Report of the Third Workshop

Summary: The meeting objectives were twofold. The first day addressed Field Programs and Algorithms. Several airborne experiments were carried out in 1999 off the U.S. East Coast measuring salinity transects across the Gulf Stream front with various radiometers. Preliminary analyses were presented and are very encouraging. Effects of surface roughness and, in some cases sun glint, are apparent, leading to progress in determining the scope of these corrections. Discussions then focused on algorithm development for satellite sensors and future airborne/field experiment planning and coordination.

The second day was devoted to Satellites and Science. As background, during 1999 the Soil Moisture Ocean Salinity (SMOS) mission was approved by the European Space Agency and began a two-year phase-A study. Another recent development is that plans are being made now to propose a high-precision salinity satellite mission to NASA this year. The status of various space agency programs and other developments were reviewed. This was followed by several scientific discussions that are pertinent to these satellite missions. We concluded with a list of some of the key science questions and open discussion of requisite measurement criteria for spatial and temporal resolution and accuracy.

The meeting adjourned after a short discussion about the next meeting and possible venue. A recommendation was made, pending a favorable outcome of a NASA salinity mission proposal, that the respective SMOS and NASA teams merge together into a common working team for both projects. Accordingly, this would ultimately replace the SSIWG, which would no longer have need to exist independently. An invitation was made by the SMOS team to have the next meeting in Europe. It was also suggested to have it jointly with the Oceanology International America conference in Miami, April 2001. The consensus was to postpone a decision of future meetings until the situation is more clear, while recognizing the need to meet again in some form by early next year.

Field Programs and Algorithms

Airborne sensors and experiments in 1999: Stephen Howden and David Levine described the preliminary results of the 1999 NASA P3 Gulf Stream surveys with ESTAR and SLFMR. Four ocean flights were made in late August and early September which included overflights of ship tracks by the RV Cape Henlopen. Sun glint was a significant error source at times. A night flight on 29 August provided an excellent comparison between surface salinity retrieved from ESTAR and that observed by the C. Henlopen (Fig. 1). Preliminary comparisons with SLFMR, which is V-pol, indicated a greater bias with ESTAR (H-pol) during higher wind conditions. Processing of both data sets is continuing and a PI workshop is planned in March. For more information, contact Stephen Howden (howden@nemo.gsfc.nasa.gov).

Figure 1: Comparison of ESTAR aircraft sensor salinity with that observed by the surface ship C. Henlopen on a transit across the western edge of the Gulf Stream.

A Passive/Active L/S-band (PALS) Microwave Aircraft Sensor for Ocean Salinity was described by Bill Wilson. Optimal retrieval accuracy and stability are the primary design criteria for this sensor. It includes dual polarized passive and active(radar) channels and a ~45o viewing angle to optimize the polarimetric information for resolving surface roughness effects. The 1999 flight tests indicate that absolute accuracy is ~1.0K, with a calibration stability of ~0.1K, which equates to about 0.2 psu in salinity. The S-band channels were marginally useful because of radar interference in the Cape Hatteras region. Simon Yueh described further the preliminary analyses of PALS test flights in 1999. Each flight day required a different absolute bias calibration which varied by about 1K. The biases were also different for H and V pol (H>V). Within one flight day, results were generally repeatable to 0.05-0.1K or 0.1-0.2 psu. The roughness effects were quite apparent and when corrected empirically using the radar data, the H and V pol salinities were very consistent, except in the inshore regions on certain flights. The plane track was offset from the ship track by ~10 km because of air traffic control restrictions. Data comparisons indicated a sharp front in the ship data that was more smooth in the PALS data (Fig. 2).

Figure 2. The upper two panels plot the L-band vertically and horizontally polarized brightness temperatures (Tv and Th) and radar measurements (sigma0). The excess brightness temperatures induced by the surface roughness have been estimated from the radar sigma0 data with a linear model. The brightness temperatures corrected for the surface roughness are plotted in green and blue in the upper panel. We have retrieved the surface salinity from the radioemter Tv and Th data with the dielectric constant model by Klein and Swift [1977]. The SSS estimated from uncorrected Tv and Th data are illustrated in the third panel from the top. The estimates from corrected Tv and Th are plotted in the fourth panel together with the ship data. The fifth panel illustrates the estimated errors due to the surface roughness. The sixth panel plots the ship and aircraft observations of SST. The bottom panel plots the aircraft and ship tracks.

The IGARRS meeting in Honolulu July 24-28 will be the next opportunity to present the results from the above experiments and is a milestone date for finishing the initial processing and analyses of these data sets.

US Navy Programs: Jerry Miller (NRL) explained a possible new US Navy requirement for coastal remote sensing of salinity. He then described the next generation SLFMR being procured by NRL for continued airborne experiments. The new sensor will have about 6x lower noise than the current SLFMR, and may include a Step Frequency Microwave Radiometer (SFMR) for roughness correction. Several experiments are planned for years 2000-2001 using both the old SLFMR this year and the new version after delivery around the end of the year.

European Tower Platform Experiments: Jordi Font described the ESA Wind and Salinity Experiment (WISE) planned for October 2000; a 1-month experiment from the Casablanca Oil Platform near Barcelona in the Mediterranean Sea. Surface observations will include S, T, wind and sea state. An air-sea interaction (ASIS) buoy from H. Graber (U. Miami) may also be provided. An L-band radiometer will be furnished by the Polytechnic University in Barcelona, being built from components of the MIRAS demonstrator. It will include step motors to adjust a range of incidence and azimuth angles, and will be oriented northward away from the sun to avoid sun glint. An additional radiometer is refurbished for this experiment by U. Massachusetts, as described by Eric Knapp. It will be a dual channel L and S-band, and dual-pol in L-band. Jacqueline Etcheto described a stereo camera system which will be used to measure breaking wave foam and 2-D wave topography. The fractional foam coverage of the surface tends to vary as the wind speed squared and the data will be used to study the effect of foam on L-band brightness temperature. Following WISE, there may be a similar experiment from a North Sea platform to evaluate lower SST and higher wind regimes.

Future Field Experiment Opportunities: Gary Lagerloef described a few other field experiment opportunities on the horizon. Hans Graber will have ASIS buoys in the Gulf of Mexico associated with various ocean surveys that may provide overflight opportunities. A major air-sea interaction experiment called EPIC is planned by NOAA/NSF in eastern tropical Pacific August, 2001. The focus of the experiment is at 10N, 95W. This will include the NOAA P3 for atmospheric dropsondes and two surface ships: NOAA Ship Ron Brown and RV New Horizon. It has been learned from R. Warner (NOAA) that NOAA has funded the installation of the SLFMR on the P3 aircraft this year, so it will be available for EPIC. Jerry Miller suggested using the SLFMR-2 if it can fit in the SLFMR mount. The Ron Brown will have a precipitation radar, so that it may be possible to make quantitative measurements of the effect of rain on the L-band measurements during this experiment. In other experiments, Stephen Howden and Chet Koblinsky intend to make a transect with SLFMR across the Labrador Sea when the NASA P3 goes to Greenland later this year. Lastly, Jerry Miller indicated that a plan for a coastal experiment from a pier in Duck, N. Carolina is beginning to be discussed. In 2001, Nils Skou, Denmark, plans to deploy an L-band radiometer in flights over the North Sea to study wind effects at various azimuth angles.

Summary of Field programs planned 2000-2001:

Satellite Algorithms and Error Sources: Jacquelin Boutin reviewed her study of errors induced by the SST and sea state on the retrieved SSS in the SMOS configuration. Errors due to wind speed effects dominate over SST. Assuming ~20 samples per pass at various incidence angles in the SMOS field of view (see http://www-sv.cict.fr/cesbio/smos), and uncorrelated errors, she obtains a net error of 0.5 psu including brightness temperature error. Ideally, with a minimum 40 km spatial resolution, errors of 0.2 to 0.1 psu are obtained by averaging over 100 km to 200 km areas respectively. Simon Yueh then reviewed a number of additional error sources and budgets, including atmosphere absorption, Ionosphere, solar and lunar reflectance, galactic reflectance, hydrogen line and surface roughness, with a residual rss error of 0.2K (0.4 psu) for the OSIRIS satellite concept. (See also the SSIWG-2 meeting report).

Satellites and Science:

Satellite-related developments in the past year: Gary Lagerloef began with a brief review of highlights:

1. The approval in May by ESA for the SMOS Phase-A study, which began September 1999 and goes for two years.
2. The initial SMOS team meeting was held in Barcelona in September,
3. The OceanObs99 conference in St. Raphael, France in October in which salinity played a prominent role in discussions and several posters were presented on remote sensing,
4. The ADEOS Symposium in Kyoto Japan in December in which Gary Lagerloef and Simon Yueh gave invited talks on salinity remote sensing from satellite and the meeting summary included a recommendation to NASDA to include salinity measurements in the future Global Climate Observing Mission (GCOM) program,
5. Discussions over the past few months concerning a high precision salinity mission proposal to make to NASA in 2000.

Upcoming events:

1. A Special Session OS13 "Scientific Applications of Surface Salinity Measurements from Space" has been organized for the Western Pacific Geophysics Meeting (WPGM) in June 27-30 2000, Tokyo. Papers which address scientific applications of satellite salinity data and help refine accuracy and resolution needs are especially encouraged. (http://www.agu.org/meetings/wp00top.html)
2. A special session on salinity remote sensing is organized for the IGARRS meeting, July 24-28, Honolulu. (Convenors David LeVine (dmlevine@meneg.gsfc.nasa.gov) and Cal Swift ( klemyk@ecs.umass.edu).
3. "Oceans from Space" meeting to be held in Venice 9-13 October 2000. This meeting is held only once every 10 years and is intended to review progress during the preceding decade and to assess late-breaking developments which may lead to substantial new capabilities during the coming decade. (See http://www.me.sai.jrc.it). Jerry Miller has agreed to organize a special session on salinity remote sensing.
4. The Oceanology International Americas meeting in April 2001, Miami, will include a NASA Ocean Investigators Workshop and an Oceanography Society meeting, and may serve as a useful venue for a salinity symposium.

CLIVAR: Chet Koblinsky presented an overview of CLIVAR, the Upper Ocean Panel (UOP) and OceanObs99, and described several relevant implementation panels, such as the Data Buoy Cooperation Panel (DBCP) and the Ship Of Opportunity Program Implementation Panel (SOOPIP) which are analogous to the TAO Implementation Panel (TIP) and others which exist under CLIVAR. The ARGO profiling float system (http://www.argo.ucsd.edu) deployment is beginning this year. Full ARGO deployment of 3000 floats worldwide with a 10 day cycle would yield an estimated ~400 simultaneous surface observations per month globally assuming a 100 km satellite footprint. This illustrates the value of a coordinated effort between satellite and in situ observations. Ragu Murtugudde reported that the CLIVAR Indian Ocean Implementation Plan is due out soon. SOOPIP and ARGO groups will meet March at Scripps and Southampton, respectively. The TAO panel is planning to meet in November.

Future plans of all these panels can be found through the CLIVAR homepage: http://www.clivar.ucar.edu/org_new.html.

Profiling Floats: Ray Schmitt described recent studies with autonomous profiling (PALACE) floats in the tropical Atlantic. Initial profile data calibrated very well with simultaneous ship CTD. Some problems in the top few bins due to delays in the pressure sensors have been corrected in recent buoy systems. Interpolated SST maps from float data agree with Reynolds SST maps, and maps of vertical salinity maxima agree with old Worthington maps. Sea surface salinity (SSS) maps showed some unexplained anomalies, possibly related to rain or Amazon River water lenses. The ARGO program has funds for pilot studies in the Atlantic and Pacific. While there is a U.S. commitment for 1000 floats and for another 2000 from international partners, the overall funding remains uncertain. Floats will cost about $12K each. The aim is to deploy about 750 per year which will maintain an array of ~3000 assuming an average 4-year lifetime. A workshop was proposed for manufacturers and users to address in situ salinity sensor technology, including new products, performance, calibration stability and biofouling. Measurement platforms requiring this technology include profiling (PALACE, ARGO) floats, moorings, ship thermosalinographs, and surface drifters (which tend to lose calibration within a month). It was noted that PALACE floats in the Labrador Sea have experienced negligible drift in salinity calibration for more than 2 years, however floats in the Pacific which were on a 3-day cycle tended to foul quickly. David Halpern commented that a primary objective for this group is to promote SSS measurements from surface drifters and to solve that problem of calibration drift and fouling. He also emphasized the cost-effectiveness of expanding the SOOP thermosalinograph network.

Model and Data Studies:

Zuojun Yu described preliminary numerical modeling results regarding SSS and surface freshwater forcing. Using a 4.5 layer model with mixed layer physics in a 15-year integration, she compared model SSS fields with Levitus climatology with forcing from CMAP, GPCP, deSilva(NODC) and NCEP/NCAR fields. The GPCP provided the best fit. CMAP showed similar P-E patterns to GPCP and higher amplitudes. The NCEP results were the worst, particularly in the SPCZ. The results indicate that the satellite-derived precipitation products of GPCP and CMAP have the correct geographic patterns and SSS data with models will have value in scaling the amplitude of the P-E data.

Ragu Murtugudde described recent results from two modeling studies using an ocean model coupled to an advective atmospheric mixed layer model. In the first study the effects of river inflow into the Bay of Bengal were examined. It was found that, except in the immediate vicinity of the river plumes, SST was not appreciably changed by the inclusion of river input. This counter-intuitive result occurs because although the mixed layer is thinner and fresher with river input penetrative radiative loss is greater, so the surface layer temperature does not rise from barrier layer effects. The second study is a simple initial step in an attempt to quantify the benefit SSS data will have on climate prediction.

In the control run, the model was run in the Pacific Basin with interannual winds and climatological precipitation, solar radiation, and cloudiness. The second run differed from the control run in that the surface SSS was relaxed to Levitus (1984) climatology. Changes relative to the control run were largest in the central equatorial Pacific, where SST was 0.2-0.8 C cooler and surface current anomalies were westward. The next step is to examine potential feedback's in an atmospheric model forced by these SSTs

Ming Ji discussed the NOAA/NCEP assimilation efforts. The goals are 1) estimate salinity variability and impacts, 2) optimize the use of altimeter data and 3) assimilate velocity data. It has been shown that 5-8 cm dynamic height differences are associated with interannual salinity variations. Methods have been developed to estimate S(z) profiles from T(z), altimeter heights and SSS using EOF techniques. Because most of the variability is near the surface, the SSS data reduces the error of pseudo S(z) profiles from these EOFs. Assimilating the data in the NCEP model produces the largest changes near the date-line, where an estimated S(z) error of about 0.5 psu over the top 130-150m of the upper ocean will contribute to about 5 cm sea level error. This adjustment is also accompanied by a significant change in zonal currents with more westward anomaly and reduced EUC in the east-central basin and more eastward anomaly in west of the date line. Broad area coverage from satellite would be much more useful that individual ship track lines, where the in situ data is used instead for bias removal of the satellite data. Data would be used on a weekly basis to update the model.

Gary Lagerloef presented figures from a new paper by Johnson and McPhaden (JGR, 2000) analyzing a large number of cross equatorial CTD transects. Significant upper layer salinity variations in excess of 1 psu are evident during the 1997-1998 Nino-Nina cycle in both the western and eastern Pacific.

Detlef Stammer reviewed a NOPP-funded program ECCO (Estimating the Circulation and Climate of the Ocean) being done by a consortium of institutions and investigators. Presently there is completed a 6-year ocean state estimation experiment on 2x2 grid, with a 10 m surface layer and NCEP flux forcing along with assimilation of various in situ an satellite data sets. The assimilation provides a calculation of the freshwater flux correction to the NCEP fields, with adjustments of 1-1.5 m/year in the warm pool and SPCZ, and opposite signs in the ITCZ. An annual change of 0.1 psu represents a ~0.1 m/year rainfall anomaly, which is an order of magnitude smaller than the NCEP corrections. Precise satellite SSS data will improve the ocean state estimation in the following ways:

1. More information about SSS time variability.
2. Constraint on surface freshwater flux.
3. Enhanced information on precipitation.
4. Constraint on mixed layer model.

Satellite Mission Status and Related Studies: Jordi Font provided an update of the SMOS mission status and related studies in Europe. The project is now in an extended Phase A study until the Phase B review in Fall 2001. The industrial aspect of the Phase A begins in June 2000. The Phase C/D review in Fall 2002, and the projected launch date is Spring 2005. There are three related ESA studies commencing:

1. The MIRAS demonstrator project proof-of-concept study for the 2-D interferometry image reconstruction.
2. A Salinity study aimed as assessing the impact of salinity observations on climate research.
3. A Soil Moisture study to determine critical requirements and science impacts for soil moisture data.

Jacqueline Boutin and Jacqueline Etcheto reviewed the ESA ITT (Invitation to Tender) proposal to assess and define science requirements and impacts of space SSS observations for modelling and climate studies. The proposal is coordinated by H. Drange of the Nansen (NERSC) Center in Bergen, and is still in review. Part 1 addresses SSS retrieval accuracy and algorithms including roughness and foam effects, rain effects, diurnal SST changes, validating emissivity models, SSS uncertainties globally and regionally. Part 2 consists of model impact studies coordinated by P. LeTraon, CNES. Using models for the tropics, the Atlantic, and the N. Atlantic-Arctic, studies will address 1) SSS sensitivity to E-P, runoff and ice melt, 2) SSS variability in space and time and relation to ocean dynamics, 3) Comparison of SSS mapped errors with model SSS variability.

Chet Koblinsky then presented a summary of evolving plans for a high-precision salinity mission to propose to NASA under the ESSP program later this year. The ESSP is a science demonstration program soliciting missions costing between $60M and $120M, allowing three years from project start until launch and a plan to provided validated measurements. Prime Mission Requirements and Minimum Mission Requirements must be specified. The objective of the team is to build and launch the mission and validate the data, which then become open to science use. Validation is a key issue, and could include independent efforts and efforts included within the project. Science requirements drive costs and evaluations must be made concerning accuracy, orbit parameters, repeat cycle, spatial resolution, fractional global coverage, co-located surface validation data, etc. Groups at JPL and GSFC are studying configuration options. Negotiations to form the proposal team are in progress.

Science Requirements: A plenary discussion ensued regarding science requirements. The following four science problems were posed and preliminary accuracy, spatial and temporal requirements were suggested as the minimum to sustain scientific progress:

1. Barrier layer effects on tropical Pacific heat flux

0.2 psu

100 km

30 days

2. Halosteric adjustment of heat storage from sea level

0.2 psu

200 km

7 days

3. N. Atlantic thermohaline circulation

0.1 psu

100 km

30 days

4. Surface freshwater flux balance

0.1 psu

300 km

30 days

The North Atlantic thermohaline circulation and convection in the subpolar seas has the most demanding requirements, and is the most technically challenging because of the lower brightness/SSS ratio at low water temperatures. It was recommended that this be established as the Prime Mission Requirement, allowing for the others to be met by Minimum Mission Requirements as appropriate.

Science Teams and Working Groups: Gary Lagerloef proposed that pending a favorable outcome of a NASA salinity mission proposal, the respective SMOS and NASA teams merge together into a common working team for both projects. Discussions on the future of SSIWG and the next meeting are reported above in the Summary.

Fini

Participant List:

Boutin, Jacqueline ( jb@lodyc.jussieu.fr)
Boyer, Tim ( tboyer@nodc.noaa.gov )
Chao, Yi ( yc@pacific.jpl.nasa.gov )
Etcheto, Jacqueline ( je@lodyc.jussieu.fr )
Flament, Pierre ( Pierre.Flament@ifremer.fr )
Font, Jordi ( jfont@icm.csic.es )
Gloersen, Per (per.gloersen@gsfc.nasa.gov )
Halpern, Dave (halpern@pacific.jpl.nasa.gov )
Hein, Jeff (jhein@pop700.gsfc.nasa.gov )
Howden, Stephen (howden@nemo.gsfc.nasa.gov )
Ji, Ming (mingji@ncep.noaa.gov )
Knapp, Eric (knapp@mirsl.ecs.umass.edu )
Koblinsky, Chet (koblinsky@gsfc.nasa.gov )
Lagerloef, Gary ( lagerloef@esr.org )
LeVine, David ( dmlevine@meneg.gsfc.nasa.gov )
Li, Fuk (Fuk.K.Li@jpl.nasa.gov )
Lindstrom, Eric (elindstr@hq.nasa.gov )
Liu, Tim ( liu@pacific.jpl.nasa.gov )
McLaughlin, David ( dmclaugh@ecs.umass.edu )
Miller, Jerry ( jmiller@nrlssc.navy.mil )
Murtugudde, Ragu ( ragu@seetha.gsfc.nasa.gov )
Njoku, Eni (eni.g.njoku@jpl.nasa.gov )
Reising, Steven ( reising@ecs.umass.edu )
Schmitt, Ray ( rschmitt@whoi.edu )
Stammer, Detlef ( detlef@fjord.ucsd.edu )
Stokes, Diane ( dstokes@ncep.noaa.gov )
Wilson, Bill ( William.J.Wilson@jpl.nasa.gov )
Yu, Zuojun ( zuojun@pmel.noaa.gov )
Yueh, Simon ( simon@stokes2.Jpl.Nasa.Gov )