Proposal No Eit Dem User Guide

User Manual:

Open the PDF directly: View PDF .
Page Count: 18

Download
Open PDF In Browser	View PDF

EIT DEM TOOL User's Guide
John Cook
Jeffrey Newmark
Naval Research Laboratory
Introduction
The Solar and Heliospheric Observatory (SOHO) spacecraft carries onboard the Extremeultraviolet Imaging Telescope (EIT), which images the Sun in a field-of-view extending out to
1.4 Ro in four wavelength channels covering a temperature range from 80,000 K to 2.5x10 6 K
(Delaboudiniere et al. 1995). The 1024x1024 CCD detector has a spatial resolution of two pixels,
where 1 pixel subtends 2.6 arcsec. The SOHO spacecraft was launched in December 1995, and
scientific observations began in January 1996.
Cook, Newmark, and Moses (1999) have developed a technique to calculate the differential
emission measure (DEM) using images from the four wavelength channels of the EIT instrument
at 171 Å, 195 Å, 284 Å, and 304 Å. An individual DEM curve, covering the temperature range
from logT = 4.6 to 6.5 in 0.1 dex steps, is constructed for each pixel in the EIT field-of-view.
We call the result a DEM map. The DEM map is determined so as to best match the original four
EIT images when synthetic EIT images are back-calculated from it. This technique uses the
CHIANTI program (Dere et al. 1997) as the atomic physics engine for its calculations.
We have placed the computer programs that produce the DEM maps, all in the IDL
programming language, in the Solar Soft collection of software products. The programs should
be run in a home IDL environment with paths set to Solar Soft, including the CHIANTI package.
In this User's Guide we briefly describe the construction technique, illustrate the use of the DEM
tool with examples for running it, and finally present a more detailed discussion of a sample
application of the DEM tool to model solar EUV irradiances, where we also discuss some further
aspects of our basic DEM tool technique in greater detail.
For further help and coaching with using our DEM tool for EIT, contact Jeff Newmark at
jeffrey.newmark@nasa.gov.

Basic idea of the DEM construction
The differential emission measure (DEM) provides a compact summary of the solar atmospheric
density and temperature structure along an individual line-of-sight, determined from emission
line intensities. The intensity of an optically thin emission line (in a simple 1-D form for this
discussion) is found from

We have developed a technique to calculate the differential emission measure using images from
the four wavelength channels of the EIT instrument at 171 Å, 195 Å, 284 Å, and 304 Å. We use
the elemental abundances of Feldman et al. (1992), and the ionization balances of Mazzota et al.
(1998) in calculating intensities.
The EIT images when processed by eit_prep are in instrument intensity units DN s -1. From a
given DEM curve the unique synthetic instrument units intensity for each channel can be found
by calculating the absolute intensity spectrum over the EIT wavelengths from the DEM curve
and then applying the instrument efficiency function. You can't go the other way, from an
individual channel instrument units intensity to a unique DEM curve, because different DEM
curves can produce the same synthetic instrument units intensity. But looking for a single DEM
curve that simultaneously reproduces all four channel instrument units intensities is more
constrained. Once you obtain that DEM curve you can calculate model absolute intensities.
We begin with a standard initial starting DEM curve (the CHIANTI active region DEM,
although our method does not depend on the exact starter choice), which produces starting
synthetic absolute intensity spectra over each EIT channel. From these synthetic spectra,
synthetic EIT channel intensities are found. We then use the observed EIT image intensities (DN
s-1) in all four channels to modify the initial DEM to produce individual DEM curves for each
pixel that will best reproduce the observed DN s -1 intensities, giving a DEM map of the entire
field-of-view. Our DEM map is an individual DEM curve for each pixel of an image, covering

the temperature range logT = 4.6 to 6.5 in 0.1 dex steps.
The intensity is linear with DEM for emission lines which are emitting in their allowed range of
electron density. The EIT channel passbands contain a number of forbidden lines, and we first
checked to see if their intensity contributions make the total passband intensities significantly
pressure-sensitive over a range of representative solar electron pressures, which they were not.
Using the initial starting DEM curve and its calculated four synthetic EIT channel intensities, we
determine for each individual pixel the scaling constants for each of the four channels which
would multiply the starting DEM to best reproduce the four observed EIT intensities. Naturally,
the four scaling constants are almost never identical: simply linearly increasing or decreasing the
initial active region DEM curve will not reproduce the four intensities simultaneously. The actual
shape of the initial DEM curve must be modified. From the response curves with temperature for
each channel, we identified the temperatures of peak response (temperature-of-formation). We
generate a scaling function vs. logT from the four individual channel scaling constants placed at
their peak response temperatures, interpolated over the logT grid. We then multiply the initial
DEM curve by the scaling values over the temperature grid, resulting in an individualized DEM
curve which now better reproduces the four observed EIT intensities. Finally, we iterate once
using this new, now individualized DEM map as the starting DEM for each pixel to obtain a final
DEM map. We found no further improvement iterating more than once. The actual technique is
slightly more complex because we must handle explicitly the presence of significant multiple
temperature response peaks in the 304 Å (three peaks which must be considered) and 284 Å (two
peaks) channels, and we do some careful bookkeeping of the multiple temperature contributions
to these last two channels in calculating the synthetic intensities and scaling constants.
The He II 304 Å line is difficult to model, but is critical to include in the DEM map construction
as part of the total EIT 304 Å channel intensity, which controls the cooler end of the DEM curve.
This line is inherently optically thick and so not appropriate for a direct DEM approach to
calculating its intensity. It is also anomalous from other, optically thin, transition region lines in
that it clearly shows coronal holes. The DEM map construction uses an empirical correction
factor for the direct He II 304 Å intensity calculated by the CHIANTI atomic physics engine for
an optically thin plasma. Cook, Newmark, and Moses (1999) obtained this scaling factor from
examining a collection of observed He II 304 Å intensity values for quiet and active regions.
The EIT 304 Å channel contains contributions from cooler transition region plasmas emitting the
He II 304 Å line, as well as from hotter coronal plasmas emitting in the same wavelength range.
The hotter contribution must be consistent with the DEM curve well-determined by the three
other coronal EIT channels, and the required consistency in the final DEM to reproduce all four
EIT channels does help to check the empirical He II scale factor, which we believe to be
reasonable. But again, this is inherently a difficult line to model theoretically.
From these DEM maps we can calculate a synthetic solar image for any emission line,
wavelength interval, or instrumental passband with measured (or assumed) passband efficiencies.
In addition we can calculate a synthetic full disk irradiance from the synthetic solar image.
Because the EIT images and the resulting DEM maps extend off-disk, these calculated
irradiances include the irradiance contribution from off-disk, which can be significant in the
EUV wavelength range.

Running the DEM tool in IDL
We assume that the user is familiar with IDL. In this section we list the initial comments sections
of the two IDL procedures employed, and give examples of program parameter choices to
perform various tasks using the DEM tool. The basic driver procedure is EIT_DEM_TOOL. It
defines internally two IDL functions and a separate IDL procedure called EIT_KCORR, which
computes the scaling factors to modify the shape of the initial DEM curve. EIT_DEM_TOOL
calls a separate IDL function EIT_LINE_MAP, whose comments section is listed after
EIT_DEM_TOOL.
;+
; NAME:
;
EIT_DEM_TOOL
;
; A user guide is included in the Solar Software distribution. For further information
; go to your local directory $SSW/soho/eit/idl/response/dem_tool/user_guide.pdf
;
:
; PURPOSE:
;
Create a Differential Emission Measure (DEM) Map, i.e. computed for
;
each pixel individually, based upon the 4 EIT channels. Optionally,
;
also returns a line/bandpass map or the irradiance for chosen line.
;
DEM maps can be from pre-computed daily map database
;
; CATEGORY:
;
Analysis
;
; CALLING SEQUENCE:
;
eit_dem_tool,files,dem_map,temp,[mk_line=mk_line],[line_map=line_map],$
;
[wmin=wmin],[wmax=wmax],[irradiance=irradiance],$
;
[no_dem_map=no_dem_map], [date = date],[leak_284=leak_284],$
;
[coefs_only=coefs_only],[nrl=nrl]
;
; INPUTS:
;
files
- Names of 4 EIT raw files in order 171,195,284,304
;
or a processed (1024,1024,4) array
;
; OPTIONAL INPUT KEYWORD PARAMETERS:
;
mk_line
- set to produce output line map
;
ion
-set to ion name for line maps
;
wmin
- minimum wavelength to consider for line maps
;
wmax
- maximum wavelength to consider for line maps
;
irradiance
- set if only want line irradiance
;
no_dem_map
- set if do not wish calculation of output DEM
;
map i.e. only wish line_map output
;
date
- read in pre-computed DEM for set date
;
leak_284
- set to use in eit_prep for the 284 A channel light leak
;
coefs_only
- set if return correction coefficients only

;
nrl
- REQUIRED to set for FITS files from NRL archive (to run the
;
NRL version of eit_prep)
;
; OUTPUTS:
;
dem_map
- differential emission measure map at
;
26 predefined temperatures logT = 4.0 – 6.5, step 0.1
;
temp
- temperatures (log)
;
; OPTIONAL OUTPUT KEYWORD PARAMETERS:
;
LINE_MAP
- output line map or irradiance
;
; COMMON BLOCKS: none.
;
; SIDE EFFECTS:
;
; RESTRICTIONS:
;
; PROCEDURE:
; Modifies starter DEM to fit observed EIT data.
; Starter DEM = CHIANTI V3 active_region for logT = 4.6 - 6.5 and
;
CHIANTI V5 quiet_sun for logT = 4.0 - 4.6
;
; SUBROUTINES:
; E_INTERP, E_FIT2, EIT_KCORR : included in this file, automatically compiled
; EIT_LINE_MAP : Separate procedure
; MK_EIT_SPEC: Included in EIT_LINE_MAP
;
; MODIFICATION HISTORY:
;
Written by: J. Newmark
October 2001
;
Modified: J. Newmark
May 2006
;
N. Rich
October 2010 Add /NRL keyword
;
;-

The IDL function EIT_LINE_MAP is called from EIT_DEM_TOOL, and itself includes a
subroutine procedure MK_EIT_SPEC which uses the CHIANTI package to perform the atomic
physics calculations.
;+
; NAME:
;
EIT_LINE_MAP
;
; PURPOSE:
;
Create a line/bandpass map based upon input differential emission
;
measure (DEM) for EIT like instrument.
;
; CATEGORY:

;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
;
:
;
;
;
;
;
;
;
;
;
;

Analysis
CALLING SEQUENCE:
line_map =eit_line_map(wave,dem,intemp,x,coefs,ion=ion,wmin=wmin,$
wmax=wmax,he_fac=he_fac,trange=trange,instr=instr,$
irradiance=irradiance,delta_wave=delta_wave, $
xyidx=xyidx,nocon=nocon,abund=abund, $
eqion=eqion,pressure=pressure)
INPUTS:
wave
dem
intemp
x

- EIT bandpass or particular line wavelength
- input DEM (1D or 3D)
- temperature array (1D) corresponding to DEM
- EIT specific line parameter

OPTIONAL INPUT KEYWORD PARAMETERS:
coefs
- coefficients to "tweak" input DEM
ion
- choice of output line if using single line
wmin
- minimum wavelength to consider for line maps
wmax
- maximum wavelength to consider for line maps
he_fac
- enhance factor for Helium II 304 over default
spectrum
trange
- array of temperature indices to consider if not
full range
instr
- set to particular instrument, default = 'eit'
set to 'dum' or anything else for simple lines
irradiance
- set if only want line irradiance
delta_wave
- wavelength bin size for subsetting wavelength
range. Default is 1 bin for wmin->wmax. Only
useable with irradiance keyword
xyidx
- Index values of array that should be used. All
others are set to zero
nocon
- Set if do not want continuum included in calculation
pressure
- Set if want spectra calculated with specific Pressure,
default = 1.0E+15
abund
- Set if want spectra calculated with specific Abundance,
default = Coronal (Feldman) et al.
eqion
- Set if want spectra calcualted with specific ionization
equilibrium, default = Mazzotta et al. ext
OUTPUTS:
line_map

- computed line/bandpass map

OPTIONAL OUTPUT KEYWORD PARAMETERS:
COMMON BLOCKS: none.
SIDE EFFECTS:
RESTRICTIONS:

; Requires Pre-computed spectra using CHIANTI Isothermal procedure
;
; PROCEDURE:
;
; SUBROUTINES: Include MK_EIT_SPEC for computing Isothermal spectra
;
; Helium enhancement = 9.5 Any change in the elemental abundances
;
employed from our Feldman default would
;
require re-examination of the Helium
;
enhancement factor
;
;
; MODIFICATION HISTORY:
;
Written by: J. Newmark
October 2001
;
Modified:
J. Newmark
May 2006
;
Modified:
J. Newmark
July 2010
;-

We strongly suggest that these programs be run using the standard default choices of atomic
parameters. Unless the user is prepared for significant work in understanding the code and time
in testing and verifying any results with alternative parameter choices, they are best avoided. In
particular, the He II 304 Å emission line is a special problem, and the parameter he_fac is chosen
for the particular elemental abundance choice we employ. We have retained several features
which we introduced for special calculations, and that have been left in the code but will not be
described in this guide. Be warned. An example is the nocon keyword. We include the CHIANTI
continuum calculation as the default setting, and advise you to also. We in past checked the
continuum contribution to the total by comparing the nocon = 0 and nocon = 1 calculations, and
leave this keyword in. For special calculations we often changed code instead of adding general
keywords, and to do some special calculations you too will have to slightly edit the code.
The most direct use of our DEM tool is to produce emission measure maps of the differential
emission measure Ne2 dl/dT (this is the type of the CHIANTI reference DEMs). In addition, the
output array dem_map gives the log DEM values. However, for most purposes we find that the
logarithmic DEM, Ne2 dl/d logT in units cm-5, is more useful, and this is what we will always
mean by DEM. This DEM is just log e10 T Ne2 dl/dT, and its log is dem_map(i,j,*)+temp(*)
+0.3622, where the output vector temp is logT. The default temperature range of temp is
nominally logT = 4.0 – 6.5 dex, in 0.1 dex steps. The temp output vector will have 26 elements.
The first six of these, temp(0:5), are for a highly schematic cool scaled quiet Sun downward
extension of the DEM smoothly joining at logT=4.6, which we added to get some representation
of cool continua when we were investigating modeling of possible wavelength coverage to
higher wavelengths than 550 Å. The dem_map output array is nominally of dimension
(1024,1024,26), the EIT 1024x1024 field-of-view at 26 temperatures. The individualized DEM
curve calculation is only performed over logT = 4.6 – 6.5, the temperature range actually
sampled by the four EIT channels. We suggest that the user simply ignore the DEM output
below logT = 4.6, dem_map(1024,1024,0:5), unless willing to perform significant checking and
verification.

You must keep in mind the difference between temperature and wavelength dependencies. The
temperature range sampled by EIT sets the temperature range coverage of the DEM maps. But
any optically thin emission from whatever wavelength that is actually formed in this temperature
range can be modeled, such as the Fe XIV 5303 Å coronal green line, if CHIANTI can compute
the atomic level populations. Solar flare lines cannot be treated, nor can hotter (temperature-offormation > logT = 6.5) active region emissions; nor can cooler (< logT = 4.6) emissions be
accurately treated at any wavelength. We have checked carefully the DEM calculations over the
default wavelength range 165-550 Å, and verified that broadband intensities produced from the
DEM maps are reasonable (“broadband” because there are some weak cooler and hotter lines in
the range that won't be accurately modeled, but they don't significantly contribute over, say, 10 Å
intervals anywhere in this wavelength range). Read the final section for further discussion.
The raw (or processed; see keywords list) EIT data must be input in eit_dem_tool. The easiest
procedure is to put the four raw files in the working IDL directory, but at least include the full
path name to the data if you store it in a separate directory.
To produce the basic dem map, simply run
eit_dem_tool,['filename1','filename2','filename3','filename4'],dem_map, temp
The input files must be in the order: 171 Å, 195 Å, 284 Å, 304 Å channel filenames.
Each time eit_dem_tool is run, the function eit_line_map looks for an IDL .save file usually
named radiance_unit_dem_con.save in the working directory, which contains information
calculated using CHIANTI necessary to the DEM map construction. If this file is not present in
the directory, such as the first time run, it is calculated and the .save file is stored. The next time
eit_dem_tool is run it does not have to repeat these calculations but simply restores the .save file.
If you later want to do a new calculation with new parameter and keyword choices, make sure to
delete the old .save file so that a new one based on your current choices is calculated.
The IDL function eit_line_map is used in two modes. It is called from eit_dem_tool in forming
the DEM map output, when it returns synthetic values for DN s-1 in each EIT channel from the
DEM map being constructed. It also is used to make line_map, which is in absolute intensity
units, when mk_line is set in eit_dem_tool.
To make a line intensity or broadband map (or the scalar line irradiance if the irradiance keyword
is set), set the mk_line and line_map keywords. The output intensity array line_map(1024,1024)
is an array of intensities in units ergs cm-2 s-1 sr-1 (the irradiance scalar is ergs cm-2 s-1). The
keywords wmin and wmax are used to define the wavelength interval summed over. In
eit_line_map an additional .save file analogous to radiance_unit_dem_con.save should be
calculated over a narrower wavelength range around the line or wavelength band, but at a higher
wavelength resolution than the default 1 Å over the 165–550 Å range, typically over an angstrom
or two with 0.02 Å resolution. We have coded in examples for three emission lines, but you will
need to edit (described next) for other lines. The CHIANTI line list can be consulted to verify
what wavelength CHIANTI assigns and what other lines are nearby. For a new line or interval
not calculated previously, explore several choices of wmin and wmax and the wavelength
choices in the call to isothermal (again, described next) to satisfy yourself that the output

intensity map is not sensitive to the exact values used (contains the core wavelength assigned by
CHIANTI, does not include any nearby possible blends of same element and ionization stage).
We give an example for the Fe XIV 5303 Å coronal green line. You can run
eit_dem_tool,[4 filenames as before],dem_map,temp,/mk_line,line_map= fe14_map,
ion='fe14',wmin=5304,wmax=5305
Before this, you may need to edit eit_line_map in the subsection defining mk_eit_spec that runs
the CHIANTI isothermal procedure. Make a copy in your working directory to edit. Look at
if keyword_set(ion) then begin
; Here are three sample line calculations for He II 304 A, Si XI 303 A, Fe XIV 5303 A
; If you want another line, add new code appropriate for new line; keyword sngl in
; isothermal call of format element_ionization stage
isothermal,303.,304.,.02,temp,waves,spec0,pressure=pressure,/noverbose,cont=cont,$
sngl='he_2',abund=abund,ioneq=eqion
save,file='he2_radiance_unit_dem_'+app+'.save',temp,pressure,spec0,waves
isothermal,303.,304.,.02,temp,waves,spec0,pressure=pressure,/noverbose,cont=cont,$
sngl='si_11',abund=abund,ioneq=eqion
save,file='si11_radiance_unit_dem_'+app+'.save',temp,pressure,spec0,waves
isothermal,5304,5305,.02,temp,waves,spec0,pressure=pressure,/noverbose,cont=cont,$
sngl='fe_14',abund=abund,ioneq=eqion
save,file='fe14_radiance_unit_dem_'+app+'.save',temp,pressure,spec0,waves
Notice that we have coded examples for three lines. No edit is necessary for these, but other lines
will need to be coded. Then re-compile explicitly in IDL with “.run eit_line_map”. The first time
you run with ion set, the three .save files for fe14, si11, and he2 (and broadband, below) will
appear in your working directory. In choosing the isothermal wavelength range and resolution,
what matters is where CHIANTI has assigned the wavelength (Fe XIV 5304.57; Si XI 303.33,
He II 303.78). Do not set wmin or wmax outside the range used in the isothermal call. Notice in
the code that if you set ion='he2' the CHIANTI optically thin intensity will be scaled by he_fac.
We also illustrate a broadband example, for the intensity over 300-310 Å (all the emissions, with
the He II 304 Å emission line the strongest component). The call is
eit_dem_tool,[4 filenames as before],dem_map,temp,/mk_line,line_map=band_300_310,
ion='broadband',wmin=300,wmax=310
We set ion='broadband', which uses an isothermal calculation over 165-550 Å with the keyword
ergs set, and includes (as with the previous isothermal call over 165-550 Å in photon units, used
to calculate dem_map) all the emissions because sngl is not set in the coded isothermal call.
The next section contains an explanation of the file infile.list in the Solar Soft distribution.

Application of the DEM tool to solar EUV irradiances
The DEM map technique has been used in several previous papers, including modeling of the Fe
XIV 5303 Å coronagraph green line observations from Sacramento Peak Observatory (Cook,
Newmark, and Altrock, 2005), comparisons between EIT and SUMER observations from SOHO
(Brosius et al., 2002), and modeling of the full sky He II 304 Å flux for application to analysis of
observations from the GAS experiment on the Ulysses spacecraft (Auchere et al., 2005a, 2005b,
2005c).
This section is adapted from the texts of an earlier paper, and a proposal to NASA, on
comparison of the EUV irradiance observed by the SEM instrument on SOHO with a synthetic
model irradiance. We have included in our Solar Soft directory the file infile.list which lists the
GSFC archive filenames of the four daily EIT images which we have used in the work described
below. This work illustrates types of analysis which can be performed using our DEM tool:
Active regions rotating across the disk modulate the solar radiative output on a 27 day timescale,
while over the 11-year solar activity cycle timescale the full-disk brightness varies with activity
level as the fractional area of active regions increases. The brightness varies more strongly for
emissions from higher-temperature plasmas, because of the increasing contrast with temperature
of active regions to quiet regions on the surface. From space the Sun can be viewed at EUV
wavelengths in a variety of emission lines spanning a range of emitting temperatures in the solar
transition region and corona. Some of these, such as the He II 304 Å emission line, are
significant to the physics of the upper terrestrial atmosphere. The EUV range is important for the
terrestrial ionosphere, where it is the dominant ionizing flux at certain levels, and controls the
dynamics of the thermosphere (Meier, 1991). He II 304 Å is also an important contributor to the
solar flux at wavelengths controlling photoionization of neutral H and He atoms in the
heliosphere. The solar EUV flux and its variability impacts a range of phenomena of physical
interest.
We have modeled irradiances over 1996-2005, which covers the solar activity cycle from the
previous cycle minimum epoch, to well past the cycle 23 maximum. For this period an accurate
EIT absolute calibration is available.
We compute a DEM map for each day using four EIT images exposed closely together in time
(20 minutes). From the daily DEM map we then calculate daily synthetic 10 Å band intensity
images over 170-550 Å. Figure 1, from Newmark, Cook, and McMullin (2008), illustrates daily
synthetic images for the two days 10 February 1997, a period of low solar activity level, and 3
January 2002, a period of higher activity, in solar cycle 23. For each day the integrated 260-340
Å intensity image and also the pure He II 304 Å line intensity image is illustrated. These two
days are marked on the SEM irradiance history plot of Figure 2 below.
The intensity image is converted to the equivalent flux at 1 A.U., or the irradiance F, where F =
/(1 A.U./R)2, where  is the average intensity out to radius R. Coronal lines can have
significant off-limb emission, and we include intensities out to 1.3 R o , which captures the great
majority of off-limb emission, while not extending into low signal-to-noise areas of the EIT
images. The EIT fluxes are observed at the location of the SOHO spacecraft, the Lagrangian Sun

-Earth L1 point. Using the spacecraft data files, this flux is converted to a standard 1 A.U. flux.
Because the spacecraft orbital path can extend slightly above or below the ecliptic plane by about
0.01 A.U., this flux is not identically the irradiance (flux toward Earth at 1 A.U.), but we still
refer to it as the irradiance. More detail is given in Newmark, Cook, and McMullin (2008).

Fig. 1. Model synthetic images of the 260-340 Å and the pure He II 304 Å intensities for 10
February 1997 and 3 January 2002.

Fig. 2. Histories over 1996-2005 of the SEM 260-340 Å observed and model irradiances.
SEM observations: + joined by line; SEM model: triangle. The two days of differing
activity levels illustrated in Figure 1 are marked by triangular pointers from below.

Newmark, Cook, and McMullin (2008) compared model irradiances to the observed
SOHO/SEM 260-340 Å irradiances (see Judge et al. 1998) over 1996-2005, finding little
systematic long-term change in the fractional systematic error (Model/SEM) ~ 0.98, and 19%
scatter (3σ). Figure 2 illustrates from this paper the time histories over 1996-2005 of the 260-340
Å observed SEM and our model irradiances. The SEM monitor should have a particularly
accurate long term absolute calibration, from a pre-launch laboratory calibration and several
tracking comparisons with calibration rocket underflights over 1996-2006 (Judge, McMullin, and
Ogawa 1999; McMullin et al. 2002). The analysis of the error in the SEM irradiances by
McMullin et al. (2002) gives a 3σ fractional error of ±30%. Newmark, Cook, and McMullin
(2008) give the error in the model SEM irradiances as ±32%. The good agreement, within the
errors of both the SEM and EIT calibrations, suggests that the calibration of the EIT observations
(Newmark et al. 2000) behind the DEM maps is accurate. While the actual strong numerical
agreement may be fortuitous considering the error estimate for both the observations and the
model results, it is still satisfying to be this close, instead of at the extremes of the error overlaps,
even though this too would represent agreement. The excellent short term rotational time scale
agreement is a strong indicator of the relative accuracy of the model.
We have already extended our wavelength coverage to 170-550 Å, and potentially can model
additional wavelengths. We discuss the wavelength range that we can model in terms of the
contributions to the solar EUV spectrum of emissions from differing temperature regions of the
solar atmosphere, and of the range of applicability of our DEM map method. Our technique is
limited to modeling optically thin emissions from plasmas over the range of temperatures logT =
4.6-6.5 that contribute to the images from the four EIT channels defining the DEM map. This is a
temperature, not wavelength, limitation, and Cook, Newmark, and Altrock (2005), for example,
have compared a DEM map model Fe XIV 5303 Å green line index with the Sacramento Peak
coronagraph green line observations at visible wavelengths. EUV irradiances over 170-550 Å
arise from this EIT temperature sensitivity range, but to shorter wavelengths below 170 Å
emissions will eventually start to arise from plasmas hotter than logT = 6.5.
In Figure 3 we illustrate the temperature contribution functions d I/d logT of the two wavelength
bands 110-120 Å and 170-180 Å for the representative CHIANTI quiet and active region DEM
curves. The integrated area under each curve is the total emission intensity in the band (not the
irradiance), plotted logarithmically. The CHIANTI active region DEM extends to logT = 6.8 and
will produce an intensity contribution at higher temperatures if hotter plasmas contribute
emission to an individual band. Figure 3 demonstrates that the 170-180 Å band can be
realistically modeled by our DEM map technique, while the 110-120 Å band will have
contributions from plasmas hotter than we treat. Flare lines at any wavelength cannot be
modeled.

Fig. 3. Temperature contribution functions for the 110-120 Å and 170-180 Å bands, using
the CHIANTI quiet (line with +) and active region (line) DEM curves. Intensity is plotted
logarithmically.

To longer wavelengths above optically thick Lyman alpha 1215 Å an optically thick continuum
and chromospheric emission lines, together with optically thin transition region and a few
coronal emission lines, contribute emissions. Above 1680 Å the Fraunhofer spectrum begins.
The wavelength range past Lyman alpha cannot be modeled by a DEM technique except for the
individual optically thin transition region and coronal emission lines present. Even to shorter
wavelengths several strong optically thick emission lines occur, especially He II 304 Å, but also
He I 584 Å and H I Lyman beta 1025 Å. We use an observationally determined multiplier to
treat the pure He II 304 Å line separately within the EIT 304 Å channel, and discussed this
treatment in Newmark, Cook, and McMullin (2008). We need to do this to utilize the EIT 304 Å
channel to tie down the cooler end of the DEM curve, but do not intend to push on similarly to
treat other optically thick lines. Finally, above 750 Å the hydrogen Lyman continuum, up to the
photoionization edge at 911 Å, is increasingly important. Although we have added a sketchy
treatment for continua emitted down to logT = 4.0, we have not yet made sufficient comparisons
with available observations to trust this at longer wavelengths where the cool continuum is a
meaningful contributor to the irradiance. We will quantitatively study extending our modeling to
additional EUV wavelengths. We do now have confidence in applying our DEM map technique

over at least the 170-550 Å range. We report our results in 10 Å bins, but can calculate any
wavelength resolution, including individual emission lines, to a resolution where assumptions on
individual line profiles become critical for accuracy.
We have included our approximate treatment for continua formed over logT = 4.0-4.6 by
extending the DEM curve to cooler temperatures than sampled by the EIT channels. We have
always included continua arising from the temperature sensitivity range directly sampled by EIT;
this is just a switch in the CHIANTI program the user sets, and makes little numerical difference
over our standard temperature range. We added the extension down to logT = 4.0 in preparation
to better model the irradiances longward of the EIT channels. When we use this treatment only
over 170-550 Å, in this range it is numerically still a small contribution.
Our DEM map technique produces synthetic daily images, and allows the determination of the
irradiance contribution arising from individual structural features, and the separate calculation of
an on-disk and off-disk irradiance component. The off-disk component becomes increasingly
significant for hotter coronal emissions, and can be appreciated in the synthetic images shown in
Figure 1. This irradiance segmentation is also useful for modeling the projection of the irradiance
viewed in the Earth direction to other, non-ecliptic, viewing directions. While the disk
component can be obtained from a Carrington format map, projected to any viewing angle (see
Auchere et al. 2005a, 2005b, 2005c), the off-disk component cannot be directly projected (short
of tomographic reconstruction) because of the infinite line-of-sight beyond the limb.
We illustrate in Figure 4 a plot of the on-disk and off-disk irradiance histories at top, and at
bottom the statistical relationship between the on-disk and off-disk components, for the 270-280
Å band. We have obtained this type of relationship over the entire 170-550 Å range for each 10
Å band. This division is also interesting because of the different behavior of the two components
with solar rotation. The disk component irradiance is modulated by the passage across the disk of
bright active regions on a 27 day timescale. The off-disk component is also modulated by bright
active regions, but by those at the east and west limbs, and is frequently modulated on a 13 day
timescale 90° out of phase with the disk component, because the line-of-sight is not limited by an
opaque disk (earlier discussed for the Fe XIV 5303 Å index in Cook, Newmark, and Altrock
2005).

Fig. 4. Top: on-disk and off-disk irradiance histories for the 270-280 Å band. Bottom:
correlation of the on-disk and off-disk irradiances.

References
Auchere, F., Cook, J. W., Newmark, J. S., McMullin, D. R., von Steiger, R., and Witte, M.
2005a, Advances Space Res., 35, 388
Auchere, F., Cook, J. W., Newmark, J. S., McMullin, D. R., von Steiger, R., and Witte, M.
2005b, Ap. J., 625, 1036
Auchere, F., McMullin, D.R., Cook, J.W., Newmark, J.S., Von Steiger, R., and Witte, M. 2005c,
Proc. Solar Wind 11 - SOHO 16, ESA SP-592, 327
Brosius, Jeffrey W., Landi, Enrico, Cook, John W., Newmark, Jeffrey S., Gopalswamy, N., and
Larra, Alejandro 2002, Ap. J., 574, 453
Cook, J. W., Newmark, J. S., and Altrock, R. C. 2005, Ap. J., 633, 518
Cook, J.W., Newmark, J.S.,and Moses, J.D. 1999, in Proceedings of the 8th SOHO Workshop:
Plasma Dynamics and Diagnostics in the Solar Transition Region and Corona, ed. J.-C. Vial and
B. Kaldeich-Schumann (Noordwijk: ESA SP-446), 241
Delaboudiniere, J.-P., and 27 co-authors 1995, Solar Phys, 162, 291
Dere, K.P., Landi, E., Mason, H.E., Monsignori Fossi, B.C., and Young, P.R. 1997, A&A
Suppl., 125, 149
Feldman, U., Mandelbaum, P., Seely, J.F., Doschek, G.A., and Gursky, H. 1992, Ap. J. Suppl.,
81, 387
Howard, R.A., Moses, J.D., Socker, D.G., Dere, K.P., Cook, J.W., and the SECCHI Consortium
2002, Adv. Space Res., 29, 2017
Judge, D.L., McMullin, D.R., and Ogawa, H.S. 1999, J.G.R., 104, 28321
Judge, D.L., McMullin, D.R., Ogawa, H.S., Hovestadt, D., Klecker, B., Hilchenbach, M.,
Mobius, E., Canfield, L.R., Vest, R.E., Watts, R., Tarrio, C., Kühne, M., and P. Wurz, P. 1998,
Solar Phys, 177, 161
Mazzotta, P., Mazzitelli, G., Colafrancesco, S., and Vittorio, N. 1998, A&A Suppl., 133, 403
Meier, R.R. 1991, Space Sci. Rev., 58, 1
McMullin, D. R., Judge, D. L., Hilchenbach, M., Ipavich, F., Bochsler, P., Wurz, P., Burgi, A.,
Thompson, W. T., and Newmark, J. S. 2002, in The Radiometric Calibration of SOHO, ed. A.
Pauluhn, M.C.E. Huber, and R. von Steiger (Bern: International Space Science Institute SR002), 135

Newmark, J.S., Cook, J.W., and McMullin, D.R. 2008. This paper can be viewed at
http://sungrazer.nrl.navy.mil/docs/sem.ps.
Newmark, J.S., Moses, J.D., Cook, J.W., Delaboudiniere, J.-P., Song, X., Carabetian, C.,
Bougnet, M., Brunaud, J., Defise, J.-M., Clette, F., and Hochedez, J.-F. 2000, SPIE, 4139, 328



File Type                       : PDF
File Type Extension             : pdf
MIME Type                       : application/pdf
PDF Version                     : 1.4
Linearized                      : No
Page Count                      : 18
Language                        : en-US
Title                           : Proposal No
Author                          : NRL  7660
Creator                         : Writer
Producer                        : OpenOffice.org 3.3
Create Date                     : 2011:09:28 04:57:10-04:00