The definition that the Royal Commission gave for BPEO was
'the outcome of a systematic consultative and decision-making procedure, which emphasises the protection and conservation of the environment across land, air and water'. The BPEO procedure establishes, for a
given set of objectives, the option that provides the most benefit or least
damage to the environment as a whole, at acceptable cost, in the long
term as well as the short term. The system of technology-based controls,
which defines this ambitious, although difficult to achieve goal, is known
in the United Kingdom as BATNEEC (best available technique not
entailing excessive cost). The approach still depends upon knowledge of
the quality standard in the environment and one difficulty in operating
this regime has been to establish emission standards for substances for
which no environmental quality standards exist. One approach to
solving this problem has been to base standards on established exposure
limits used to protect employees in the workplace (workplace exposure
limits) under health and safety legislation. Because the impacts in the
wider environment may involve longer exposure time to more vulnerable
individuals it has been considered appropriate to apply safety factors of
up to 100-fold to the workplace exposure limits. Standards formulated in
4

'Best Practicable Environmental Option', 12th Report of the Royal Commission on Environmental
Pollution, HMSO, London, 1988.

this way have been widely criticized by plant operators as being
excessively restrictive.
The IPC regime applied in United Kingdom under the 1990 Environmental Protection Act has been influential in the development of a
similar but more wide ranging system of controls now to be introduced
throughout Europe. An integrated pollution control and permitting
regime is to be adopted under the Integrated Pollution Prevention and
Control (IPPC) Directive adopted in 1996 (96/61/EC) and due for
implementation progressively from 1999. This directive is built around
the concept of best available technology (B AT) which in practice is very
similar in concept to BATNEEC. Controls on emissions under the
directive will be based upon BAT, which describes processes or operations that are 'the most effective and advanced' of those available and
which are selected 'bearing in mind the likely costs and benefits and the
principles of precaution and prevention'. They should be capable of
emission limit values designed to prevent impact on the environment,
and if that is not practicable, minimize emissions and the impact on the
environment as a whole. If BAT standards do not meet environmental
quality standards then standards tighter than BAT must be set to ensure
that the environmental quality standard is not breached. In practice BAT
will be established on a Europe-wide basis in the form of standard
documents drawn up by groups of process specialists. The European
adoption of the BAT concept is of great significance for environmental
regulation and will extend beyond the IPPC directive to other directives
and is expected to drive common European environmental standards.
2.4

Integrating Limit Values and Quality Standards

The IPPC Directive, as described above, brings together the approach of
limit values based on best available techniques (BAT) and strict environmental quality standards and will apply to a wide range of processes and
industries, considerably wider than the IPC regime in the United
Kingdom. This recognition that two approaches can coexist has been
long coming and has been most evident in the development of European
water quality regulation where three strands are now about to merge.
2.4.1 Use-related Approach. Most of the early EC Directives on water
laid down environmental quality standards for individual parameters
designed to protect specific water uses such as abstraction for drinking
water supply (75/440/EEC), bathing (76/160/EEC), fish life (78/659/
EEC), and shellfish life (79/923/EEC). Member States are required to
designate waters under the Directives. For designated water the standards in the Directive are legally binding.

2.4.2 Uniform Emission Standards. The European Directive on dangerous substances (76/464/EEC) first recognized the dual approach and
allowed Member States to apply either uniform emission standards or
environmental quality objectives for the control of listed substances,
which are toxic, persistent in the environment, and bioaccumulative.
2.4.3 Sectoral Approach. A sectoral approach has been applied more
recently in the Directives on titanium dioxide (78/176/EEC and others),
nitrate pollution from agriculture (91/676/EEC), and urban waste water
treatment (91/271 /EEC). These directives require specific control measures appropriate to the special need in each circumstance, including, in
the latter two cases, the designation of zones where special preventative
measures must be taken.
It\ is the objective of the European Commission to simplify the many
individual directives covering water into a Water Resources Framework
Directive, which, together with the IPPC Directive, will provide a more
consistent framework within which increasingly integrated environmental management can operate. Table 3 gives a list of the large number of
current EC directives directly relating to water quality, many of which
will be modified or replaced by the proposed framework directive and by
the introduction of the IPPC regime.
2.5

Specifying Standards

Quality standards may be specified in a number of different ways, for
example:
• as absolute standards to be achieved at all times;
• as standards to be met most of the time, in this case usually
expressed as a percentile compliance; or
• as standards to be met as an average over a period.
Depending on circumstance, in particular the frequency of change and
the vulnerability of the environmental target, the period over which an
average may be taken can vary greatly. Some air quality standards
specify 15 min averages, whereas others may be set as an annual average.
Examples are shown in Table 4, which lists some of the United Kingdom
air quality objects for specific pollutants.
Absolute standards, if translated directly into emission limits, place
high demands on process designers and their operators and also on
monitoring systems. There are some circumstances, in terms of protecting human health from toxic substances, where absolute standards are
necessary, but in many cases properly specified and controlled percentile

Table 3

List of current EC Directives relevant to water pollution

Directive Number EC Directive title
75/440/EEC
79/869/EEC
76/464/EEC

Quality of surface water abstracted for drinking
Pollution caused by the discharge of certain dangerous substances
into the aquatic environment

76/160/EEC

Quality of bathing water

78/176/EEC

Waste from the titanium dioxide industry

78/659/EEC

Quality of fresh waters needing protection or improvement in order
to support fish life

79/409/EEC

Directive on the conservation of wild birds

80/68/EEC

Protection of groundwater against pollution caused by certain
dangerous substances
Procedures for the surveillance and monitoring of environments
concerned by waste from the titanium dioxide industry
Limit values and quality objectives for mercury discharges by the

82/883/EEC
82/176/EEC

chlor-alkali electrolysis industry
85/513/EEC

Limit values and quality objectives for cadmium discharges

84/491/EEC

Limit values and quality objectives for discharges of hexachlorocyclohexane
Limit values and quality objectives for mercury discharges by sectors
other than the chlor-alkali electrolysis industry
Limit values and quality objectives for discharges of certain dangerous substances included in List I of the Annex to Directive 76/464/
EEC (DDT, PCP)

84/156/EEC
86/280/EEC

88/347/EEC

Amending Annex II to Directive 86/280/EEC on limit values and
quality objectives for discharges of certain dangerous substances
included in List I of the Annex to Directive 76/464/EEC (aldrin,
dieldrin, endrin, hexachlorobenzene, hexachlorobutadiene, chloroform)

90/415/EEC

Amending Annex II to Directive 86/280/EEC on limit values and
quality objectives for discharges of certain dangerous substances
included in List I of the Annex to Directive 76/464/EEC (dichloroethane, trichloroethane, perchloroethane, trichlorobenzene)

91/676/EEC

Protection of waters against pollution caused by nitrates from
agricultural sources

91/271 /EEC

Urban waste water treatment

Table 4

Examples of UK air quality objectives

Pollutant

Objective

Nitrogen oxides

To achieve the standards of 150 ppb hourly mean, and 21 ppb
annual mean, absolute compliance, by 2005
Ozone
To achieve the standard of 50 ppb, measured as a running hourly
mean, 97th percentile compliance, by 2005
Particulate matter To achieve the standard of 50 ^g m~3, measured as a running 24 h
mean, 99th percentile compliance, by 2005
Sulfur Dioxide
To achieve the standard of 100 ppb, measured as a 15 min mean,
99.9th percentile compliance, by 2005

standards provide an adequate level of environmental protection at less
cost, both to operator and regulator. It is possible to set emission limits
incorporating both concepts. Discharge consents for sewage works in the
United Kingdom are set on a percentile compliance basis, but this can be
supported for critical determinands by a so-called upper tier, set as a
multiple of, say, three times the percentile limit value, which acts as an
absolute standard.
Because of the wide variety of standards and methods of expressing
them it is not a trivial matter to understand their impact and even to
know, when a standard is reformulated in a different way, whether it
represents a tightening or a relaxation. A good example in the United
Kingdom has recently emerged with the changes made by the United
Kingdom Government to its Nitrogen Dioxide Air Quality Standard. In
December 1996 the environmental magazine ENDS carried an article
'UK Goes It Alone on Weaker NO2 Standard', claiming that a revised
standard to be set in the National Air Quality Strategy was 'considerably
less stringent' than that set in the draft strategy. The new standard, now
published2 is an hourly mean NO2 concentration of 150 ppb with one
hundred per cent compliance. This revision was based upon advice from
the Government's Expert Panel on Air Quality Standards. The previous
draft standard had been an hourly mean concentration of 104.6 ppb but
with 99.9 per cent compliance and was based upon the WHO recommended standard. It is not a matter of simple assertion, based on a
comparison of the numerical values, that revised United Kingdom 1997
is less stringent because of the need to take account of the tighter
percentile compliance.
It is a matter of principle in setting standards that the criteria are those
of protecting public health and the environment. Depending on the
availability or certainty of toxicological or epidemiological evidence, the

approach to standard setting will be different. In those circumstances
where it is possible to identify a no-effect threshold, at or below which
effects are most unlikely, even for sensitive population groups, then this
concentration becomes the standard. In other circumstances it will not be
possible to establish a zero risk level, in which case it is necessary to
follow a risk-based toxicological approach to derive acceptable risk
exposure limits. This approach will apply particularly with carcinogenic
substances but, as with the case in setting the United Kingdom air
quality standard for particulates, the same approach can be applied for a
non-carcinogenic substance, if the available data shows that a no-effect
threshold cannot be identified.
Issues of costs and benefits and technical feasibility are clearly relevant
to standard setting but should not, in principle, affect the setting of the
numerical standard. They may influence the political judgements about
the way objectives are set and the time-scale over which they need to be
achieved. Inevitably questions of cost and feasibility frequently get
bound up in political debates arising from environmental standards.
An example is the standard for nitrate in water. A standard of
50 mg I""1 nitrate as NO 3 was set as part of the EC Drinking Water
Directive (80/778/EEC). This has now become the environmental standard for nitrate in water because the same numerical standard has been
translated into the European Directive for the Control of Nitrates from
Agriculture (91/676/EEC). The standard of 5OnIgP 1 is based upon
epidemiological evidence relating to human health. It has faced many
challenges but it has not been changed. When the standard was first set in
the early 1980s there was not a great deal of information about the level
and variability of nitrate in rivers and groundwaters, although most
countries had examples of extreme values, not uncommonly exceeding
200 mg I" 1 NO3, most often from shallow groundwater sources. The
intensive monitoring that has been done over the last fifteen years has
shown that, while nitrate concentrations in surface and groundwaters
regularly and widely exceed 5OnIgP 1 NO3 and are still increasing,
exceedance of 100 mg P 1 NO3, which was the original World Health
Organization standard, is rare. The impact of setting a standard of
5OnIgP 1 NO3 rather than 100 mg I" 1 NO3 has been to impose
significant costs on the water industry in Europe to treat or blend many
water supplies. As member states of the European Union progressively
implement the Nitrate Directive, significant changes in agricultural
practice in vulnerable areas will become necessary. The best medical
evidence is that the standard is correct at 50 mg P ] but there is no doubt
that had the earlier 100 mg P 1 NO3 standard been preserved the
compliance costs would have been dramatically less. The European
Drinking Water Directive sets a guideline value of 2 5 m g P ! NO3,

which influenced a number of member states to set objectives of less than
50 mg 1"l NO3 but the controversy about the standard has now put this
objective in question. For example, the Austrian Drinking Water Decree
of 1989, made before Austria was a member of the EU, set an objective of
achieving a nitrate standard of 30 mg I" 1 NO3 by 1999. This has now
been abandoned on grounds of high cost of achievement and a lack of
epidemiological evidence to support the tighter standard.
Case Study 1: Urban waste water treatment (Germany). This case study
illustrates the interrelation between environmental and technologybased standards and how, in practice, new discharge limits can be
imposed.
A sewage treatment plant in Germany serving the 60 000 population of
the town of Bensheim (Hessen), on the east side of the Rhine Valley
needed to be upgraded to include a two-stage activated sludge process.
The receiving stream was very small and provided little dilution for the
effluent, even in wet weather. During the dry season the effluent from the
treatment plant is the source of the stream.
The municipal authorities were notified by the regulatory agency that
additional tertiary treatment would be needed, and that the following
standards would be set as annual means in the revised discharge consent:
Chemical oxygen demand
Biochemical oxygen demand
NH 4 as N
NO3 as N

20 mg 1~l
5 mg 1~l
2mgP1
gmgr1

The above values are significantly more stringent than the legally binding
minimum (Best Technical Means) requirements for sewage treatment
plants the German Federal Government has established. They were
imposed because of the need to meet the minimum quality standards
relating to the target biological class II of the German system shown
below in a situation of low or zero dilution.
Chemical oxygen demand
Biochemical oxygen demand
NH 4 as N

> 6 mg 1 ~ l
2-6 mg l~l
<0.3mgl"" 1

The municipality of Bensheim, which owned and operated the plant,
appealed against this notification, arguing that the requirements were
too demanding for their treatment plant and exceeded the minimum
technology-based requirements they were obliged to meet. In particular
the standard for chemical oxygen demand of 20 mg l~l, which compared
with 90 mg I""1 Federal minimum requirement, would impose substan-

tial additional costs for the construction and operation of a filtration
plant. The appeal delayed progress during which time the existing plant
continued to affect water quality in the receiving stream.
Following the enactment of the European Urban Waste Water
Treatment Directive (91/271 /EEC) it became necessary to upgrade two
other sewage treatment plants serving nearby communities of 10 000 and
20000 inhabitants. A study of the situation showed that the economically and environmentally optimal solution would be to close down the
two smaller plants, serving the nearby municipalities of Lorsch and
Einhausen, and extend and upgrade only the Bensheim sewage works.
Lorsch did not agree to the connection but it was decided to enlarge
the Bensheim plant and to connect Einhausen to it. In accordance with
the Best Technical Means in sewage treatment for this size of plant, the
following standards were set in the discharge consent:
Chemical oxygen demand
Biochemical oxygen demand
NH 4 as N
NO 3 as N
Ptotai

30 mg \~l
6 mg \~l
2mgl~1
lOmgr1
0.2 m g r 1

The plant will now operate under these consent conditions before an
evaluation takes place, possibly leading to further measures being
imposed in order to achieve the biological water quality class II.
Case Study 2: Discharge of dangerous substances (United Kingdom).
This case study concerns emissions of the dangerous substance pentachlorophenol (PCP) from textile finishing industries in the Irwell catchment in North-West England. PCP is controlled as a dangerous
substance under European Directive 86/280/EEC, within the framework
of Directive 76/464/EEC (see Table 3). Member States are required to
control emissions of PCP through an authorization process.
At the time Directive 86/280/EEC came into force, monitoring of the
Irwell catchment showed a number of rivers to contain levels of PCP
exceeding environmental quality standards. There were consequential
effects upon river biota. Monitoring of PCP levels downstream from
sewage treatment plants gave PCP values in the range 3 to 21 fig I""1.
Investigations revealed 16 textile finishing companies in the River Irwell
catchment with effluents containing PCP, all discharging their trade
effluents to sewers.
Discharge consent standards from the sewage works to the river were
required to ensure that the environmental quality objective of 2.0 /ig l~l
as an annual mean value was met in the receiving water. Because of the
limited dilution in the River Irwell catchment, conditions for sewage

works effluents were set with maximum PCP concentrations between 2
and 6 fig 1~l. This in turn led to water company trade effluent control on
the discharges from the individual works to the sewer being set at levels
of between 6 and 10 fig l~ l in order to meet the environmental quality
standard in the river. It is notable that these levels are much more
stringent than the limit value of 2000 /^g I" 1 specified in the Directive.
As a result of measures taken by the textile industry to reduce PCP
emissions, annual mean levels of PCP in the River Irwell catchment are
now between 0.06 and 0.15 jag I" 1 . The reductions in PCP in the
discharge were achieved by better waste water treatment facilities in the
textile factories but, in particular, by better quality control on the raw
materials being taken in to the factories. Imported material in particular
was found to have varying and often high levels of PCP.
A very similar set of circumstances existed in the Worcestershire Stour
in the English Midlands, as a result of discharges from the carpet
industry centred on Kidderminster. Again, quality control on imports,
plus better effluent treatment, encouraged by the setting of appropriate
discharge controls, has improved the discharge and significantly
improved water quality which now meets the environmental quality
standard and has improved the diversity of aquatic life in the Stour.
2.6

Remediation Targets

The discussion so far has related to the setting of environmental
standards to prevent pollution. A key question is whether the same
standards necessarily apply when pollution has already occurred and a
judgement is required on whether, to what level, and with what priority
remediation should be carried out. These issues come into sharpest focus
in relation to the remediation of contaminated land, which also impacts
on the protection of water quality. Risk-based approaches to the setting
of targets are increasingly being applied in this sector. In remediating
contaminated land the most important factors relate to the protection of
the human health of those using the land, for the protection of water
quality draining from the land and for the protection of biota and
habitat. The earliest experience of contaminated land remediation at any
scale were those carried out in the United States in the 1980s under the
impetus of the 'Superfund' legislation, where federal finance was allocated for clean-up of what were perceived to be the major environmental
problems. Massive remediation projects were commissioned using both
landowner and federal funds to try and achieve environmental standards
often equivalent to natural ambient quality. These were, in many cases,
unachievable because of lack of available technology and also of
questionable necessity at the high cost specified. The same philosophies

began to be applied also in Europe. An example is the 'multi-functionality' approach adopted by the Dutch Government in its Soil Protection
Act of 1987, which established a remediation regime based upon cleanup standards intended to make polluted soil fit for any use.
The United Kingdom Government's approach has been very different.
This is set out in the provisions of the 1995 Environment Act and the
related draft guidance5 which requires remedial action to land contamination only where:
the contamination poses unacceptable actual or potential risks to
health or the environment; and
there are appropriate and cost-effective means available to do so,
taking into account the actual or intended use of the site.
This has been called the 'fitness for use' approach and leads to userelated targets being set for remediation of land only where it can be
shown that there is risk of environmental harm to people, water or biota.
In other words land which is contaminated but where the contamination
is not mobile and placing a human or environmental receptor at risk
need not be remediated unless or until circumstances and change and risk
arises. This therefore leads to the concept of latent contamination; not
desirable but not a priority for action. This approach has led to the
development of a concept of guideline values for soils in different
environmental settings, which are used to define risk and therefore set
criteria for remediation.
The United Kingdom approach has been widely criticized both inside6
and outside the United Kingdom but it is finding increasing favour. In a
major change of policy the Dutch Government announced, in May 1997,
that it would be following a 'fitness for use' approach and that its
priorities for remediation would be based more on redevelopment needs.
The prime reason for this change is the rising cost of the remediation
programme as more data become available about the scale of soil and
groundwater contamination, now estimated to be 100000 sites in The
Netherlands with a clean up cost of US$50 billion. This is a hundred-fold
increase on estimates made in the early 1980s when the policy was set. In
addition there is the realization that the problem will take many years to
solve, during which time the size of the problem will increase as polluted
soils continue to contaminate groundwaters. Significant changes of
approach are now being adopted in the USA, for example in California.
In late 1966 the state authorities changed its programme of tackling
5

'Draft Statutory Guidance on Contaminated Land', Department of Environment, London, 1996.
'Sustainable Use of Soils', 19th Report of the Royal Commission on Environmental Pollution,
HMSO, London, 1996.

contamination caused by underground fuel tanks from a comprehensive
remediation programme attempting to tackle every known site to one
based upon known problems or identified risk to human health or water
supply.
Protocols to provide a framework for the application of what has
come to be called Risk-based Corrective Action (RBCA) are now under
active development in many countries. These techniques use a tiered
approach to target setting for remediation, whereby decreasingly precautionary targets are set, as more information is available from survey
and monitoring.
2.7

The Principles of No Deterioration and Precaution

Pollution control regulation based around fitness for intended use has
been seen by many as philosophically unsatisfactory because it is seen to
give preference to the short term exploitation of the environment rather
than its protection over the longer term. In the case of use-related
standards there will be a threat, it is argued, of a lower level of protection
and a decline in quality of the environment, if a use on which an
environmental standard was based were to be discontinued. These
concerns are met by two underpinning principles. The first is of no
deterioration, which means that, where a quality standard has been set,
even if on a use-related basis, no new release to the environment would
be permitted which would allow the quality standard to fall below the
established threshold. The precautionary principle establishes that,
where doubt exists as to the environmental consequences of a proposed
new action, an approach based upon caution, supported by the available
evidence and avoiding risk of irreversible impact, should be adopted. In
various formulations these principles are incorporated in pollution
control law in most developed countries. Its application in Europe is
described further in Section 3.2
3 LEGISLATION TO CONTROL AND PREVENT POLLUTION
Pollution prevention legislation has evolved along a variety of routes in
different countries. Even in the USA and Europe, where there is strong
unifying legislation, there is still great diversity in the priorities given to
implementing various aspects of the legislation. There is even greater
diversity in the number and relationships of the agencies that carry out
these tasks. This section will deal with the general features of pollution
control legislation and the following section with the agencies charged
with enforcing them.

3.1

Origins of Pollution Control Legislation

Pollution control legislation in all countries emerged from concerns
about the impacts of both industrial processes and the disposal of
human wastes on human health. These gradually evolved to increasingly
embrace a specific environmental agenda distinct from health and safety
legislation. The duties were vested in a number of agencies, such that
there was no integrated control linking pollution of air, water, and soil
and land but most countries, in various ways, have now achieved a
significant degree of integrated multimedia environmental management.
The linkage between control of environmental pollution and health and
safety has recently come into closer focus, particularly in relation to the
regulation of major industrial plant. Incidents like the major industrial
plant disaster at Seveso in Italy led the European Community to adopt
the Control of Industrial Major Hazards and Accidents Directive
(CIMAH) (82/501/EEC). This is being replaced in 1999 by the Control
of Major Accidents and Hazards Directive (COMAH), the so-called
Seveso II Directive (96/82/EC), which is aimed at ensuring that operators
of industrial plant take all necessary measures to prevent major accident
hazards to both people and the environment. It will require joint
competency of health and safety and environmental regulators for its
implementation in European Member States. In some countries the
emphasis on protection of human health has led to environmental
regulation being closely linked to public health legislation. In an international context much of the impetus for pollution control initiatives are
driven by international and national health agencies. The World Health
Organization, as a source of health standards, is a key agency in the
setting and reviewing of environmental standards that depend upon
human health criteria.
In the United Kingdom the first significant step in the management of
industrial pollution came with the passing of the Alkali Act in 1863. This
created the Alkali Inspectorate, a direct forerunner, via Her Majesty's
Inspectorate of Pollution, of the Environment Agency and the Scottish
Environmental Protection Agency (SEPA). In 1874 a subsequent Act
introduced the concept of 'Best Practicable Means' to prevent noxious
emissions and render them harmless, a direct forerunner of the technology-based direct emission controls and BAT. The two Acts were
consolidated into the Alkali Act 1906 which was finally and completely
repealed only in 1996, when all remaining industrial process control
covering emissions to air came under the Environmental Protection Act
1990.
The first significant controls on water pollution in the United
Kingdom came with the Rivers Pollution Prevention Act of 1876 which

required sewage discharges to 'be rendered inoffensive' prior to discharge
to an inland watercourse. The Act was poorly enforced and did nothing
to address water pollution from industrial sources. A Royal Commission
on Sewage Disposal investigated the continuing problem of water
pollution from domestic sewage and reported in 1912. It recommended
emission standards for BOD and suspended solids related to the dilution
available in the receiving watercourse. Its recommendations, although
never given statutory force, had a very significant impact on standards. A
system of discharge controls to rivers was not established until the Rivers
(Prevention of Pollution) Act 1951, which was the first time industrial
pollution was addressed. These developments led, through successive
legislation and agencies, via the National Rivers Authority and the
Scottish River Purification Boards, to the water pollution control duties
of the Environment Agency and the Scottish Environmental Protection
Agency.
Control of pollution to land came very much later in the United
Kingdom in 1972 with the introduction of the Deposit of Poisonous
Wastes Act, which led quickly to more comprehensive legislation for
control of waste to land administered by Local Authorities (Control of
Pollution Act 1974, Part I). These provisions were extended and
incorporated in the Environmental Protection Act 1990, Part II and
responsibilities for waste regulation passed from Local Authorities to the
Environment Agency and SEPA in 1996.
Legislation to control and remediate land contamination not specifically covered by waste management or water pollution legislation,
although provided for in the Environment Act 1995, is not expected to
be fully implemented in the United Kingdom until 1999.
3.2

Trends in European Environmental Legislation

Within the European Community it is possible to see a significant
evolution in the style and scope of environmental legislation. The
Treaty of Rome, which established the European Community, saw
standardization of legislation on the environment as an aspect of
standardization of trading across the community. The Single European
Act of 1987 made various amendments to the Treaty, in particular
providing a new Chapter on the Environment in its own right (Articles
13Or, 130s and 13Ot). Article 13Or establishes that European action
relating to the environment shall take place to:
• Preserve, protect, and improve the quality of the environment
• Contribute towards protecting human health
• Ensure a prudent and rational utilization of natural resources.

Table 5

Treaty of the European Union (Article 13Or): Key principles in
environmental regulation

Sustainable

Policies should contribute to the goal of sustainable development
High level of protection Environmental damage should as a priority be rectified at
at source
source rather than by treatment at the point of use
Precautionary

Where outcomes are uncertain, particularly if they are likely to
be irreversible, then there should be a presumption in favour of
a cautious approach

Integrated

Decisions on environmental impact should have regard to
impacts and options for all media and be taken in an
integrated and holistic way

Subsidiarity

Decisions should be taken at the lowest level and at the most
local scale reasonable in the circumstances and recognize that,
although principles of environmental protection are general,
the diversity of situations applying in different regions will lead
to different practices being applied

It also establishes guiding principles for environmental management
shown in Table 5.
The early directives primarily focused on the control of toxic
substances in the environment, usually in relation to one medium
only and in general focused on point source pollution. The regulations
were based around specific limits on emissions linked to an environmental quality standard set for a range of listed substances. As
already described in Section 2 of this chapter, this approach provoked
a strong debate with those countries whose legislation, particularly in
the field of water pollution control, had evolved along a route of
environmental capacity, that is the ability of the environment to
assimilate polluting loads. More recently there has become increasing
international acceptance that the two approaches are not mutually
exclusive and can work in parallel, with the emission control concept
being more specifically relevant for toxic and persistent substances,
particularly those for which there is an international or global agenda
to seek reduction. One feature of European environmental legislation
that has precipitated this convergence of thinking is the move away
from specific process or substance controls to broadly integrated
environmental directives. The most significant in this regard is the
Integrated Pollution Prevention and Control Directive (96/61/EC)
which provides for a wide framework of integrated and environmental
management of a range of industrial, commercial, and agricultural
activities.

The Commission has announced the intention of incorporating a wide
range of existing directives relating to water management and water
pollution into a Water Framework Directive. This directive proposes to
combine, for the first time in any European legislation, the ecological and
chemical quality of surface waters and the chemical quality and quantitative status of groundwaters. It is proposed to develop a series of criteria
to define 'good' status for each of these, with a target of achieving this by
2010. This draft directive specifically embraces both emission control
and environmental quality objectives and will incorporate most of the
earlier dangerous substances and daughter directives. A further feature
of the proposed directive is to require monitoring, reporting, and action
plans to use the river basin rather than political boundaries as the
reporting unit.
In the area of waste management the European Commission adopted
Waste Framework Directives (75/442/EEC and 91/156/EEC). These
directives, together with the directives on defining hazardous wastes and
their status (91/689/EEC and 94/31/EC), define waste, encourage its
reuse, and control its movement and disposal. The packaging directive
(94/62/EC) seeks to further encourage waste minimization and resource
conservation by constraining waste production. A specific directive
relating to the landfilling of waste has been under consideration for a
number of years but has yet to be adopted. There is increasing
awareness of the complexities of the environmental and economic
arguments, which drive wastes to be recycled, deposited in landfills or
incinerated. The legislative framework for each of these, currently
included within three different directives or proposed directives, needs
to be closely harmonized if the right outcomes for the environment are
to be achieved.
3.3

Reporting Environmental Performance

One characteristic feature of most early pollution control legislation,
including the early European Directives, was that they were weak on
requirements for monitoring, reporting, and verification. It was often left
to the discretion of the pollution control agency to determine the extent
to which the implementation of legislation was enforced. Obligations to
undertake environmental monitoring were also few, thus limiting the
judgements that could be made about the efficacy of the pollution control
regime. This lack of a requirement for monitoring both implementation
and effectiveness has been seen to be a major stumbling block in both the
uniform application of environmental control regimes and the objective
assessment of future legislative needs. Although governments have been
slow to react to these concerns, public opinion has not. Pressure of public

opinion has become a major force in demanding and achieving better
environmental monitoring and better reporting. The outcome has been
much more rigorous attention being given to monitoring and reporting
in both European, and consequently in national legislation.
The European Commission first sought to address this problem by
promoting in 1991 the Standardized Reporting Directive (91/692/EEC)
to provide better information across Europe on compliance with thirty
major environmental directives. More recent directives have included
their own specific requirements for reporting. In 1995 a European
Environment Agency (EEA) was established with the specific initial
objective of reporting on the state of the environment in Europe
integrating, and over time forcing, standardization of national environmental databases and collating and reporting on specific environmental
topics. One of its first actions was to co-ordinate an extensive report on
the state of the environment in Europe, the Dobris Assessment.7 Such
assessments will be repeated from time to time. The EEA has established
a network of topic centres in various research and environmental
institutes around Europe specializing in different issues, including fresh
water, air quality, air emissions, and nature conservation. The Agency is
also establishing a European Environment and Conservation Network
(EIONET) to provide a database for European environmental information much of which will be assembled from the reporting requirements of
European directives together with the wider environmental monitoring
carried out by national agencies across Europe. These moves parallel in
many aspects the existing databases established by the EPA in the USA,
which has provided a model for many of these initiatives on public
information on the environment.
3.4

Pollution Control and Land Use Planning

A further aspect of legislation, which closely relates to the prevention of
pollution, is the control, which all developed countries have, in one form
or another, as part of their urban and rural infrastructure planning. The
United Kingdom has not been alone in relying for many years on
planning controls rather than pollution control legislation to prevent
the creation of future pollution threats. The convergence between these
different aspects of environmental management was recognized by the
United Kingdom Government in 1994 by publishing guidance on liaison
between planning and pollution control agencies.8 In many cases pollution prevention and the protection of the environment will be significant
7

'Europe's Environment: The Dobris Assessment', European Environment Agency, Copenhagen,
1995.
8
'Planning and Pollution Control', Planning Policy Guidance PPG 23, HMSO, London, 1994.

issues to be addressed in the Environmental Assessment required as part
of an application for planning permission.
An Environmental Assessment needs to be undertaken in support of
planning approval for a wide range of development which may have a
significant effect upon the environment. The statutory requirement
comes from the Environmental Assessment Directive (85/337/EEC),
first agreed in 1985 but modified extensively to increase the range of
topics coming within the scope of the Directive. An Environmental
Assessment requires the applicant to prepare an Environmental Statement and to assess, and make available for public scrutiny and comment,
the environmental aspects of a proposed new development. Environmental Statements are required to consider issues such as visual appearance and impact on the landscape as well as issues of environmental
quality. They are also required under a revision of the original directive
to contain an appraisal of options through an 'outline of the main
alternatives studies and an indication of the main reasons for the scheme
chosen, taking into account the environmental effects'.
Environmental assessments are in general complementary to and not
in duplication of the fulfilment of environmental protection legislation
and they provide an objective basis for decisions between alternative
strategies for meeting infrastructure development needs. There has been
concern expressed in many countries, including the United Kingdom,
about the lack of integration of environmental assessments under
planning regulations and BPEO assessments under pollution control
regulations. The concerns arise on account of the additional burden
placed on promoters of new schemes because of the duplication of
activity in satisfying the demands of different regulatory regimes (planning and pollution prevention), the potential for conflicting outcomes,
and uncertainty because of conflicts in timing between the two processes.
The European Commission has reacted to this criticism by making
provision within the IPPC directive for members states to run Environmental and IPPC assessments in parallel if they wish.
A further criticism of Environmental Assessments is that they take
place too late in the evolution of environmental strategies. Many of the
environmental issues and impacts caused by individual projects such as
roads or industrial developments stem from earlier decisions taken at a
higher decision-making level that cannot be tested at the project level.
This leads to the concept of Strategic Environmental Assessments
(SEAs) which could be carried out at an earlier stage at this higher level
and which would help to integrate environmental considerations into
policies from the outset.

4
4.1

POLLUTION CONTROL AGENCIES
Structure and Organization of Pollution Control Agencies

As we have seen the pollution control legislation in the United Kingdom
has developed piecemeal, whereby successive legislation has led to new
agencies being created, usually, but not always, independent of both
national and local government structures. Since 1996 these various
strands have now combined in the United Kingdom to form the
Environment Agency and the Scottish Environmental Protection
Agency, with comprehensive duties for pollution control as free standing
public bodies independent of government, known as Non-Departmental
Public Bodies (NDPBs). These agencies share some of their powers and
duties with Local Authorities particularly on contaminated land and
also, in England and Wales, on air pollution control. They carry out
environmental protection duties with a wide strategic and operational
remit, ranging from national policy development and direct interface
with Government and the European Community through to the management of the local field inspectorate. The Agencies operate to environmental boundaries, based on river catchments, rather than local political
boundaries and are fully independent of Local Authorities. The Agencies
combine in one organization environmental management of the whole
water cycle (integrated river basin management) and a wide range of
pollution control duties (integrated pollution control). They have direct
powers of prosecution through the courts without having to rely on any
form of political or Government legal intermediary or prosecution
service. The system in the United Kingdom is unique in Europe for its
comprehensiveness of technical functions, degree of independence, and
method of organization.
Many countries in Europe have created unified inspectorates but few
of them link water management and the full range of pollution control
duties, and few integrate their operational activities over river basins.
Most, especially in federal states, are part of the regional government
structure.
The UK approach has its supporters and its detractors. It is seen to
maximize the possibility that regulatory decisions will be taken consistently, objectively, and on the basis of sound science, independent of
local political influence. As national organizations the Agencies can, if
necessary in technical collaboration, ensure that, when required, the
highest level of technical expertise can be brought to bear on a problem.
In addition, because the Agencies have the bulk of their staff located in
their field inspectorates they can be responsive to local needs and can
easily collaborate in local partnerships with other agencies.

At a policy level, regional as well as national environment decisions
are intimately linked with social and economic decisions. Whereas it
is right that regulatory decisions should be taken on the basis of
sound science and technology, they should not be taken so remotely
that there is no public confidence in them and no public understanding of the technical rationale. The general public tend not to
understand the technical complexity of issues but have strong views
about the protection of the environment. If not involved they can feel
that their voice is not carrying due weight, compared with the
technically informed lobbyists from industry, who they might feel
have a greater ability to influence regulatory decisions if they are not
open to scrutiny. The issue of balance between sound science and
public perception and understanding are a challenge for environmental regulators in every country and are a natural and welcome
consequence of greater public knowledge of and concern for the
environment.
There is also no consistency from country to country as to the best way
of managing the environment in an integrated way. Some countries
combine management of natural habitat within the same agency as
pollution prevention and they might also combine related regulatory
activities like the protection of drinking water and aspects of public
health protection, which although not part of the natural environment,
have very strong links to it.
Countries also differ in the way in which they manage the water cycle
in comparison with other environmental media. The Environment
Agencies in the United Kingdom combine both the integrated management of pollution control and the integrated management of the water
catchment including water resources, fisheries, and flood defence within
one organization and within common operational units. This approach
is not replicated in many other countries in the world where basin
management, if it exists, is usually independent of pollution control or
is managed in discrete operational units. Achieving the optimum solution to complex environmental problems depends on an integrated
management of environmental media and a wide range of technical
specialisms. Organizations, which are the subject of regulation because
their activities impact on the environment, do not like dealing with a
plethora of agencies and officials. Both of these arguments favour the
creation of large multi-functional inspectorates but, as the above discussion reveals, there is no common view about the best way of organizing
these complex organizations so that they are effective, efficient and enjoy
public confidence.

4.2

Forestalling Pollution

Pollution control agencies have their origins in the need to enforce
pollution control legislation. Historically, pollution control legislation
has been reactive in character, setting standards, prosecuting breaches,
and thereby forcing corrective action and setting an example to others.
The philosophy behind the BAT approach to environmental regulation
is that it sets standards not only for emissions but also for waste
management and emission abatement technology. Compliance with
environmental standards is more about having processes that conform
to the needs of the environment rather than focusing on end of pipe
regulation. This leads to the concept of the enforcement notice by which a
pollution control inspector can issue a legally enforceable demand for a
procedure to be modified so that acceptable standards of environmental
protection can be achieved and risk of pollution forestalled. As a result
of the Environment Act 1995 a wide range of enforcement notice powers
are now in force, covering many aspects of industrial, commercial and
agricultural practices.
Enforcement notices are general powers that can be used at any
location where they are relevant without geographical restriction. There
are situations, however, where environmental risks are of a localized and
specific character and can be best addressed by special powers which
would not be appropriate for general application. A good example of
this is the restriction of land use activities within the catchment or
gathering grounds of water supply sources. There is a long tradition in
legislation in European countries for setting protection zones around
points of groundwater abstraction. This statutory protection zone
approach has not been applied in the United Kingdom largely because
of its inflexibility and a consultative approach using planning as well as
pollution control law, based upon a general groundwater protection
policy9 has been preferred. Recently, however, a system of protection
zones has been established under the EC Nitrate Directive (91/676/EEC)
which requires the creation of Nitrate Vulnerable Zones. These are zones
where agricultural activity has caused or is in the process of causing
pollution of either surface water or groundwater resources in excess of
the EC Nitrate Standard of 5OnIgI" 1 . Different countries in Europe
have implemented the directive in different ways. In Denmark, where
because of the prevailing geology water resources are particularly
vulnerable to agricultural activities and where virtually all of the public
water supply is taken from groundwater, there is no value in applying
9

'Policy and Practice for the Protection of Groundwater', HMSO, London, 1998 (National Rivers
Authority, 1991).

localized protection zones. Therefore Denmark has designated the whole
country as a Nitrate Vulnerable Zone. In the case of the United
Kingdom, and in most other Member States, there is significant local
variation in the scale of the problem. Therefore zones have been defined
on water quality and hydrological criteria to identify the catchments of
both ground and surface water supplies, where land use change is
necessary to mitigate the impact of agricultural practices on the concentration of nitrate in water resources. Case study 3 describes the approach
adopted in the United Kingdom.
Another example of targeting pollution control to localized problem
areas, this time in United Kingdom national legislation, is the setting up
under the United Kingdom National Air Quality Strategy2 of Air
Quality Management Areas (AQMAs). These are areas where air
quality objectives are unlikely to be achieved by existing regulatory
measures and special action, through a local action plan, is necessary to
force improvements in air quality through the existing planning and
regulatory regime. Case study 4 describes the first steps being taken in the
United Kingdom to set up these zones.
Case Study 3: Control of nitrate pollution from agricultural sources. The
1991 European Nitrate Directive is specifically directed towards the
control of nitrates derived from agricultural sources. The directive
applies to:
(1) surface waters used for human consumption, which currently have,
or are expected in the future to have, nitrate concentrations in
excess of 50 mg I" 1 NO3;
(2) groundwaters used for human consumption, which currently have,
or are expected in the future to have, nitrate concentrations in
excess of 50 mg \~l NO3;
(3) waters, which are eutrophic on account of their nitrate content.
There are three steps in the process to meet the requirements of the
directive.
— Firstly Member States, using information from special monitoring
programmes if sufficient data is not otherwise available, must
identify the water bodies—rivers, lakes or groundwaters—to
which the above criteria apply.
— Secondly they must identify the areas of land where agricultural
practice is considered to be the cause, or potential cause, of the
nitrate problems identified. These are called Nitrate Vulnerable
Zones (NVZs).

— Thirdly they must set up a series of controls or limits on agricultural practice to apply in the defined NVZ, which will limit nitrate
leaching and protect the water bodies identified.
Different Member States have adopted different approaches. Some
countries, such as Denmark, have designated the whole country as a
Nitrate Vulnerable Zone, which has removed many of the technical and
administrative problems experienced in other Members States in carrying out the first two stages. The demands of the directive are onerous and
are strongly contested by farmers who have lobbied strongly in all
Member States to influence the way the directive has been implemented.
In 1997 the European Commission published10 an assessment of the
implementation of the Directive and, to varying degrees, found fault
with procedures adopted in every Member State.
In the United Kingdom NVZs were defined by identifying the catchment areas of all river reaches or points of groundwater abstraction
where nitrate concentrations exceeded 50 mg P 1 NO3. In addition, for
groundwater sources, assessments were made of those showing rising
trends in nitrate concentration to allow for the long lag times when
nitrate leaches into underground strata. Catchments for surface waters
were defined from the topography; catchments for groundwater sources
were defined by using groundwater models to work out flow directions.
Where there were a number of adjacent groundwater sources with
coalescing zones of pumping, these were combined and the NVZ
boundaries where possible followed geological boundaries. No waters
were identified which were considered eutrophic under the restricted
definition of the directive.
This exercise led to the definition of 68 NVZs in the first phase in
England and Wales. These are shown in Figure 1. The directive requires
a four year revision which is likely to lead to further designations in 1998
as a result of more comprehensive data which has been collected in the
intervening period. Also in 1998, the restrictions on agricultural practice
came into force, requiring limitation on the amount and timing of
fertilizer use and also limiting or controlling other activities which have
been shown to cause excessive nitrate leaching.
Case Study 4: Introducing Air Quality Management Areas. The United
Kingdom Air Quality Strategy2 set air quality targets to be achieved by
the year 2005. The general expectation of Government is that these
targets will be met by national programmes of emission controls and
enforcement, planning controls, and a transport policy which reduces
10

'The Implementation of Council Directive 91/676/EEC concerning the protection of waters
against pollution caused by nitrates from agricultural sources', European Commission
COM(97)473, Luxembourg, 1997.

Nitrate Vulnerable Zones

Figure 1 Nitrate vulnerable zones

reliance on road transport for both social and commercial transport.
There remains considerable debate about both the standards and the
time-scale in which they are expected to be achieved but there is no
disagreement that there are some areas where national measures alone
will not be sufficient to ensure that the targets are met. Local authorities
are to be required to review and assess air quality in their areas. Where
they believe that air quality objectives will not be met by 2005 by national
measures alone, they must declare an Air Quality Management Area
(AQMA) and draw up a specific and locally relevant action plan
designed to achieve compliance.
The sorts of measures that might be put in place, some of which would
require the grant of additional powers to local authorities, are:

traffic management to control the number and frequency of vehicle
movements and to control the movement of vehicles with unsatisfactory emissions. There is a wealth of experience worldwide on the
use of flow control, restricted access, pedestrianization, parking
controls, automatic traffic control systems, the encouragement of
other modes of travel, and giving priority to high occupancy
vehicles to reduce traffic density, although the benefits in terms of
air quality are not well documented.
more restrictive controls of point source emissions and problem
industries within the AQMA.
control on sale as well as use of unauthorized fuels,
roadside vehicle emission tests with fixed penalty charges for
offenders.
access restriction for vehicles to sensitive areas, perhaps permanently or limited to certain times of the day or the year,
use of planning powers to restrict developments likely to impede
achievement of air quality objectives. Except in special circumstances the impact of planning controls is likely to be felt only over
a long time-scale although there is already an increasing number of
cases where development plans are being opposed on grounds that
traffic generated by the development would affect air quality. The
existence of an Air Quality Management Area would greatly
strengthen such an argument.
The ability to specify an action plan for an Air Quality Management
Area and to choose between the various options is limited by the lack of
data and predictive capability. For this reason, in advance of the main
legislation, there has been a programme to establish 14 trial areas in the
United Kingdom covering a variety of geographical locations and air
quality problems. The emphasis of the work in these areas has been on
data collection, validation of new monitoring techniques, and assessment
of models, in each case focused on the particular air quality problems
characterizing that area. Many of the areas have covered urban air
quality issues but problems in more rural areas, for example north-east
Derbyshire and Cornwall where particulates from quarrying are a
particular concern, have been included.
4.3

Other Regulatory Action

Even with effective legislation and enforcement it is inevitable that, from
time to time, pollution incidents will occur. They may arise because of a
breach, accidental or malicious, of pollution control law or they may be
due to unplanned circumstances, such as a road traffic accident involving

a vehicle with a chemical load causing spillage that pollutes a river or
aquifer. Regardless of the question of breach of environmental law the
first requirement is to remedy the immediate situation. For this reason
the Environment Agencies maintain an emergency response capability to
attend accidents and emergencies where the environment might be
placed at risk. Emergencies vary in scale and require collaborative
action with other agencies but, in all cases, the objectives are to protect
human life and health, protect property, and protect the environment.
This action will be combined with investigation of the causes of the
pollution.
An increasing role for environmental agencies is to encourage and
promote environmental best practice, not as a substitute for the enforcement of pollution control regulations but to extend the influencing role
beyond the negative and minimalist approach of avoiding prosecution.
To this end the Environment Agencies have worked extensively with
others to promote better environmental practice in industry, commerce,
and agriculture, using the benefits of their extensive technical expertise
and, on the ground, contact through environmental protection staff.
These initiatives have not only concerned themselves with minimizing
emissions but also with conservation of raw materials and environmental
resources and the minimization of waste. These sorts of activities are
increasingly coming within the scope of environmental legislation, in
particular through the EC directive on packaging (94/62/EC) implemented in the United Kingdom through the producer responsibility legislation. These wider objectives of environmental best practice in business
are fundamental to the development of Environmental Management
Systems, which are discussed in Section 6.
5 ECONOMIC INSTRUMENTS FOR MANAGING
POLLUTION
5.1 Alternatives to Pollution Regulation by Permit
In most countries, including the United Kingdom, operations (abstraction of water, discharge of gaseous or liquid effluents, movements and
disposals of waste) which have a potential to impact on the environment
require a permit (variously called an authorization, licence, or consent)
from the pollution control agency. At least part of the cost of operating
the pollution control permitting system is recovered by a system of
charges for issuing and maintaining the permits. A criticism of this
approach is that these permits can be regarded as licences to exploit the
environment rather than a means of protecting it. The charges for
permits are generally low compared with the costs of investing in

environmental management of pollution control systems that would
minimize or remove the need for the permit. The costs also often bear
only a limited relationship to the impact of the regulated activity on the
environment. In general the charging schemes will only offer a limited
incentive to responsible operators and no encouragement to the discharger to go beyond the minimum requirements of their consents even if
there was financial advantage to them doing so. Economic signals to
business about an improved environment are therefore rather weak.
As a result, commentators11 have suggested that the goal of sustainable development might be advanced if there were a change of emphasis
in favour of economic instruments as an additional means of achieving
environmental goals. Various forms of economic charge and incentive
have been shown to reduce the total impact on the environment and to
ensure that there is a more optimal take up of the capacity of the
environment to absorb pollution. Examples of these economic instruments are various forms of environmental or green taxes, which levy a
charge on particular products or processes to discourage their use or
encourage recycling. Closely allied to this is the concept of tradeable
permits whereby authorizations to use environmental resources, either to
discharge pollutants to the environment or to abstract water, are tradeable assets, which can be bought and sold like any other commodity.
This approach is intended to ensure that such environmental capacity as
is available, over and above limits set by environmental quality objectives, is used to the greatest economic benefit, thereby giving scarce
environmental resources a value, which is lacking under conventional
regulatory systems.
These ideas have been widely canvassed in academic and policy circles
and the most surprising thing is how little they have actually found their
way into existing regulatory regimes, especially in the United Kingdom.
Two reasons are quoted for the lack of readiness to introduce green taxes.
The values of environmental benefits are hard to assess and therefore it is argued that there is considerable uncertainty as to how well
any given tax will work, or even if it will have unexpected
detrimental effects.
It is a taxation principle in the United Kingdom that taxes should
not be 'hypothecated', that is the decisions about tax raising and tax
expenditure should not be linked. It has been argued that environmental taxes not spent to the benefit of the environment would not
command public acceptance. Green taxes have their critics who
11

'Report of the House of Lords Committee on Sustainable Development', HMSO, London, 1995,
Vol. 1.

question whether the basic motivation is to improve the environment or to be a convenient source of additional revenue for national
treasuries.
A further incentive to development in green taxation is that economic
theorists have begun to speculate that taxation to penalize pollution and
force investment into environmental improvement will not have a
negative impact on industry, which has been the previous assumption,
but will be a stimulus to the economy. This is the so-called 'double
dividend' of green taxation, both generating revenue and stimulating
growth and jobs in the environmental industries.12 These arguments
remain controversial and demonstrate the difficulty there has been in
getting environmental issues to a central position in policy thinking.
However during 1997 the UK Government made a number of statements
indicating an increased commitment to green taxation and in November
1997 published a consultation paper on economic instruments for water
pollution.13 Case study 5 below gives examples of situations around the
world where green taxes are in use.
In the United Kingdom various modest taxation measures with green
credentials have been established (VAT on domestic fuel, differentials in
duty between leaded and unleaded petrol). In October 1996 a landfill tax
was introduced which places a levy on the tonnage of waste deposited to
land. A further innovation of this tax is that arrangements can be made
for a proportion of the income to be returned, through a charitable trust
format, to finance environmental improvement projects, thus in some
way addressing the issue of returning the benefit of environmental taxes
to the environment. Early indications of the impact of this tax are that it
is sending important price signals, which over time are likely to shift
thinking away from waste production to various forms of waste
conservation. In some parts of the country there is evidence that it has
also increased the amount of illegal or opportunistic waste disposal
although this is likely to be a transitional impact as the industry comes to
terms with the new financial regime and is not expected to persist. In the
light of this experience it is now being suggested that an even larger tax
would send a stronger conservation signal.
One of the inherent objectives of a landfill tax is to increase costs of
landfilling and so direct wastes to various forms of recovery or reuse,
either as the product itself or through incineration as waste-to-energy
schemes. This is driven by the 1991 European Waste Framework
Directive (91/156/EEC) which required Member States to 'take appro12

'Environmental Policies and Employment', OECD, Paris, 1997.
'Economic Instruments for Water Pollution', Department of Environment, Transport and the
Regions, London, 1997.

priate measures to encourage firstly the prevention or reduction of waste
... and secondly the recovery of waste ... or the use of waste as a source
of energy'. This waste management hierarchy also features strongly in the
UK strategy.3 Preventing the production of waste is the main objective of
producer responsibility legislation in its various forms. A scheme
managed by the Environment Agencies to control and to meet reduction
targets in packaging materials was introduced in the United Kingdom in
1997, and many similar schemes, extending to a wider variety of
materials, are being developed in other European Member States.
Many recycling schemes fail to achieve their potential of preventing
wastes going to landfill because the economic incentive is missing, that is
the environmental waste hierarchy is not matched by a parallel economic
waste hierarchy. The likely way forward is that waste management
decisions will be based upon a hierarchy that is influenced by both
economic and environmental costs. Sustainability can be built into this
equation by using fiscal measures to affect the economic costs. Life Cycle
Analysis, which quantifies the environmental impacts 'from cradle to
grave' of products or processes, is one of the tools being used to help
assess waste management options and the influence of economic incentives upon them.
5.2

Tradeable Permits

Attempts to establish systems of trading in environmental licences have
moved very slowly despite the fact that this approach is strongly
favoured by governments who favour market orientated solutions
rather than conventional regulatory solutions. One of the problems of
instituting a system of tradeable permits is that it has to operate within a
defined environmental unit so that, as emission or abstraction authorizations are bought and sold, the net environmental impact can be
controlled without regulatory intervention. Trading of water rights
within a catchment or within an estuary fits this concept well. In many
cases the overall net impact on the environmental system of moving the
rights to abstract or discharge from one location to another might be
assumed to be small. However a way needs to be found to prevent
unacceptable impact in the situation, for example, that a right to abstract
or discharge at the bottom of a catchment were sold for use nearer the
headwaters where the environmental impact might be unacceptable. It is
harder to apply the same ideas to the trading of air emission permits
because the same concept of a bounded environmental pool either does
not exist or cannot easily be identified or may not be under a common
jurisdiction. In practice any scheme involving tradeable environmental
permits is only likely to be robust in the heterogeneous environmental

conditions which exist in most of Europe over small geographic areas
where the number of potential traders is likely to be small. Likely
exceptions are water abstraction in areas of intense agricultural use for
irrigation and discharges to estuaries.
Case Study 5: Examples of economic instruments with environmental
benefits. There are three categories of economic instrument, which can
be identified:
taxation schemes which are targeted at products or services and
which are designed to reduce use of potentially harmful substances
or otherwise reduce environmental impacts but where the revenue
generated is added to general taxation.
There are a number of well established models, in particular taxes
on acid gas wastes in Sweden, on toxic waste in Germany and on
various aspects of water pollution in The Netherlands. Nordic
countries have led the way in this regard and, for example, Sweden
is said to raise in total the equivalent of 6% of its GDP from green
taxes. Other taxes which have been, or are currently, under
consideration either at a European or Member State level are taxes
on VOCs, sulfur, and carbon, the last often called an energy tax.
taxation schemes which may have a desirable environmental
outcome in their own right but where some or all of the revenue
derived from the taxation is also devoted to environmental
improvement through subsidizing or promoting environmental
improvement.
The DETR report13 quotes examples from Germany and France
where water pollution charging schemes are used to provide
incentive for environmental improvement. In Germany discounts
on the charging tariff are available if the discharger meets or
exceeds the standards required under the best technical means
standards. In France a similar scheme also provides rebates for
meeting best technical means standards and generates revenue for
investment in wastewater treatment and pollution abatement
projects. Revenue from environmental taxation in Norway is
funding a scheme to separate carbon dioxide from the Norwegian
offshore gas fields and to reinject it back into the reservoir
formation to minimize the atmospheric impacts,
tradeable permits; systems of licensing and charging for environmental capacity (to discharge pollutants or to use renewable natural
resources) which can be traded and which, by creating a market,
encourage the optimal use of the environmental capacity.
Permit trading is well established in the field of water resources in
arid areas. Water use permits for crop irrigation depending upon

water transfer schemes incorporate various forms of rights trading,
annual licence, and even a rights auction to optimize the economic
return from the water transfer. Well established examples exist in
the south-west USA (California, Colorado River) and in Australia
(Snowy Mountain Scheme). Trading of water pollution rights is
much less well established. A number of studies have shown
theoretical cost savings and environmental benefits but in practice
the achievements have been very limited. One scheme for BOD
trading on the Fox River in Wisconsin was established in 1981 but
only one trade has taken place. In the field of air quality the greatest
consideration in the United Kingdom has been given to sulfur
trading as a means of controlling SO2 and meeting international
obligations with minimum commercial impact. To a degree this was
achieved by the IPC authorizations issued to power stations which
allowed a degree of trade between stations but the full application
of a trading approach has not been implemented because it could
not be guaranteed to be compatible with BATNEEC. The second
international sulfur protocol requires that the UK reduce sulfur
emissions progressively by 80% from a 1980 baseline by 2010. It is
now apparent that this will be achieved within the existing pollution
control regime without recourse to other means and the tradeable
permit approach will not be further extended. On the global scale,
however, one emission trading agreement has been reached. The
negotiations at the UN Kyoto 1997 World Climate Conference on
the control of greenhouse gas emissions involved a trade in national
emission quotas, which will allow the USA to offset its target
reductions by acquiring additional quota from other nations,
particularly Russia.
A United Kingdom Government report published in 1993 said 'although
theory suggests that economic instruments will save resources, the
limited use of economic instruments to date means that there is no
widespread body of empirical evidence to demonstrate this in practice'.
This view remains valid.
6 PUBLIC AND COMMERCIAL PRESSURES TO IMPROVE
THE ENVIRONMENT
6.1 Environmental Management Systems
So far we have considered the management of environmental quality in
terms of what Governments or agencies of Government can do through
legal and financial incentives to encourage companies and individuals to

not harm, and where possible, to improve the environment. These are
widely held public goals, so why do organizations not set environmental
objectives for themselves independent of or in addition to the demands of
regulation? The answer is, of course, that increasingly they do, both to be
certain that they achieve compliance with regulation and to meet wider
goals, particularly if they see competitive advantage or cost savings
accruing. The object of this section is to review various aspects of
environmental management systems and procedures, which are adopted
by organizations to meet this objective. The term Environmental
Management System (EMS) is the generic term used to describe such
procedures and is the subject of international standard ISO 14001,14
which from 1997 has replaced the equivalent British Standard BS7750.
The key reasons why a company would wish to develop an EMS are:
the requirement for regulatory compliance
the need to establish systems to ensure that compliance is maintained
financial incentives
market place pressures
green credentials
An important factor in the development of environmental management
systems has been the influence which companies have on their trading
partners, the so-called supply chain pressures. Major manufacturers and
traders—particularly in the automotive industry, which has been in the
forefront of quality management, supermarkets, and other parts of the
retail trade—have been a major force, through their own environmental
policies, in encouraging and often obliging their suppliers to adopt an
approved EMS.
The principal formalized EMS in Europe is the European Communities EMAS (Ecomanagement and Audit Scheme). This can be achieved
by meeting the ISO 14001 standard but also requires a verified public
environmental policy to meet the requirements.
The take up of EMAS within Europe has been very variable as shown
in Table 6. In four countries no sites had been registered by mid-1997,
more than two years after the scheme was initiated. The greatest take up
is in Germany which is generally believed to be because companies in
that country believe that registration under EMAS will reduce the formal
regulatory burden upon them. The position of the Environment Agencies in the United Kingdom is that they encourage companies to establish
14

'Environmental Management Systems—Specifications with Guidelines for Use', EN ISO 14001,
CEN, Brussels, 1996.

Table 6

Growth in number of EMAS Registered Sites in
EU Member States

Austria
Belgium
Denmark
Finland
France
Germany
Ireland
The Netherlands
Spain
Sweden
UK

February 1996

April 1997

6
2
3
O
3
113
1
2
O
1
9

56
3
17
4
7
518
2
13
4
45
26

Source: EC Official Journal.
None registered in Greece, Italy, Luxembourg and Portugal as at
1 April 1997. The scheme commenced in 1995.

an EMS but that this, in itself, is not seen as a substitute for environmental regulation. It is notable that one of the first companies to achieve
BS7750 accreditation in the United Kingdom was prosecuted for causing
water pollution shortly afterwards.15
There is good evidence from many companies that investment in EMS
is repaid in both tangible benefits, such as reduced operating costs, and
intangibles, for example improved safety, improved public perception,
and less risk of prosecution. However the achievement of EMAS or
ISO 14001 accreditation and the ongoing maintenance of the necessary
systems is a significant investment in time and resources, which many
smaller organizations are unwilling to commit. The benefits, in reduced
operating costs and reduced risk of causing pollution, of adopting
environmental best practice are still important to business and bring
with them an improvement to the environment. Therefore, following the
adage that 'prevention is better than cure', pollution control agencies are
increasingly devoting effort to education in pollution prevention and
good environmental management through special campaigns, on-site
advice and promotion of waste minimization and recovery projects.
6.2

Public Opinion and the Environment

There is a high level of general public awareness of environmental issues,
which is itself a positive force for environmental improvement. In rare
cases the weight of public opinion can be decisive. This was most
15

'First BS7750 holder to be fined for pollution', ENDS Magazine, Feb. 1997, 265, 43.

dramatically shown in the case of the proposed deep sea dismantling of
the disused Brent Spar oil platform by the Shell Company. This so
excited public opinion that a boycott of petrol sales by the company
throughout Europe, led by the environmental pressure group Greenpeace, forced the company to abandon their proposal. Less spectacular,
but with significant effect over the years, has been the close independent
monitoring of environmental issues by environmental interest groups,
providing reports to the European Commission of failures to implement
directives or prosecute breaches. Publicly accessible registers of environmental monitoring data provide an important source of public information on the state of the environment and demonstrate accountability to
the public for the way it is regulated.
There is also an increasing interest and demand by the public for a
close involvement in decisions affecting the environment. Public consultation takes place on applications for the issue of permits to release
waste into the environment. The style of this consultation has changed
significantly as public awareness and interest in issues affecting their
environment has increased. Pollution control agencies are increasingly
being required to engage in debate about the technical issues surrounding the granting of permits and provide greater public documentation of
their decisions and the reasons for them. Whereas traditionally newspaper advertisements were the medium to make the public aware of new
applications, some pollution control agencies are beginning to make this
information available on the Internet. Many issues which attract public
concern are those for which the technical assessments are complex and
involve judgements of level of risk often involving widely varying public
perceptions of what is acceptable. Regulatory decisions must be based
upon sound science but the decisions are being made on behalf of the
public and for the wider public good. Therefore pollution control
agencies cannot ignore issues of public perception of risk and must,
through consultation and information, seek to maintain the confidence
of the communities they serve while making regulatory decisions that
contribute to the goal of environmental improvement.

QUESTIONS
1. Discuss with examples the concepts of objectives, standards, and
limits in relation to environmental quality.
2. Compare and contrast the relative merits of pollution control
strategies based respectively upon environmental quality objectives
and uniform discharge standards, illustrating with examples as
appropriate.

3. Discuss in the context of air, freshwaters, and soil the appropriateness of applying a 'fitness for use' approach to environmental
quality.
4. Discuss the appropriateness and effectiveness of using taxation
measures as an instrument to control pollution.
5. Describe the concept of integrated pollution control. Discuss the
factors which would need to be taken into account in evaluating the
pollution impact of a lead battery works whose activities involve
discharge to air and water, and contamination of soils within the
works boundary. If the primary aim is to protect the health of the
public, what considerations would be necessary to minimize the
effect of the works in the optimal manner?

Index
Index terms

Links

A
Absorption

297

Accidental emissions

270

Accuracy

321

Acid aerosols

365

deposition

366

hydrolysis

mine drainage

108

mist

373

neutralizing capacity
precipitation
rain

339
322
373

93
106
66

Acid-conjugate base pair

Acidification

of lakes

283

of soils

370

Activity

369

176
296

181

coefficient

Adiabatic lapse rate

Adsorption reactions

208

Adsorption

139
196

Aerodynamic resistance

246

Aerosol

Agent Orange

216

Aggregation

189

Agricultural chemicals

213

182

189

249

This page has been reformatted by Knovel to provide easier navigation.

437

438

Index terms
Air Accumulation Factor

Links
262

pollutants

336

358

Quality Management Areas

423

424

quality monitoring

quality standards

quality strategy

400

sampling methods

291

Air-land exchange

244

Air-sea exchange

247

Albedo

373

253

Algal blooms

125

Alkalinity

Alkylation

176

Alkyllead additives

119

Alkylphenolethoxylates

132

Alkylphenols

338

Alligator

381

Aluminium

107
188

Alveoli

340

Alzheimer’s Disease

108

Ambient air monitoring

274

environmental monitoring

161

162

164

152
367

181
373

182

167

247

259

185

274

Ammonia

Ammonium sulfate

Amphibians

368

Anaerobic ponds

111

Analytical activity

Anion adsorption

210

Anoxic sediments

177

185

186

190

Anoxic waters

157

190

191

259

This page has been reformatted by Knovel to provide easier navigation.

191

439

Index terms
Antarctica
Antarctic ozone ‘hole’

Links
141
176

AOT40

364

Aquatic mixing

309

Aqueous phase chemistry

193

178

179

Antimony

systems

149

254
62

Aquifer

110

Aragonite

159

164

Argon

153

154

Arsenic

112

122

171

176

Arsenopyrite

112

Asthma

361

Atlantic Ocean

149

150

157

171

Atmospheric cycle

238

Atmospheric dispersal

305

fallout

212

photochemistry

stability

Atrazine

217

B
Barite

181

Barium

152

BATNEEC

403

Bauxite

108

Benthic macroinvertebrates

368

Benzene
Benzo[a]pyrene

171

188

134

251

338

Beryllium

185

Best available technology

404

This page has been reformatted by Knovel to provide easier navigation.

180

440

Index terms

Links

Best Practicable Environmental Option (BPEO)

403

Best Technical Means

409

Bicarbonate

151

Bioaccumulation

118

Bioalkylation

142

Bioavailability

Biogenic volatile organic compounds

Biogeochemical cycles

237

Biological oxygen demand

130

reduction processes

258

score

398

Biomagnification

118

Biomarkers

349

Bioremediation

192

Bipyridyl herbicides

209

Birds

368

172
257

229

Bisulfite

Black Sea

157

Blood lead

346

392

Boron

151

152

Boundary layer
layer resistance

239

246

BPEO

419

Bromide

153

Bromine

151

Bryophytes

355

Buffer strips

125

152

This page has been reformatted by Knovel to provide easier navigation.

245

441

Index terms

Links

C
Cadmium

169
188
344

170
233

171
337

Caesium

119

169

282

caesium-137

173
341

185
342

119

Calcite

151
180

159
184

164

178

179

Calcium

151

152

153

164

184

188

367

Carbamate

216

350

Carbon dioxide

13
158
164

140
159
165

150
161
197

153
162
253

155
163

155

Carbon dioxide cycle

Carbon disulfide

Carbon monoxide

Carbon tetrachloride

155

Carbonate compensation depth

164

system

Carbonic acid

Carbonyl sulfide

Carcinogens

Cariaco Trench

157

Cation exchange capacity

183

Cereals

337

Ceriodaphnia dubia

112

CFCs

2
246

Chemical oxygen demand

130

weathering
Chernobyl reactor accident

155

204

79
290

This page has been reformatted by Knovel to provide easier navigation.

442

Index terms

Links

Chlor-alkali plant

125

Chloride

153

172

Chlorine

151

152

Chlorinity

147

Chlorobenzenes

338

Chlorofluorocarbons

Chlorophenols

335

Chromium

152

Clay minerals

181

Climate change

Coastal waters

385

Cobalt

152

Colloids

108

Complexation
Compliance
Concentration
factor

153

204

173

185

148

188

142

150

151

155

169

188

198

173

177

185

82
268
74
262

Condensation

298

Conductivity

147

Congruent dissolution
Conservative behaviour
Constant ionic strength equilibrium constant
Contaminated land
soils

78
200

219

199

Contamination
Copper

200
152

169

186
CORINAIR

Coriolis force

148

Correlation spectrometer

324

This page has been reformatted by Knovel to provide easier navigation.

443

Index terms
Critical loads
loads assessment

Links
69
68

Crops

199

Crude oil

383

Cumulative stability constants

Cyanide

118

Cyanobacteria

125

231

D
2,4-D

216

Davies equation
DDT

77
217

Debye-Hückel theory

345

249

Decomposers

367

Density fractionation

295

Deposition of particles

342

velocity

244

Diagenesis

117

185

Dialysis

108

Diatom

106

communities

369

Dibenzofurans

Dieldrin

376

Diethylstilbestrol

381

Differential absorption laser

325

Diffuse layer

Diffusion tubes

297

Dig and dump

226

Dimethylarsinic acid

113

Dimethyl disulfide

372

This page has been reformatted by Knovel to provide easier navigation.

375

381

444

Index terms
Dimethyl sulfide

Links
36

151

153

165

166

216

235

377

381

197
Dimethyl sulfoxide
Dioxin

167
53

Dipole moment
Diprotic acids

142
88

Diquat

209

Directional deposit gauge

295

Disease

114

Dissociation fractions

Dissolved solids

255

Dobris Assessment

418

Dogwhelk

380

Dolphins

378

Dose

332

Dramsfjord

157

Drinking Water Directive

408

water

263

Droplets

Dry deposition

Dust samples

246

304

E
EC water quality standards for inland waters

323

Economic instruments

427

Eddy diffusivity

247

Effect

332

Eggshell thinning

375

Electron acceptors

activity

transfer

101

This page has been reformatted by Knovel to provide easier navigation.

445

Index terms
Emission

Links
48

inventories

limits

402

Endocrine disrupters

132

374

380

Enforcement notice

422

Enrichment factor

150

Environmental Assessment Directive

419

322

402

compartment

237

Management System

433

partitioning

264

Quality Standards

131

quality

268

reservoirs

239

Standards

401

Water Monitoring

279

Episodic events
Equilibrium constants
Essential trace elements

398
77
4

Ethylbenzene

134

Ethylene

362

Eulerian grid modeling
European Environment Agency
Monitoring and Evaluation Programme

34
418
35

Eutrophication

Expert Panel on Air Quality Standards

Exposure

332

Exposure-response relationships

333

Extended Debye-Hückel equation

336

F
Faecal pellets
Fast-response continuous monitors

141
313

181

This page has been reformatted by Knovel to provide easier navigation.

446

Index terms

Links

Ferric hydroxide

113

Ferrihydrite

109

Filtration

293

Fish

368

Fitness for use approach

412

Flocculation

189

Fluorides

362

Fluorine

151

Fluorocarbons

152

Flux

244

Foodstuffs

263

Forest decline

373

Free troposphere

239

Freons

382

Frequency of sampling
Freshwater
ecosystems

311
71
367

Fugitive emissions

Furans

270

G
Garden soils

234

Gasoline lead

264

Gastrointestinal tract

341

Gastropods

379

Gasworks sites

230

Gaussian model

plume dispersion
plume model
Geochemical anomalies
computer models

305
34

308

201
106

This page has been reformatted by Knovel to provide easier navigation.

447

Index terms
Geographical Information Systems
Geostrophic wind speed
Germanium

Links
326
11
176

Gibbsite

Glemeds

377

Global climate

Global energy balance

Global Environment Monitoring System

Global pollution
Global warming potential

240
14

Goethite

109

Gold

177

Grassland

337

Green taxes

428

Greenhouse effect
gases

14
12

Ground-level concentrations

274

Groundwater

112

Güntleberg approximation

H
Halite

144

Halmyrolysis

179

Halocarbon
Halocline
Halons

2
146
25

Hard water

115

Hardness

Hazards

219

Haze

Health effects

357

388

This page has been reformatted by Knovel to provide easier navigation.

448

Index terms
Heavy metal
Helium

Links
52

214

233

154

159

251

234

154

Henry’s Law

Homeostatic control

349

Hospital admissions

360

Human exposure

263

Humic substances

205

143

159

383

Hydration
Hydrocarbons

361

Hydrogen

155

Hydrogen bonding

142

182

Hydrogen sulfide

259

Hydrolysis

143

Hydroperoxy radical

Hydrothermal

140

Hydroxyapatite

128

Hydroxyl radical

Hyper-keratosis

112

164

151

171

167

242

I
Illite

204

Impingement

293

Imposex

194

Incongruent dissolution
Indian Ocean

80
148

Inner sphere complexes

Inorganic complexes

Instrumental analytical methods

321

Intake design

291

Integrated Pollution Control (IPC)

403

Pollution Prevention and Control (IPPC)

379
149

404

This page has been reformatted by Knovel to provide easier navigation.

172

184

449

Index terms

Links

Intellectual deficits

393

Internal loading

126

Internet

435

Ion pairs

Ionic strength

IPPC

105

419

Directive

404

Iridium

177

Irish Sea

247

Iron

121
178
186

Irrigation water

114

Isokinetic sampling

292

152
181
188

Isoprene

Itai-Itai

113

233

204

172
182
189

K
Kaolinite
Kesterson Reservoir

115

Krypton

154

Kyoto

L
Lagrangian trajectory modeling
Lakes

34
257

sediments

369

119

Land contaminants

214

Land Use Planning

418

Landfilling

232

235

This page has been reformatted by Knovel to provide easier navigation.

176
184
190

177
185

450

Index terms
Landfill sites
tax

Links
273
429

Lanthanum

152

Lapse rate

Lead

52
171
197
260
391

lead-210

113
173
234
341

152
185
250
346

169
186
257
354

164

181

184

119

Leaded paintwork

262

Leaded petrol

261

Leblane Process

123

Lichens

355

Life Cycle Analysis

430

Lifetimes

240

264

Ligands

Liming

129

Limits

400

Line sources

269

Lithium

140

152

358

London smog

140
176
241
342

Long-path absorption measurements

324

Los Angeles

359

Lysocline

164

M
Magnesium

151
188

deficiency

374

152

This page has been reformatted by Knovel to provide easier navigation.

451

Index terms
Manganese

Marine aerosol

Links
121

140

152

153

171

172
181
188

173
182
189

176
184

177
185

178
186

155
340

174
388

186

242

247

birds

384

environment

383

mammals

378

Mathematical modeling

Mean ion activity coefficient

Measurement
of gaseous air pollutants

267
314

Mediterranean Sea

145

157

158

Mercury

117
175

142
176

152
188

Mesophyll tissues

344

Metal cations

210

344

354

155

complexation

ion mobility

pollution

388

toxicity

Metalliferous mining

233

Metals

339

in dusts

288

Meteorological data

318

Methaemoglobinaemia

130

Methane

Methanesulfonate

167

Methylation

113

Methyl bromide
Methyl tert-butyl ether
Methylchloroform

23
133
25

243

This page has been reformatted by Knovel to provide easier navigation.

452

Index terms

Links

Methyl iodide

155

Methylmercury

118
390

Methyl radicals

Minamata

118

MINTEQA2

106

Mirabilite

144

Mixed acidity constant

341

346

388

390

Modeling of environmental dispersion

303

Molybdenum

185

Monitoring

267

protocol

317

Monomethylarsonic acid

113

Monsoons

148

Montreal Protocol

339

Multi-functionability approach

412

Mussel Watch

352

N
National Air Quality Strategy

National Radiological Protection Board

278

National Survey of Air Pollution

276

Neon

154

Neutralization

129

Nickel

152

169

171

173

185

Nitrate

59
168

60
186

129
259

157
408

167

186

259

radical

167

Nitrate Vulnerable Zones

422

Nitrite

157

This page has been reformatted by Knovel to provide easier navigation.

453

Index terms
Nitrogen
deposition

Links
154

259

373

dioxide

242

340

344

oxides

155

259

246

145

164

169

188

190

249

250

264

334

338

340

156

365
N-Nitroso compounds
Nitrous oxide
Non-conservative

130

198
Non-conservative behaviour

142

Non-sea salt sulfate

151

Nonylphenol

132

North Sea

196

Number of sampling sites

316

Nutrients

O
Objectives
Occult deposition

400
64

Ocean basins

257

Octanol-water partition coefficient

211

Oil

383

spills

385

Organic contaminants
Organic material

210

215

139

147

149

158

164

176

186

345

378

Organochlorine pesticides

352

Organochlorines

212

216

Organophosphate

216

350

Organophosphorus

131

Orthophosphate

126

This page has been reformatted by Knovel to provide easier navigation.

381

454

Index terms
Outer sphere complexes

Links
82

Over-abstraction

398

Oxides of nitrogen

260

Oxygen

140

142

153

154

155

156

157

169

171

186

13
344

19
354

56
360

245
363

325
373

depletion

142

layer

150

157

168

212
384

218
386

231

339

Ozone

P
Pacific Ocean

149

Pacific oyster

379

PAHs

53
352

Palladium

177

Paraquat

209

Parathion

217

Particle formation

Particles

249

Paniculate matter

360

Partition coefficients
Parts per billion
Pasquill stability class
PCBs

264
5
307
53

132

217

219

235

345
379

347
381

375

376

377

Pedogenic processes

202

Peritachlorophenol

410

Permeability

206

Peroxyacetyl nitrate

This page has been reformatted by Knovel to provide easier navigation.

455

Index terms

Links

Peroxybenzoyl nitrate

Pesticides

131

211

352

85
153
161
181
189

139
157
162
182

140
158
163
183

142
159
164
186

Phenol

338

Phosphate

125

167

168

170

Phosphorus

122

152

169

Photic zone

142

156

158

164

344

Photochemical ozone

Photochemical ozone creation potential (POCP)

Photochemical smog
Photosynthesis
PHREEQC
Pinene

152
160
177
188

168

359
98
106
38

Pit lakes
Pitzer equation

110
77

Placer gold

117

Planned emissions

270

Plants

362

Platinum

177

Plume rise

Plutonium

248

282

PM10
Podzolization

107

Point of zero charge

181

Point sources

269

Polar stratospheric clouds

340

182

This page has been reformatted by Knovel to provide easier navigation.

360

361

456

Index terms
Pollution
control permit
Polonium

Links
190

191

197

200

152

172

184

188

156

157

177

427
185

Polychlorinated biphenyls see PCBs
Polychlorinated dibenzodioxins

Polycyclic aromatic hydrocarbons, see PAHs
Pore waters

121

Potassium

151

Potential density

146

Potential temperature

142

Precaution

413

Precision

321

Prerequisites for monitoring

316

Presentation of data

326

Primary minerals

Primary pollutants

Primary producers

367

Productivity

129

Protection zone

422

Proton transfer

101

Purification

134

Pycnocline

146

Pyrite

104

pε

145
322

135

140

186
pε–pH diagrams

100

Q
QSAR

335

Quality assessment

398

Quality control procedures

317

This page has been reformatted by Knovel to provide easier navigation.

457

Index terms
Quartz

Links
80

R
Radon

277

Rain

255

Rainout

Rainwater composition

Reclamation

226

Recycling

430

Red Sea

145

Redox conditions

207

Redox equilibria

100

Redox intensity

Redox processes

Redox-sensitive elements

Reedbed

135

Remediating contaminated land

411

Remote sensing of pollutant

324

Reproduction

374

Residence time

142

Resistances to transfer

252

Respirable particles

341

Respiration

253

151

181

Respiratory disease

358

Respiratory tract

340

Restoration

128

Revelle factor

164

Risk Assessment

347

Risk-based Corrective Action

413

River water

255

Riverine suspended particulate matter

254

360

This page has been reformatted by Knovel to provide easier navigation.

197

458

Index terms

Links

Rubidium

140

Ruthenium

177

S
Saanich Inlet

157

Salinity

142

143

145

146

147

148
187

150
188

151
197

153

162

Salinization

115

Scandium

152

Scavenging coefficient
Scavenging ratio

64
252

Sea breezes

Sea-salt

Seal populations

185

378

Secondary minerals

Secondary pollutants

135

Sediment

110

258

cores

123

soil and biological monitoring

285

Sedimentation

294

velocities

Selenium

115

142

Sellafield

247

283

Semen quality

381

Sewage effluent

382

Sewage sludge

114

Sigma-tee

146

Silicate

167

Silicic acid
Silicon

176

218
168

170

169

181

81
152

This page has been reformatted by Knovel to provide easier navigation.

459

Index terms

Links

Silver

177

Site investigation

220

Smectites

204

Smelting

110

233

Smoke
particles

186

358

Sodium

151
172

152
173

184

Soil

201

262

370

and sediment sampling methods

302

cleaning

226

constituents

203

formation

202

organic matter

205

profile

202

properties

206

Solubility

Solubility product

Solvay process

123

Solvents

235

Soot

188

153

175

176

115

Source inventory

319

Source monitoring

271

Southern Ocean

148

Speciation

171
339

Specific conductance

Specific Ion Interaction Theory

Stability field

104

Stability limits of water

100

Standards

400

172

174

402

This page has been reformatted by Knovel to provide easier navigation.

460

Index terms
Stomata
Stratosphere
Stratospheric ozone
depletion
Strontium
Subsidence inversion
Subsidiarity

Links
337

341

239

325

142
151

152

30
416

Sulfate

Sulfite

Sulfur

151

176

Sulfur dioxide

41
344

50
355

Sulfurous acid

Superfund

411

Supersaturation

153

Supply chain pressures

433

Surface charge

181

Surface complex

Surface resistance

246

Surface tension

143

Surfactants

143

Suspended particulates

299

Synthetic pyrethroids

131

106
358

T
2,4,5-T

216

Tailing ponds

110

TBT

195

196

197

TCDD

216

217

377

TCDF

216

217

Temperature inversions

217

This page has been reformatted by Knovel to provide easier navigation.

151
362

340

461

Index terms
Tertiary treatment
2,3,7,8-Tetrachlorodibenzodioxin

Links
135
4

Tetraethyl lead

260

Tetramethyl lead

260

Thallium

176

Thermocline

145

Thorium

152

185

180

Titanium

152

153

Toluene

134

Tin

Total suspended particulate matter

Toxic Organic Micropollutants (TOMPS)

Toxicity
tests

185

299
333

334

Tradeable permits
Trajectories

430
28

Transfer velocity

252

Tributyltin

194

Trichloroethene

133

Trigger concentrations

224

Tropopause

198

379

225

Tube wells

112

Turbidity

117

U
Underground fuel tanks

413

Uniform emission standards

402

Units of concentration

403

405

Upwelling

149

Uranium

152

Uranium dioxide

153

168

This page has been reformatted by Knovel to provide easier navigation.

462

Index terms

Links

US ambient air quality standards

323

US National Air Sampling Network

298

UV radiation

142

V
Vanadium

152

Vegetation

337

Vehicular emissions
Vermiculite

47
205

Vienna Convention

Visibility

Vitellogenin

133

Vivianite

126

Volatile organic compounds (VOCs)
Volatilization

382
57

141

W
Washout

coefficient

253

Waste disposal

232

Waste Framework Directive

429

Waste management hierarchy

430

Waste treatment

135

Waste waters

135

WATEQ4F

106

253

Water
meadows

125

Quality Objectives

401

resources

431

sampling methods

300

table

112
This page has been reformatted by Knovel to provide easier navigation.

463

Index terms

Links

Water Framework Directive

417

Weathering

179

of rock

255

Weddell Sea

149

Wet deposition

193

195

198

152
186

169
188

170
189

171
250

Wetlands

128

Wheal Jane tin mine

109

Wind shear

192

X
Xylene

134

Z
Zinc

122
185

Zooplankton

368

This page has been reformatted by Knovel to provide easier navigation.

Source Exif Data:
File Type : PDF File Type Extension : pdf MIME Type : application/pdf PDF Version : 1.6 Linearized : No XMP Toolkit : 3.1-701 Modify Date : 2006:05:10 01:56:41-04:00 Create Date : 2006:05:10 01:56:07-04:00 Metadata Date : 2006:05:10 01:56:41-04:00 Creator Tool : ABBYY FineReader Format : application/pdf Document ID : uuid:0d974feb-cf19-4043-a58d-89f2c589c077 Instance ID : uuid:e046bd0d-7156-4777-85a9-657716366171 Has XFA : No Page Count : 485 Creator : ABBYY FineReader Producer : Acrobat Distiller 5.0 (Windows)
EXIF Metadata provided by EXIF.tools

An Introduction To Environmental Chemistry And Pollution

Navigation menu

Versions of this User Manual:

Views

Navigation