Sage O2Argo Manual
User Manual:
Open the PDF directly: View PDF 
.
Page Count: 20
| Download | |
| Open PDF In Browser | View PDF |
Last updated: 15 Apr, 2019 SAGEO2-Argo SOCCOM Assessment and Graphical Evaluation for Oxygen Modified for use with International Biogeochemical Argo Floats Software User Manual V1 Prepared by: Tanya Maurer, Josh Plant, Ken Johnson MBARI 7700 Sandholdt Rd. Moss Landing, CA 95039 tmaurer@mbari.org 1 Last updated: 15 Apr, 2019 LICENSE AGREEMENT Note: This software is made available under the MIT License. Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. Authors using the software should acknowledge: “Software was developed and made freely available by the Monterey Bay Aquarium Research Institute in support of the Southern Ocean Carbon Climate Observations and Modeling (SOCCOM) project.” 2 Last updated: 15 Apr, 2019 TABLE OF CONTENTS 1. Context .................................................................................................................................4 2. GUI Setup ............................................................................................................................5 2.1 System requirements ......................................................................................................5 2.2 Toolbox requirements ....................................................................................................5 2.3 Your local float data repository .....................................................................................5 2.3.1 Setting up your float data repository ....................................................................5 2.3.2 Routines to support file acquisition ......................................................................7 2.4 GUI install......................................................................................................................7 3. Running the GUI .................................................................................................................8 3.1 File selection ..................................................................................................................8 3.2 Visualization preferences ...............................................................................................9 3.3 Gain computation using reference datasets .................................................................11 3.3.1 Gain computation using NCEP reanalysis .........................................................11 3.3.2 Gain computation using WOA2013 ...................................................................12 4. Modifying the GUI to suit your needs ...............................................................................12 4.1 Variable input file types ...............................................................................................12 4.2 Exporting gain values for use in DAC processing – suggested workflow .................13 4.3 Incorporating shipboard reference data .......................................................................14 4.4 A note on in-air data storage in Argo *BRtraj files ....................................................15 5. Concluding remarks ..........................................................................................................16 6. References .........................................................................................................................17 3 Last updated: 15 Apr, 2019 1. Context It has been shown that Aanderaa oxygen optodes (in widespread use on board biogeochemical Argo profiling floats) suffer from pre-deployment storage drift and can under-estimate oxygen concentration by up to 20% once deployed (Takeshita et al, 2013; Johnson et al., 2015; Bittig et al., 2018). Many floats are now capable of taking in-air measurements while at the surface which allow for post-deployment calibration of the oxygen optode, bringing data accuracy closer to 1% (Johnson et al., 2015; Bittig et al., 2018). While this is the preferred correction method, post-deployment optode calibration can also be performed using the World Ocean Atlas (WOA) climatology (Takeshita et al., 2013) or by comparing float data to high-quality shipboard measurements. However, currently the amount of adjusted oxygen data available on the Argo GDACs is limited, with the U.S. serving the majority of the Argo Bfiles containing corrected oxygen data. Increasing the amount of science-quality oxygen data available to users in real-time requires simple and consistent correction methods that can be adopted globally across DACs for operational implementation. SageO 2 is a MATLAB Graphical User Interface (GUI) developed at MBARI to assist in deriving oxygen optode gain corrections for use in real-time processing. In this version of the software, oxygen concentrations are corrected using a multiplicative gain factor, G: [𝑂2 ]𝑐𝑜𝑟𝑟 = 𝐺 × [𝑂2 ]𝑟𝑎𝑤 (1) Details related to the calculation of G using different reference datasets follow Johnson et al. (2015) and are described in section 3.2. The code library used to support GUI functionality and calculate optode gain was initially built as part of the Southern Ocean Carbon Climate Observations and Modeling (SOCCOM) float processing workflow at MBARI. The original version utilized raw incoming *.msg files from SOCCOM APEX and NAVIS floats. Modifications to the GUI were recently made to support its use for other float types within the Argo float Global Data Assembly Centers (GDACs) and at other research institutions. The SageO2-Argo software is now available here and can be used to visualize float oxygen data from Argo netCDF files in comparison to WOA climatology and NCEP reanalysis products (used to estimate atmospheric oxygen partial pressure along a float track) in order to derive float-specific gain correction values. These correction factors can then be integrated into the DAC processing stream and used to populate the adjusted oxygen parameter (DOXY_ADJUSTED) within Argo Bfiles. Please note that the software in its current state has undergone limited external testing, is provided as-is, and may require modification to suit the needs of each DAC employing its use. Part of this limitation is due to the inconsistencies in how in-air oxygen data is stored within *BRtraj.nc files across the GDAC database. Additionally, visualizing comparisons to bottle data is highly dependent on the structure of supporting shipboard data files at respective DACs. It is our hope that open access to the software will allow for further testing and improvements to the GUI. Additionally, we hope that the sharing of these tools will help improve the accuracy of the global BGC Argo oxygen dataset and bring the global BGC Argo community closer to maintaining internal consistency 4 Last updated: 15 Apr, 2019 with regards to the quality control adjustments of float oxygen data served on the Argo GDACs. 2. GUI Setup 2.1 System requirements The SageO 2 -Argo GUI was written using the MATLAB programming platform for Windows, release R2015b. Be aware that backwards compatibility and performance on other platforms (Linux, Mac) has not been fully tested. MATLAB must be properly installed and licensed on your machine before proceeding. 2.2 Toolbox requirements There are two freely-available external MATLAB toolboxes that must be downloaded prior to GUI use. The GUI itself was built within the framework of MATLAB’s “GUI Layout Toolbox” which supports the construction of complex layouts with graceful resizing capabilities. Links to download/install information for each required toolbox are listed in Table 1 below. BE SURE TOOLBOX DIRECTORIES AND FUNCTIONS ARE PERMANENTLY ADDED TO YOUR MATLAB PATH. Table 1: External Toolboxes Required for SageO 2 -Argo Toolbox Download Notes GUI Layout Toolbox https://www.mathworks.com/ma tlabcentral/fileexchange/47982gui-layout-toolbox Note the two separate download options for MATLAB versions before and after R2014b. Nctoolbox-1.1.3 https://github.com/nctoolbox/nct oolbox Be sure to permanently add the toolbox setup to your startup.m file. See notes under “setup” at the download link location. The MATLAB NaNsuite, m_map1.4, and SEAWATER functions are also called within the software, but these toolboxes have been included within the code repository under …\ARGO_PROCESSING\MFILES\nansuite\ …\ARGO_PROCESSING\MFILES\m_map1.4\ and …\ARGO_PROCESSING\MFILES\MISC\, respectively. 2.3 Your Local float data repository 2.3.1 Setting up your float data repository The SageO 2 -Argo GUI was built to access specific netCDF and text files, namely the Argo formatted *.BRtraj.nc, *meta.nc, and *Sprof.nc files found on the Argo 5 Last updated: 15 Apr, 2019 GDACs, as well as a textfile conversion of the *Sprof.nc file (ODV*.TXT, produced by the user, see section 2.3.2). Whether you intend to use the GUI to visualize incoming data from a single float, or multiple floats, you will want to have all files organized in a single repository on your local machine or network. By default, the software assigns this directory to …ARGO_PROCESSING/DATA/ARGO_REP/. Figure 1 provides an example of this repository containing a number of Coriolis floats. Each subfolder in the repository is float-specific and holds all float files necessary for data visualization (see Figure 2). Figure 1: Screenshot of an example float data repository on a Windows system. Figure 2: Screenshot showing the contents of a single float subdirectory within the float data repository in Figure 1 (Note, both *.Mprof and *.Sprof are shown. Only one is required. Argo “synthetic” Sprof files may soon fully replace the older merged format, *.Mprof, and the Sprof files are what are currently being supported within the code). 6 Last updated: 15 Apr, 2019 2.3.2 Routines to support file acquisition Within the code library there are routines to assist you in organizing your input files. If you do not already have the Argo NetCDF formatted files on your local machine for your floats of interest, you can (a) download them manually from the Argo GDAC (note that at the present moment only the ifremer GDAC is supplying merged Sprof files), or (b) use the MATLAB function “get_ARGOifremer_files.m” located in …/ARGO_PROCESSING/MFILES/GUIS/SAGE_O2Argo/ for automated file retrieval. Additionally, you will need to generate the ODV*.TXT file using “mprofmat2ODV.m” or “ARGOsprofmat2ODV.m” located in …/ARGO_PROCESSING/MFILES/GUIS/SAGE_O2Argo/. This part of the setup is necessary in order to convert profile data within the Sprof file into the format used in SageO2-Argo. 2.4 GUI install Once you have verified your system requirements, have downloaded and installed the necessary toolboxes, and have set up your data repository, it is time to install the GUI software. This step defines and adds all necessary paths, and also downloads NCEP and WOA files to local repositories. If you haven’t done so already, clone the GUI repository “ARGO_PROCESSING” from github here: https://github.com/SOCCOMBGCArgo/ARGO_PROCESSING and place it somewhere on your local machine (for example C:\Users\USER\Documents\MATLAB\, where USER is the username of the machine). Install steps are as follows: 1) Open MATLAB 2) Navigate to …\ARGO_PROCESSING\MFILES\GUIS\SAGE_O2Argo\ (in other words, change your MATLAB “current folder” to this location). 3) At the command-line, run INSTALL_sageO2Argo (see Figure 3). 7 Last updated: 15 Apr, 2019 Figure 3: Screenshot of MATLAB command line after successful GUI install. 4) You should now be able to launch the GUI by typing: >> sageO2Argo 8 Last updated: 15 Apr, 2019 3. Running the GUI 3.1 File Selection Upon typing “sageO2Argo” at the MATLAB command prompt, you should see the empty GUI interface as shown in Figure 4 below. The left-most column of panels is where the user can select options and actions while the right panel (including “RAW and “QC” tabs) will display the results of these actions. Note that both the main window and each individual input panel can be resized to more easily fit your screen by selecting and dragging panel edges with your mouse. Once here, click on the “Select Float” button near the top left of the interface. This should open a directory dialog box pointing to the location you designated as your data repository during the install process. From here, you can highlight the float subdirectory of interest and click “select folder”. Note that it may take a number of seconds before input data begins to load, depending on the speed of your machine. You may watch the MATLAB command prompt for progress updates. Figure 4: The SageO2-Argo GUI interface. 3.2 Visualization preferences Figure 5 shows an example of the GUI screen once your float data is loaded. The first input panel in the left panel column, titled, “Float Data Specs” will allow you to view your float’s trajectory on a map (see Figure 6), and change the range of profiles, range of 9 Last updated: 15 Apr, 2019 pressures, or threshold distance for crossovers comparisons with high quality shipboard oxygen measurements. Here, we use the GLODAPv2 data set as the shipboard reference (Olsen et al., 2016). Within the second input panel, titled, “Plot Type”, the user is able to toggle between different data views. The “Surface” and “Deep” buttons refer to the data shown in the bottom two time series panels. This is useful for visualizing the oxygen, temperature and salinity time series in relation to your calibration data, represented in the top two time series panels. You can also toggle the “Profile” button. This is where profile data (depth on bottom axis) can be viewed in reference to GLODAPv2, and is where bottle data, if available and implemented into the GUI, would be visualized. The third input panel, titled “Reference Data”, is where the user can toggle between different reference data sets. Section 3.3 describes these options in more detail. The fourth input panel, titled “Oxygen Gain Adjustments”, allows for automatic computation of drifts in gains among user-assigned breakpoints (designated by Cycle number). See Section 3.3.3 for further details. The last input panel, titled “Apply Oxygen Gain Adjustment” allows you to apply the mean gain value (shown in blue to the right of the second subplot) to the data. The pushbutton generates a ODV*QC.TXT file within your float directory which includes the corrected oxygen data. Note that the example shown in Figure 5 is for Coriolis float 6900889. For this float, the DOXY_ADJUSTED variable is already populated within its respective Sprof file. If interested in float data managed by other DACs, we always recommend using the adjusted data provided by the DAC, as they hold the most knowledge with respect to floats under their management, and may employ more sophisticated oxygen adjustment methods (ie, as outlined in Bittig et al., 2018). 10 Last updated: 15 Apr, 2019 Figure 5: The SageO2-Argo GUI interface showing Coriolis float 6900889. (a) oxygen data (pO 2 or %sat) and select reference data (NCEP pO 2 is the default), (b) gain values derived from the top panel (mean gain over the lifetime of the float is shown in blue to the right), (c) oxygen data (umol/kg) from within designated depth and profile range, and (d) temperature and salinity data from within designated depth and profile range. Figure 6: Map window display for Coriolis float 6900889 11 Last updated: 15 Apr, 2019 3.3 Gain computation using reference datasets There are two options for reference datasets within the GUI with which to compare float data and derive gain corrections: (1) NCEP reanalysis surface pressure (converted to oxygen partial pressure), and (2) World Ocean Atlas (WOA) 2013 surface water percent saturation. Either reference can be used for visualizing or computing gain factors at each cycle, although NCEP is the preferred method, assuming your float has in-air measurement capabilities, because it is independent of ocean climatology. Details and relevant equations used in the computation of gain factors specific to each reference dataset are described below. 3.3.1 Gain computation using NCEP reanalysis The preferred product for oxygen gain computation within the SageO2-Argo software is NCEP/NCAR Reanalysis-1 six-hourly surface pressure (https://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.surface.html; Kalnay et al., 1996). This is a Gaussian gridded product with units of Pascals, which are converted to hectopascals (millibar equivalent) prior to proceeding. Within the GUI, NCEP atmospheric surface pressure (PNCEP ) values are then converted to oxygen partial pressure based on the assumption that water vapor is 100% saturated at the sea surface. The calculation follows equation (2) below. The water vapor pressure (𝑝𝐻2 𝑂 , in hPa) is calculated using equation (3), where T represents optode temperature in degrees Celsius (Argo parameter TEMP_DOXY) (Aanderaa Operating Manual TD269, 2009). 𝑝𝑂2 = (𝑃𝑁𝐶𝐸𝑃 − 𝑝𝐻2 𝑂 ) × 0.20946 [ 𝑝𝐻2 𝑂 = 𝑒 52.57− 6690.9 )−4.681×𝑙𝑛(𝑇+273.15 )] 𝑇+273.15 ( (2) (3) The sensor gain that is estimated from air oxygen for each individual profile, i, is then computed using equation (4), as outlined in Johnson et al. (2015): 𝑔𝑖 = 𝑝𝑂2 𝑁𝐶𝐸𝑃 /𝑝𝑂2 𝐹𝐿𝑂𝐴𝑇 (4) where 𝑝𝑂2 𝑁𝐶𝐸𝑃 follows from equation (1) and 𝑝𝑂2 𝐹𝐿𝑂𝐴𝑇 comes directly from the PPOX_DOXY parameter (reported in millibars) within the float’s BRtraj file. In this version of the software the overall gain factor, G, used is then the mean of the n individual g values (equation (5)). Bittig et al. (2018) has described a more complex calculation that can correct for seawater carryover on the sensor during optode sampling while in air. However, at this point in time relatively few floats make the series of measurements required to apply this correction. Implementation of the Bittig protocol has not yet been incorporated into this version of the software, but could be at some future date. 12 Last updated: 15 Apr, 2019 𝐺= ∑𝑛 𝑖=1 𝑔𝑖 𝑛 (5) Be aware that the GUI is designed to retrieve NCEP data off of the web for the most up-to-date products. However, if internet connectivity is slow or unavailable, you may modify the target assignment in the function …/ARGO_PROCESSING/MFILES/FLOATS/getNCEP.m (see lines 96-98) to point to a local directory. We have included pre-downloaded data within the GUI repository (see …/ARGO_PROCESSING/DATA/NCEP_TEMPORARY), although the pres.sfc.gauss.2018.nc file would need to be downloaded to reflect the current date. 3.3.2 Gain computation using WOA2013 If your float is incapable of taking in-air optode measurements, an optode gain correction factor can be derived within the SageO2-Argo GUI using WOA percent saturation. Within the software repository, percent saturation WOA 2013 monthly climatology data (1 degree spatial resolution) are stored for comparison against float data. Percent saturation from the float is calculated following equation (6) below, where the solubility of oxygen (𝑂2𝑆𝑜𝑙 ) is computed as a function of temperature and salinity following Garcia & Gordon (1992) and using solubility constants from Benson and Krause (1984) (see equation 8 and Table 1 in Garcia & Gordon, 1992). Individual gain values, 𝑔𝑖 , are then computed using equation (7), where %𝑆𝑎𝑡𝑊𝑂𝐴 and %𝑆𝑎𝑡𝐹𝑙𝑜𝑎𝑡 represent the mean WOA and mean float percent saturation values for the upper 25 m of the profile, respectively. %𝑆𝑎𝑡 = [𝑂2 ]/[𝑂2𝑆𝑜𝑙 ] × 100 𝑔𝑖 = %𝑆𝑎𝑡𝑊𝑂𝐴 /%𝑆𝑎𝑡𝐹𝑙𝑜𝑎𝑡 (6) (7) Again, in this version of the software the overall gain factor, G, used is then the mean of the individual g values found at each cycle (equation (5)). 3.3.3 Average versus drifting gain Assessment of post-deployment optode stability has given rise to variable results within the literature (Johnson et al., 2015; Bushinsky et al., 2016; Bittig and Kortzinger, 2015; Bittig and Kortzinger, 2017). A recent paper by Bittig et al. (2018) provides a thorough review on this topic, and also presents more recent analysis suggesting that individual optodes may exhibit significant post-deployment drift within +/- 0.6%/yr, while at the same time suggesting the need for further research on the subject. 13 Last updated: 15 Apr, 2019 There is an operational simplicity in utilizing a single average gain correction in the real-time adjustment of optode data. However, characterizing the amount of optode drift between user-defined change points is currently possible within the SageO2-Argo GUI, and was recently put into practice for select floats within the SOCCOM fleet. To auto-calculate the drift in computed gain over the time-series (using select reference data), click the “ADD ROW” button under the “Oxygen Gain Adjustments” input box, and enter the ending cycle for which to compute drift (cycle 156 was entered in the example shown in Figure 7 below). The computed offset (initial gain), b, and slope (drift), m, are calculated using a model I regression of computed gain against cycle time. The gain value applied at each cycle (following equation 1) then becomes: 𝑔𝑖 = 𝑏 + 𝑚(∆𝑇) (8) where ∆𝑇 is the time elapsed since the first cycle (or time at which the drift began). Figure 7: The SageO2-Argo GUI interface showing Coriolis float 6900889 with computed drift in gain. Within the GUI there are two methods to test whether or not the computed drift is worthy of implementing, in a statistical sense. Upon auto-computation of the drift (after cycle selection from the user), a two-tailed T-test is performed to assess whether the calculated slope is significantly different than zero (results are returned on screen 14 Last updated: 15 Apr, 2019 within a pop-up window). Additionally, on the right-side panel of the interface the GUI reports the computed Bayesian Information Criteria (BIC) (Schwarz, 1978) following Equation 9 below, where SSR represents the sum of squared residuals of the model, K is the number of model parameters, and n represents the number of data points. The BIC weighs the number of predictors within a model against the goodness-of-fit, allowing the user to prevent over-fitting of the data (the model with the lowest BIC is always preferred). 𝑆𝑆𝑅 𝐵𝐼𝐶 = log ( 𝑛 )+ 𝐾 log 𝑛 𝑛 (9) 4. Modifying the GUI to suit your needs 4.1 Variable input file types As previously mentioned, this GUI was designed to read profiling float data stored within Argo *.nc files. However, depending on the data manager’s needs, one might wish to modify the GUI to read directly from a float’s incoming msg files. This is how the SageO2 GUI was originally set up for processing and adjusting data for the SOCCOM fleet, and SOCCOM’s “internal” version of the GUI software (managed by data managers at MBARI) can be made available upon request. The mfile …/ARGO_PROCESSING/MFILES/GUIS/SAGE_O2Argo/getall_floatdata_sO2Argo. m is one of the main files requiring modification when switching to alternative input files. 4.2 Exporting gain values for use in DAC processing – suggested workflow Oxygen gain correction values derived using the SageO2-Argo GUI should be applied to raw incoming oxygen data from the float using equation 1 to generate the DOXY_ ADJUSTED parameter, and the resulting adjusted data should be submitted to the GDAC within a float’s B-files. For newly deployed floats, a gain correction value can be derived as early as the fifth cycle, and then applied to subsequent incoming cycles in near real-time, although for such floats it is suggested that gains be revisited periodically throughout the float’s life. Currently, the GUI allows the user to apply the derived gain to the oxygen data and store it within a “corrected” ODV*QC.TXT file. However, automatically exporting average optode gain values outside of the GUI for incorporation into Argo netCDF files is not currently part of the GUI functionality; this task has been left to the user, as each DAC manages a unique processing workflow. However, the schematic in Figure 8 provides an example of the type of workflow employed at MBARI. The use of floatspecific “QC text files”, which store current and historical QC adjustment values for each sensor, are read into the processing stream so that adjusted data parameters get 15 Last updated: 15 Apr, 2019 populated in near real-time. Depending on the user’s needs, initializing and appending gain adjustment values to such a file could easily be performed within the GUI with the addition of a pushbutton and associated callback function. Figure 8: Schematic of potential workflow for incorporating optode adjustments into Argo data files. 4.3 Incorporating shipboard reference data If bottle data from deployment casts are available for your floats and you are interested in incorporating this data into the GUI, you will need to review the functions …/ARGO_PROCESSING/MFILES/FLOATS/get_shipboard_data.m as well as .../ARGO_PROCESSING/MFILES/GUIS/SAGE_O2Argo/redraw_PROF_sageO2Argo .m These functions were built in reference to shipboard data in WOCE Hydrographic Program (WHP) Exchange format, used by CCHDO: https://exchangeformat.readthedocs.io/en/latest/introduction.html and key off of specific associated parameter names, as defined here: https://exchangeformat.readthedocs.io/en/latest/parameters.html. 16 Last updated: 15 Apr, 2019 The easiest approach to incorporating your bottle data into the GUI (requiring the smallest number of code modifications) would be to follow the same parameter naming conventions and formatting as is used in the WHP Exchange format. 4.4 A note on in-air data storage in Argo *BRtraj files As mentioned, the preferred method for deriving oxygen gain correction values involves comparing in-air optode measurements with NCEP reanalysis. Current BGC Argo guidelines suggest storing in-air optode data (as partial pressure of oxygen, in millibars) within BRtraj files as PPOX_DOXY, identified by the general measurement code (MC) 1100 for in-air oxygen measurements (relative measurement codes also apply, see Scanderbeg et al., 2015, Annex F). However, this scheme is not sufficient for programmatically identifying “true” in-air data due to the variability in measurement acquisition among platforms. Table 2 lists examples of various operational “in-air” optode sampling schemes, all of which are currently identified by either MC 1090 or 1099. However, you will notice that some of these measurements are taken in surface water and are not “true” in-air samples. Table 2: Examples of various operational “in-air” optode sampling schemes Example float Platform Coriolis 6902740 Provor CTS4 Coriolis 6900889 NAVIS AOML 5904693 APEX AOML 5904662 APEX In-Air Sample Scheme * Unpumped nearsurface data (optode mounted on a stick, so some samples taken in air) “Surface sequence” (5 samples in surface water followed by 10 samples in air) “Surface sequence” (4 samples in surface water followed by 8 samples in air) Single in-air reading during float telemetry phase MC used in BRtraj file 1100-10 = 1090 1100-10 = 1090 1100-10 = 1090 1100-1 = 1099 *Specific example for float listed (in other words, not necessarily consistent within listed platform type) In the current version of SageO2-Argo, we have imposed a pressure threshold of 0.1 dbar associated with PPOX_DOXY measurements (MC=1090 or 1099) as a potential method to identify “true” in-air data. Users may be interested in modifying this 17 Last updated: 15 Apr, 2019 approach, based on more thorough knowledge of their float’s behavior. Additionally, at ADMT18 it was suggested that the measurement codes used to identify in-air oxygen data within the BRtraj files become more explicit to better represent the activity of the float during in-air measurement acquisition (for example, separate MC for in-water versus in-air surface sequence samples). Thus, be aware that future changes to in-air MCs within the BRtraj files would require code updates. 5. Concluding remarks The SageO2 GUI has been used successfully at MBARI for managing oxygen gain adjustments for over 100 operational floats. It is our hope that the recent modifications to the software for application to Argo floats can assist external DACs in either deriving oxygen gain factors directly, or in developing similar tools to support their respective workflows, ultimately increasing the amount of adjusted oxygen data submitted to the GDAC in real time. We welcome any comments, questions, or suggestions. Additionally, if you would like to help us improve the software by contributing to the code repository on github please let us know! 18 Last updated: 15 Apr, 2019 6. References Aanderaa Data Instruments AS, 2009. TD 269 Operating Manual Oxygen Optode 4330, 4835. Benson, B.B. and D. Krause. 1984. The concentration and isotopic fractionation of gases dissolved in freshwater and seawater in equilibrium with the atmosphere. Limnol. Oceanogr. 29: 620-632. Bittig, H.C., A. Kortzinger. 2015. Tackling oxygen optode drift: near-surface and in-air oxygen optode measurements on a float provide an accurate in situ reference. JAOT. 32: 1536-1543. Bittig, H. C., and Körtzinger, A. 2017. Technical note: update on response times, in-air measurements, and in situ drift for oxygen optodes on profiling platforms. Ocean Sci. 13, 1– 11. doi: 10.5194/os-13-1-2017. Bittig HC, Körtzinger A, Neill C, van Ooijen E, Plant JN, Hahn J, Johnson KS, Yang B and Emerson SR (2018) Oxygen Optode Sensors: Principle, Characterization, Calibration, and Application in the Ocean. Front. Mar. Sci. 4:429. doi: 10.3389/fmars.2017.00429. Bushinsky, S. M., Emerson, S. R., Riser, S. C., and Swift, D. D. 2016. Accurate oxygen measurements on modified Argo floats using in situ air calibrations. Limnol. Oceanogr. 14, 491–505. doi: 10.1002/lom3.10107. Garcia, H.E. and L. Gordon. 1992. Oxygen solubility in seawater: Better fitting equations. Limnol. Oceanogr. 37(6), 1307-1312. Johnson, K.S, J.N. Plant, S.C. Riser, D. Gilbert. 2015. Air oxygen calibration of oxygen optodes on a profiling float array. JAOT. 32: 2160-2172. Kalnay, E. M., and Coauthors. 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437–470, doi:10.1175/1520-0477(1996)077,0437:TNYRP.2.0.CO;2. NCEP, 2015: NCEP/NCAR Reanalysis 1. National Centers for Environmental Prediction, accessed 2 April 2015. [Available online at http://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.html.] Olsen, A., R. M. Key, S. van Heuven, S. K. Lauvset, A. Velo, X. Lin, C. Schirnick, A. Kozyr, T. Tanhua, M. Hoppema, S. Jutterström, R. Steinfeldt, E. Jeansson, M. Ishii, F. F. Pérez and T. Suzuki. The Global Ocean Data Analysis Project version 2 (GLODAPv2) – an internally consistent data product for the world ocean, Earth Syst. Sci. Data, 8, 297–323, 2016, doi:10.5194/essd-8-297-2016. 19 Last updated: 15 Apr, 2019 Scanderbeg Megan, Rannou Jean-Philippe, Buck Justin, Schmid Claudia, Gilson John, Swift Dana, Sato Kanako. 2015. Argo Trajectory Cookbook. http://archimer.ifremer.fr/doc/00300/41152/ Schwarz, Gideon E. 1978. Estimating the dimension of a model. Annals of Statistics, 6 (2): 461–464, doi:10.1214/aos/1176344136, MR 0468014. Takeshita, Y., T. R. Martz, K. S. Johnson, J. N. Plant, D. Gilbert, S. C. Riser, C. Neill, and B. Tilbrook. 2013. A climatology-based quality control procedure for profiling float oxygen data. J. Geophys. Res. Oceans, 118, 5640-5650, doi:10.1002/jgrc.20399. 20
Source Exif Data:
File Type : PDF File Type Extension : pdf MIME Type : application/pdf PDF Version : 1.5 Linearized : No Page Count : 20 Language : en-US Tagged PDF : Yes Author : Tanya Maurer Creator : Microsoft® Word 2010 Create Date : 2019:04:15 16:06:22-07:00 Modify Date : 2019:04:15 16:06:22-07:00 Producer : Microsoft® Word 2010EXIF Metadata provided by EXIF.tools