April 6-7, 2020
Workshop objectives were to:
(1) Review progress of the Salinity Continuity Project (SCP) in the past two years;
(2) Develop consistent metrics for evaluating sea surface salinity retrievals;
(3) Gather feedback on NASA sea surface salinity retrievals from users;
(4) Discuss outstanding issues and the way forward;
(5) Identify synergistic activities of salinity science investigators with the SCP; and
(6) Enhance community building and supporting community research.
Expected outcomes were:
(1) Defining the next phase of infrastructure team activities under the SCP; and
(2) In support of the previous outcome, identifying the main roles of various institutions and synergistic activities of the Ocean Salinity Science Team (OSST).Agenda
Documents: 21Tang, W., Yueh, S., Fore, A., and Hayashi, A.
[06-Apr-20]. An overview of efforts to explore the potential of SMAP SSS in monitoring the Arctic Ocean was provided. The latest JPL SMAP algorithm (version 4.3) experimentally allows SSS retrieval in Level 2 cells with matchup sea-ice concentration (SIC) up to 3% (without SIC), as opposed to 0.5% SIC previously used (version 4.2). Key elements of a proposed future empirical SIC algorithm were also presented including plans to implement a pre-processor for routine Level 2B SSS retrieval. Lee, T.
[06-Apr-20]. Study concluded that the consistency between SMAP RSS version 4 SSS and Argo data is approaching that between Aquarius version 5 and Argo (evaluated at 1° grid, monthly scales). It suggested that better understanding of error in Argo gridded data will improve the estimate of satellite SSS uncertainties. While bias, standard deviation, and RMSD are common metrics to evaluate satellite SSS using in-situ products, additional metrics should also be examined. Vazquez, J., Gomez Valdes, J., and Bouali, M.
[07-Apr-20]. A new co-location strategy developed for the derivation and comparison of SSS gradients was presented. Comparison data from two Saildrone deployments – California/Baja in 2018 and Gulf Stream in 2019 – and RSS SMAP (version 4) and JPL CAP (version 4.2) salinities were shared. Overall, comparisons of gradients for SSS show correlations of <0.2 for the two regions. In the Gulf Stream, SSS showed clear relationships to major frontal features.
View the movies that accompany this presentation: RSS SMAP Gradient
and JPL SMAP Gradient
. Kao, H.-Y., Schanze, J., Le Vine, D., and Dinnat, E.
[07-Apr-20]. Presentation covered comparisons for different match-up approaches between SMAP satellite-retrieved and in-water salinity measurements. Using recommended criteria (i.e., "all-in-box" averages, search radius of 50 km; search window of +/- 2.5 days), two temporal search approaches were covered: in-situ centric and satellite-centric. Several spatial search approaches were discussed: averaging all points, closest point in space, closest point in time, and weighted functions. Level-2 triple point analysis results were also shared. Meissner, T., Wentz, F., and Manaster, A.
[06-Apr-20]. Overview of Salinity Continuity Processing System (SCPS) was provided along with updates in Remote Sensing Systems data product version 4 (e.g., improved land correction). Much of the talk focused on analyzing and correcting biases in the Southern Ocean, which may be related to wind direction. Le Vine, D., Lang, R., Dinnat, E., Soldo, Y., de Matthaeis, P., and Zhou, Y.
[06-Apr-20]. Brightness temperature and potential salinity retrieval over the Great Salt Lake (UT) by SMAP were discussed. The Great Salt Lake is near saturation (ranging from 120 psu to 280 psu). This provides motivation to examine the dielectric constant using Debye and polynomial model functions at extreme salinities relative to the open ocean (i.e., above 40 psu). Lang, R., Zhou, Y., Dinnat, E., and Le Vine, D.
[06-Apr-20]. Dielectric measurements since 2018 were presented. Also, the advantages of using a Debye formula, where the conductivity term can be determined by the imaginary part of the seawater measurement data, were discussed. The authors obtained good agreement between the retrieved SSS and Argo data by using a George Washington University Debye model and found that it is more stable than the previously used polynomial model. Le Vine, D., Soldo, Y, and Dinnat, E.
[07-Apr-20]. The presentation provided an overview of the issues associated with satellite-derived and in-water salinity measurement match-ups. The talk focused on Aquarius and SMAP SSS, including statistics for several match-up options (e.g., closest point of approach, CPA, in space and/or time; CPA with averaging; and "all-in-box" sampling). The overall recommendation is using the "all-in-box" method for both SMAP and Aquarius with a radius of 50 km and time of +/- 3.5 days. Vinogradova Shiffer, N.
[06-Apr-20]. Presentation parsed the roles of the NASA Ocean Salinity Science Team (OSST) in terms of infrastructure and research activities. Goals, accomplishments, priorities, and future plans for the OSST were discussed. Broadening the use of salinity data beyond the OSST – e.g., in other Earth Science disciplines – was also addressed. Manaster, A.
[07-Apr-20]. An overview of the Salinity Continuity Processing System (SCPS) system was provided including a flow chart of the salinity retrieval algorithm. The talk focussed on the wide variety of ancillary data input sources including wind speed, atmospheric profiles, sea surface temperature, reference salinity, rain rate, and sea ice. A key part of this process is monitoring quality of incoming ancillary data, which is an integral part of the SCPS. Future plans include developing commonality between SMAP and Aquarius ancillary inputs for studying decadal changes in salinity. Tsontos, V., Vazquez, J., and Jiang, Y.
[07-Apr-20]. Recent updates at the NASA Physical Oceanography Distributed Active Archive System (PO.DAAC) related to salinity were presented. This includes Salinity Continuity Program (SCP) archival and distribution milestones: 7 new satellite data and 14 new in-situ datasets were released. An overview of science and user support services for the SCP was given along with web-based tools for enhanced NASA field campaign support. Authors also asked OSST and SCP members to consider contributing their own datasets to PO.DAAC. Schanze, J., Sabia, R., Guimbard, S., Reul, N., Lee, T., Le Vine, D., Dinnat, E., Vinogradova-Shiffer, N., Bingham, F., Kao, H.-Y., Carey, D. et al.
[07-Apr-20]. An overview of the Salinity Pilot Mission Exploitation Platform (Pi-MEP) was provided. This online platform for quality control of satellite salinity data is part of a collaborative effort between the European Space Agency (ESA) and NASA. Pi-MEP supplements U.S. salinity validation efforts and offers powerful statistical tools and hundreds of datasets. A key element of the collaboration will be unifying match-up criteria (e.g., in-situ centric, 50 km search radius, +/- 3.5 day window). Schanze, J., Sabia, R., Guimbard, S., Reul, N., Lee, T., Le Vine, D., Dinnat, E., Vinogradova-Shiffer, N., Bingham, F., Kao, H.-Y., Carey, D. et al.
[07-Apr-20]. An overview of the Salinity Pilot Mission Exploitation Platform (Pi-MEP) was provided. This online platform for quality control of satellite salinity data is part of a collaborative effort between the European Space Agency (ESA) and NASA. Pi-MEP supplements U.S. salinity validation efforts and offers powerful statistical tools and hundreds of datasets. A key element of the collaboration is unifying match-up criteria (e.g., in-situ centric, 50 km search radius, +/- 3.5 day window). Meissner, T. and Mears, C.
[06-Apr-20]. Even a small amount of sea ice creates a large source of error in salinity retrievals. An update was given on sea-ice flagging mitigation for SMAP salinity retrievals (version 4). In the future, a far-sidelobe correction (similar to what is done for land) could be implemented; however, sea-ice concentration information is not reliably accurate. Results from case studies were shared on potential improvements near the sea-ice zone. Grodsky, S., and Vandemark, D.
[06-Apr-20]. The main objective of this study was to look at SMAP-buoy comparisons for three buoys: off the coast of Cape Cod, Gulf of Maine, and Irminger Sea (40 to 60°N). Several years of SMAP and in-situ time series of SSS were examined, from which SSS error time series were estimated. SSS retrieval errors at some of these buoys were increased by land contamination and presence of cold water. Bingham, F.
[07-Apr-20]. Subfootprint variability (SFV) is a source of "error" for satellite SSS because of a mismatch of scale between the satellite and in-situ measurements used for validation. Sources of SFV includes surface flux (e.g., precipitation), internal ocean variability, fronts, and mean gradients. The presentation focused on statistical analysis of data collected during the SPURS-1 and SPURS-2 field campaigns, regions of low and high precipitation, respectively. Menezes, V.
[07-Apr-20]. An assessment of north Indian Ocean dynamics using SMAP SSS products (RSS V4 and JPL V4.2) was presented. In addition, the author discussed whether SMAP can capture SSS in the Red Sea, which is challenging due to technical and other issues. In much of the Red Sea, SMAP has good signal-to-noise ratios and can capture the seasonal intrusion of fresher and colder Gulf of Aden water. On the other hand, SMAP performance is much worse in the Gulf of Aden and Gulf of Oman. Drushka, K., Jacob, M., Asher, B., and Jones, L.
[07-Apr-20]. Vertical salinity gradients generated by rain affect satellite SSS retrievals and affect water cycle processes. The presentation provided an update to the Rain Impact Model (RIM) developed for Aquarius, which over-predicted sea surface freshening in higher winds (over 4 m/s). The updated model (known as "PRIM") parameterizes vertical diffusivity based on wind, rain, and time and could potentially be used for SMAP retrievals. Yu, L.
[06-Apr-20]. Evaluation of the mean and variability of six SSS products in the subpolar North Atlantic were presented. Mean data products differ in low SSS retrievals near the Greenland coast and in the Labrador Sea. For seasonal cycles, observed patterns differ vastly among products. Changes in precipitation (P) minus evaporation (E) patterns indicate that atmospheric freshwater input was a major source of a 2016 low-salinity event in the North Atlantic (that was not captured by SMOS). Meissner, T.
[07-Apr-20]. A brief update on instrument calibration for Version 4 of SMAP processing was provided. RSS performs its own calibration on radiofrequency interference (RFI) filtered antenna temperatures from SMAP Level-1 B files. This includes ocean target calibration and corrections for antenna spillover, antenna patterns, and reflector emissivity (which is monitored for possible drift). Future changes of back-end calibration system in SMAP L1B processing will necessitate new analysis and possible calibration changes in salinity processing. deCharon, A., and Taylor, L.
[06-Apr-20]. New features on the "NASA Salinity" website were highlighted. This included new website sections, "Which Salinity Data Are Best For You?" and "Learn More." An overview of the "How Salty (or Fresh) Are You?" quiz and associated e-brochures was given. Participants were encouraged to submit one-page "Research Highlights" for inclusion on the website. Also, next steps in website development were shared.