In order to retrieve the sea surface salinity (SSS) using L-band radiometry from space, a model for the dielectric constant of sea water is needed. Uncertainties in the relationship between the dielectric constant, the sea surface temperature (SST) and the SSS are difficult to tackle because of two main challenges: 1/ measuring the dielectric constant in laboratory with a high accuracy ; 2/ discriminating, from satellite retrieved parameters, dielectric constant model flaws from geophysical/instrumental effects (roughness, sea ice, etc…).
The dielectric constant models of Klein and Swift (1977) (KS) and Meissner and Wentz (2012) (MW) are respectively used in the current SMOS and AQUARIUS/SMAP processing algorithms. Recent laboratory measurements by the University of Washington (UW) (Lang et al (2016)) and the associated parametrization (Zhou et al. (2017)) indicate a dependency for low SST close to MW but also suggest SSS dependencies at moderate to warm SST that differ from both MW and KS.
Waldteufel et al. (2004) have shown that using the angular dependency of SMOS Tb, it is possible to infer information about the modulus of the dielectric constant, the so-called Acard parameter. Acard and SSS retrieved from SMOS measurements are compared with the ones derived from in situ fields of SSS and SST and KS, MW and UW relationships. ACARD derived from SMOS Tbs supports a SST dependency at 34pss different from KS, consistent with MW and Zhou relationships. Comparisons of SMOS ACARD with MW and UW ACARD point out weaknesses of relationships obtained from a limited set of measurements:
-UW measurements taken every 5°C, at 30, 33, 35, 38pss are too scarce to constrain SSS variations in UW relationship between 34 and 37pss; in this range of SSS, SMOS ACARD agrees with MW ACARD,
-SMOS ACARD is closer to UW ACARD than to MW ACARD for ‘extreme’ SSS and SST values observed over the ocean (less than34pss, higher than 38pss at 30°C),
These results have been obtained for wind speeds between 5 and 9m/s, after thoroughly filtering outlier measurements; no dependency with wind speed within this range has been observed. SMOS ACARD are in most cases within the error bars of UW laboratory measurements. Taking into account these results, parametrization adjustment of the L-band dielectric constant based on SMOS ACARD observations are discussed.


Klein and C. T. Swift, “An improved model for the dielectric constant of sea water at microwave frequencies,” IEEE Trans. Antennas Propagat., vol. AP-25, no. 1, pp. 104–111, 1977.
Lang, R., Y. Zhou, C. Utku, and D. Le Vine (2016), Accurate measurements of the dielectric constant of seawater at L band, Radio Sci., 51, 2–24, doi:10.1002/2015RS005776.
Meissner T. and F. J. Wentz, “The emissivity of the ocean surface between 6 and 90 GHz over a large range of wind speeds and earth incidence angles,” IEEE Trans. Geosci. Remote Sens., vol. 50, no. 8, pp. 3004–3026, Aug. 2012.
Waldteufel, P., J. L. Vergely, and C. Cot, "A modified cardioid model for processing multiangular radiometric observations," IEEE Transactions on Geoscience and Remote Sensing, vol. 42, pp. 1059-1063, 2004.
Zhou, Y., R. H. Lang, E. P. Dinnat, and D. M. Le Vine, L-Band Model Function of the Dielectric
Constant of Seawater, IEEE TGRS, Vol. 55, No. 12, 2017.

Area D - Board 368

The Bay of Bengal (BoB) receives a lot of freshwater in association to the southwest monsoon, either directly from rain or indirectly through the Ganges-Brahmaputra and Irrawaddy rivers. This large freshwater input in a small, semi enclosed basin results in some of the lowest sea surface salinities (SSS) in the tropical band. The monsoon winds also drive an energetic circulation, which yields intense SSS gradients. Previous studies suggest that the strong haline stratification in the BoB inhibits the vertical mixing of heat and nutrients, with potential impacts on cyclone intensity, climatological rainfall or productivity. It is hence important to describe the salinity distribution and its variability accurately in the BoB.
Previous studies also suggested the development of a seasonal freshwater tongue (or “river in the sea”) along the east coast of India after the monsoon, due to southward advection of the Ganges Brahmaputra freshwater plume by the East Indian Coastal Current (EICC). They also indicated a strong SSS interannual variability in response to circulation anomalies associated with the Indian Ocean Dipole (IOD). Those studies were however either based on modelling or sparse in situ data. Early attempts to describe the BoB SSS variability using remote-sensing were hampered by the strong biases in coastal regions in early SMOS products.
In this talk, we will show that the latest SMOS/CATDS product and SMAP data compare well with in situ data in the BoB, including close to the coasts (with SMAP being better quality but SMOS offering a nine-years long record). Those two datasets confirm the seasonal development of the “river in the sea” after the southwest monsoon, along the east coast of India. The 9 years-long SMOS record indicates that the southward export of freshwater is modulated by the IOD, with positive IOD events yielding less southward export of freshwater, and hence negative SSS anomalies in the northern BoB and positive SSS anomalies in the southwestern BoB. They also indicate that meso-scale eddies play an important role in stirring the SSS field, and exporting freshwater away from the fresh coastal region.

Area D - Board 369

A main goal of the Climate Change Initiative Sea Surface Salinity (SSS) project is to provide climate users with calibrated SSS long time series, with qualified uncertainties, and at the highest possible spatial and temporal resolution in order to fully benefit from space measurements, relatively to what is accessible from in situ observations (e.g. Argo floats). This is undertaken by merging SSS from different space missions.
Three spatial missions are providing SSS observations since 2010: SMOS (Soil Moisture and Ocean Salinity mission), Aquarius and SMAP (Soil Moisture Active Passive mission). Following the approach used in the SMOS SSS processing at Centre Aval de Traitement des Données SMOS (Boutin et al. 2018), we propose a strategy to standardize the error estimates by using self-consistency criteria: by comparing SSS acquired closely in time under different observation geometries and, in the case of CCI+SSS, by different sensors, it is possible to extract random and systematic uncertainties. We find that the individual SSS random errors strongly depend on the Sea Surface Temperature and are at first order in very good agreement with the ones expected from the radiometric uncertainty of each instrument. They are used to estimate the final error over merged Level 4 SSS. The systematic uncertainty estimation allows correcting individual SSS before merging. Two main biases have been identified: a static bias which is associated to land contamination and Radio Frequency Interferences; a seasonal bias with a latitudinal dependency. The amplitude of these biases depends strongly on the passes (ascending or descending), on the earth position and on the acquisition geometry (beam number, cross-swath location, aft or fore acquisition relative to the sub-satellite point). This poster will present SSS level 4 products from SMOS ESA v662, SMAP RSS v3.0 and Aquarius v5 products, resulting from inter sensor relative calibration. Some comparisons with external in-situ SSS data, such as PIRATA moorings in the tropical Atlantic Ocean, will be presented. They illustrate the impact of adding Aquarius SSS (2011-2015) and then SMAP SSS (2015-) onto SMOS SSS (2010-) for long term studies over regions with very high natural SSS variability like the Amazone plume and over regions with much lower variability.

Boutin J., Vergely J.L., Marchand S., D'Amico F., Hasson A., Kolodziejczyk N., Reul N., Reverdin G., Vialard J. (2018). New SMOS Sea Surface Salinity with reduced systematic errors and improved variability. Remote Sensing of Environment, 214, 115-134. Publisher's official version : http://doi.org/10.1016/j.rse.2018.05.022 , Open Access version : http://archimer.ifremer.fr/doc/00441/55254/

Area D - Board 370

Two L-Band (1.4GHz) microwave radiometer missions, the Soil Moisture and Ocean Salinity (SMOS) and the Soil Moisture Active and Passive (SMAP) missions, currently provide salinity measurements in the first centimeter below the sea surface. At this depth, salinity variability at hourly temporal scales is dominated by the impact of precipitation. ARGO sea surface salinity measured at about 5 m depth, where the signature of rain is really less important. Thus, ARGO SSS used for SSS space-based estimates validation may be strongly different of SMAP/SMOS measurements due to rain effect (Boutin et al, 2013). For this reason, a better understanding of rain imprint on SSS is a key aspect of L-Band SSS validation. Furthermore, the strong impact of rain on SSS satellite estimations motivate to use L-Band SSS to complement rainfall information from rain satellites (as from Global Precipitation Measurement (GPM) constellation).

To these ends, a small wave rider called Surpact (Reverdin et al, 2013) was developed for air-sea investigations. It was designed to attach to a drifter or a mooring and to float upon the surface waves in order to measure sea state and rain rate using acoustic sensor as well as temperature and salinity at a small fixed depth from the surface. During SPURS-2 field campaign, in 2017, the Surpact was deployed three times in two different modes: with high and low temporal resolution. These deployments have allowed to measure a set of rain events in different wind speed conditions. The shorter of these 3 deployments has allowed to observe a very high freshening of 9 pss. In summer 2018, the Surpact was deployed by Mirai vessel (JAMSTEC) in West Pacific. In addition to several low wind rain events, very high wind speed conditions and slight freshening was encountered during Trami and Kong-Rey typhoon.

The purpose of this poster is to present combined measurements of salinity, temperature, acoustic intensity and wave spectrum obtained with the Surpact during the field campaigns cited above. Relationship between freshening and rainfall derived from Surpact are compared with relationship derived from satellites (Supply et al, 2018) and case study with collocations between Surpact, rain satellites and SSS satellites are illustrated.



Boutin, J., Martin, N., Reverdin, G., Yin, X., & Gaillard, F. 2013. Sea surface freshening inferred from SMOS and Argo salinity: Impact of rain. Ocean Science, 9, 183–192. https://doi.org/10.5194/os-9-183-2013



Reverdin, G., S. Morisset, D. Bourras, N. Martin, A. Lourenço, J. Boutin, C. Caudoux, J. Font, and J. Salvador. 2013. Surpact: A SMOS surface wave rider for air-sea interaction. Oceanography 26(1):48–57, http://dx.doi.org/10.5670/oceanog.2013.04.



Supply A, Boutin J, Vergely J-L, et al. Precipitation Estimates from SMOS Sea-Surface Salinity. Q J R Meteorol Soc. 2018.144 (Suppl. 1):103–119. https://doi.org/10.1002/qj.3110

Area D - Board 371

Two L-Band (1.4GHz) microwave radiometer missions, the Soil Moisture and Ocean Salinity (SMOS) and the Soil Moisture Active and Passive (SMAP) missions, currently provide salinity measurements in the first centimeter below the sea surface. At this depth, salinity variability at hourly temporal scales is dominated by the impact of precipitation. The dependency of the salinity freshening with the instantaneous rain rate (RR) observed between 50°S and 50°N, with SMOS and SMAP salinities, is very similar.

We investigate the influence of rain history on salinity anomalies. By using rain rates retrieved from several microwave satellites measurements including Advanced Microwave Scanning Radiometer 2 (AMSR-2), and Special Sensor Microwave Imager Sounder 17 (SSMIS-17 and SSMIS-16) and by taking advantage of their different crossing times, we estimate the temporal cross-correlation function between salinity freshening and rain rate for different time lags in various tropical and high latitudes regions. Whatever the region, the magnitude of the salinity anomaly associated with precipitation is dominated by the instantaneous RR for each area. The apparent correlation between salinity anomaly and rain history can be explained by RR auto-correlation.

The relationship between salinity anomaly (∆S) and RR is then investigated in six regions, with RR provided using three different algorithms (the Unified Microwave Ocean Retrieval Algorithm (UMORA), the Goddard profiling algorithm (GPROF) and Integrated MultisatellitE Retrievals for GPM (IMERG)). Differences in RR distribution between the various algorithms lead to differences of up to a factor 2 in ∆S versus RR slopes. For a given RR product, we also observe that part of the variability in ∆S versus RR relationships is related to the variability in wind speed regimes as detected by SMAP wind speed.

These results are compared with Sea Surface Salinity (SSS), rain and wave measured in-situ during field campaign as SPURS-2 2017 cruise and Mirai 2018 cruise using collocation methods.

Area D - Board 372

Atmosphere ocean interactions give rise to physically distinct bodies of water that form at the ocean surface, these water masses are distinct not only in their temperature and salinity properties but also in their formation history. Walin's simplistic formulation defines water masses as being formed due to variations in heat and freshwater fluxes at the surface of the ocean which cause a shift of water in temperature, salinity and thus density space - a transformation of the water mass through different density classes. The rate of change of this transformation results in either an accumulation or a diminishing amount of water between two subsequent density surfaces - positive and negative formation respectively.

To track the spatial and temporal variability in water mass formation we have estimated formation rates in three ocean basins; the North Atlantic, North Pacific and Southern Ocean. Due to the strength of the Western Boundary Currents (WBCs) in the North Atlantic and North Pacific heat loss and deep convection during winter results in the homogenisation of a large portion of the water column to specific salinities and temperatures. Advection via the respective currents extension result in this homogeneous water column spreading over a large horizontal area. A similar water mass of homogeneous properties develops in the Southern Ocean as the Isotherms plunge rapidly towards the north across the Sub-Antarctic Front (SAF). The subsequent formation of a seasonal thermocline during the onset of spring acts as a ceiling to these water masses which are then recirculated within the subtropical gyres of the respective ocean basins. The re-emergence of these water masses during subsequent winters acts to alter winter-time climate as well as climatic oscillations.

We separate satellite derived data sets for heat and freshwater fluxes into bins of satellite derived sea surface salinity (SSS) and temperature (SST). This results in a satellite-derived estimate of water mass formation in thermohaline, density and geographic coordinates. The three coordinate systems give additional degrees of comparison between satellite derived results and literature. Also, the sensitivity of estimates of formation to perturbations in the underlying data sets of SSS and SST are explored using a Monte Carlo simulation. Computing statistics ultimately gives rise to an uncertainty in the geographic distribution of the water masses derived from satellites.

Density Flux - a key variable in calculating water mass formation rates - is particularly sensitive to perturbations in heat and freshwater fluxes, and although in the current study heat and freshwater Flux data sets have been kept constant, they’re inclusion as additional parameters in the Monte Carlo simulation would allow for a more complete assessment of the propagation of uncertainties in the underlying data sets to estimates of formation. We plan to include additional satellite data sets for SSS, SST, heat and freshwater fluxes which would allow for a comparison between various satellite sources and literature in order to pinpoint differences which arise through various satellite sources. Through the utilisation of parameters such as the Turner angle and the Brunt vaisala frequency we can understand the evolution of surface water masses into intermediate and deep water masses - a key aspect to accurately assessing climate variability.

Area D - Board 373

SMOS ocean salinity products are continuously improved at the Ocean Salinity Center of Expertise for Centre Aval de Traitement des Données SMOS (CATDS CEC-OS). Before being implemented in the near-real-time processing chain (CATDS-CPDC) and before being distributed to the users via the CATDS portal, they need to be systematically validated. This systematic validation consists of two main stages: (1) Comparison with ship transects and calculation of typical statistics; (2) Comparison with the In-Situ Analysis System (ISAS) monthly fields, in terms of both mean spatial maps and time series of key statistics parameters. The key statistics parameters are computed both over the global ocean and for individual zones of interest. Thus, both the mean spatial patterns and temporal variability in various regions are evaluated.
This poster focuses on the evaluation of the most recently released CEC-OS products : Debiased SMOS SSS L3 v3, available at the CATDS (http://www.catds.fr/Products/Available-products-from-CEC-OS/L3-Debiased-Locean-v3, http://doi.org/10.17882/52804#57467) and comparison with previous version. The L3 products are available for 9-day and 18-day Gaussian averaging. Both v2 and v3 products contain a bias correction mitigating SMOS systematic errors in the vicinity of continents and seasonally-varying latitudinal systematic errors (Boutin et al. 2018), based on internal consistency of SMOS SSS retrieved in various locations across swath. The v3 product additionally considers seasonal variation of salinity natural variability that brings critical improvement in strongly varying areas, like river plumes, as well as an additional bias correction improvement at high latitudes.
The standard deviation of the difference between the 18-day v3 product and the ship measurements varies between 0.65 pss for small distances from the coast (Dc < 100 km) and 0.20 pss at Dc > 1000 km. The v3 product thus offers a considerable improvement close to the coast, compared to the previous, already bias-corrected, v2 version (with 0.82 pss at Dc < 100 km and 0.21 at Dc > 1000 km). At high latitudes (50N to 80N), the new v3 product also performs better: the mean bias decreases from 1.0 to 0.8 pss at Dc < 100 km and from -0.5 to -0.2 pss between the v3 and v2 product, as compared to the ship measurements.

Area D - Board 374

The CATDS (Centre Aval de Traitement des Données SMOS) is the French ground-segment facility in charge of the generation, calibration, archiving and dissemination of the SMOS level 3 and level 4 science products. It processes Sea Surface Salinity (named Ocean Salinity, OS) and Soil Moisture (SM). CATDS is also providing services to users and scientists.

More specifically, it is in charge of:
- Processing of the SMOS level 3 and 4 science data from level 1B data received from the ESA Data Processing Ground Segment (DPGS),
- Re-processing of the SMOS level 3 and 4 data,
- Catalogue and archive of the level 3 and 4 products, including the auxiliary data, the calibration data, and the level 1B data used for the processing,
- Calibration and validation of the SMOS L3 and L4 products,
- Dissemination of the SMOS level 3 and level 4 products to the users,
- Assistance and support to SMOS L3 and L4 users,
- Potential feedback towards DPGS (new algorithms, calibration functions, and software).

CATDS involves two main components:
- A production centre (called C-PDC) which routinely produces and disseminates L3 and L4 data.
- 2 expertise centres (called C-EC) which host the definition of algorithms, assess the quality of the products and provide specific information to users.

The C-PDC processing chain which generates SM and OS L3 data from L1B products is composed of several processors. Some of them are based on the ESA DPGS’s prototypes. The main difference with level 2 processes is that DPGS processing is based on half an orbit, whereas C-PDC processing is based on several days of data.
One of the goal of the C-PDC L3 processors is to select correct input data and to reject the dubious ones. These L3 processors perform temporal analysis, time aggregations and spatial and temporal averages.

The SSS L3 products generated and distributed are:
- Daily maps
- 10 day and monthly averaged maps
- Debiased daily maps
- 10 day and monthly averaged debiased maps
- 9-day averaged debiased maps (simple average – work in progress)
- 9-day averaged debiased maps (gaussian average – work in progress)
In debiased products, the SSS is corrected from land-sea contamination and latitudinal biases. Each product is referenced with a DOI.

A web site (www.catds.fr) presents the products and gives information related to the operational status. It includes an online catalogue which lists, describes and gives access to the products. The website also includes a tool (maps.catds.fr) to search, browse and visualize the main fields of the generated products.

Users can access the products either through FTP or through the Sipad, an interactive Web based tool which allows aggregations and sub-settings.

A single point of contact (support@catds.fr) operated by the production center allows users:
- To obtain support either technical or scientific on the products and their usage
- To give feedback about the products

We present here the CATDS production centre with a focus on SSS processing chains and on the exploitation status. We also describe the L3 data dissemination process and the way to retrieve CATDS L3 products.

Area D - Board 375

The thermodynamics (air-sea fluxes, ...) and dynamics (currents, mixing, ...) of the ocean surface mixed layer are intimately linked with the horizontal density (buoyancy) and associated density compensated temperature/salinity (so called spiciness) structure of the ocean surface. These parameters are derived from both temperature and salinity measurements through the Ocean Thermodynamic Equation of State (TEOS-10, 2010). The Sea Surface Temperature (SST) and Sea Surface Salinity (SSS) contrasts have been shown to vary as a function of location, season, at all time and space scales. Thus, both consistent SST and SSS measurements are needed to monitor density and spiciness variability.
Before Argo array, the sparse scientific cruises and Thermosalinograph (TSG) transects from both scientific vessels and ship of opportunity were the main platforms providing consistently both SST and SSS measurements. Since the early 2000’s, the Argo array of autonomous profilers floats provides temperature and salinity profiles every 10 days in 3°x3° box. Whereas, this design is able to resolve monthly SSS and SST at large scale (>300 km), the ocean surface synoptics scales are much smaller than the main in situ capabilities.
Although, consistent SST daily maps from satellite at increasing resolution are available since late 70’s (up to O(1 km)), no comparable SSS measurements existed before 2009. The Soil Moisture and Ocean Salinity (SMOS) mission (European Space Agency) launched in November 2009 provides measurements of Sea Surface Salinity (SSS) over the global ocean at 45 km resolution every 3 to 5 days. Thus, the SMOS SSS products can be used to monitor the distribution of mesoscale features of SSS (larger than 50-100km, with 10 day to monthly averages). In addition, since 2015, Soil Moisture Active and Passive (SMAP/NASA) completed the SSS satellites constellation with a similar spatial resolution and 2-3 days cycle.
However, both SMOS and SMAP SSS still experiences large biases and noise on various time and space scales, related in particular to instrument calibration, to image reconstruction, and to radiofrequency interferences. Using a SMOS, SMAP SSS and the global Argo array of in situ measurements as reference data base, a statistical approach and Optimal Interpolation are used to characterize and to correct a large part of the biases and reduce the noise. Finally, a 25-km and weekly Optimal Interpolated (OI) mapping provides improved SMOS SSS and SMOS/SMAP SSS L4 global fields. Combined with SST products of comparable spatio-temporal resolution, remote sensed L4 density and spiciness products are derived and provided for the first time. We present a validation of this new products based on comparisons with SSS derived from TSG measurements from ships of opportunity.

Area D - Board 376

The interannual variability of the South Indian Ocean sea surface salinity maximum (SISSS-max) is studied using a combination of satellite and in situ data. Properties of the feature examined include the size, strength and location, the forcing via wind stress and surface freshwater flux, subduction, and changes in the underlying subtropical circulation.

The SISSS-max is more elongated than other ocean basins, and centered 10° poleward of the other southern hemisphere SSS-max features. Its center lies near (30°S, 85°E). It sits in a zone of stagnant convergent flow at the surface. Unlike other ocean basins, water entering the SSS-max at the surface due to Ekman transport comes mainly from the poleward side of the feature. The SISSS-max salinity has a weak seasonal cycle, less than 0.12 in amplitude, with maximum SSS in summer, February-April.

Deeper, the thermocline is nearly flat, and the flow is very weak and equatorward at 150 m. Water subducted into the subsurface salinity maximum spreads slowly to the northeast. Unlike the South Pacific SSS-max, the entire South Indian Ocean basin south of 20°S is dominated by net evaporation.

Results indicate that the SISSS-max has been getting larger (by a factor of 2), saltier and has been moving poleward (by about 1.5°) and eastward (by about 5°) since the start of satellite observations of SSS. Surface currents, including Ekman transport driven by wind stress, are very weak suggesting that changes observed in the SISSS-max are a not result of changing winds or surface circulation.

Area D - Board 377

The Algerian Basin is a key component of the general circulation in the Western Mediterranean Sea. The presence of both fresh Atlantic water and more saline Mediterranean water gives the basin an intense inflow/outflow regime and complex circulation patterns. Energetic mesoscale structures that evolve from meanders of the Algerian Current into isolated cyclonic and anticyclonic eddies dominate the area, with marked repercussions on biological activity. In recent years, several studies have demonstrated the advantages of combined use of autonomous underwater vehicles, such as gliders, with remotely sensed products. In this framework, an important contribution could come from a new generation of enhanced sea surface salinity (SSS) products, i.e. Soil Moisture and Ocean Salinity (SMOS), Soli Mosture Active Passive (SMAP) and Aquarius.
Here we present a preliminary study to assess the advantages of using the enhanced SMOS SSS products obtained through a new methodology using a combination of debiased non-Bayesian retrieval, DINEOF (Data Interpolating Empirical Orthogonal Functions) and multifractal fusion with high resolution Sea Surface Temperature (OSTIA SST) maps for improving their spatial and temporal resolution in the Western Mediterranean Sea (Olmedo et al., 2018). The Algerian Basin SSS described by the new SMOS maps have been compared with the in situ high resolution glider measurements collected in the framework of the ABACUS series observational program conducted in the Algerian Basin during fall 2014 – 2017 and spring 2017 – 2018 (Cotroneo et al., 2016; Aulicino et al., 2018).
The achieved results show that i) the enhanced SMOS products well represent the salinity patterns described by in situ SSS; ii) the mesoscale structures described by SMOS in the Algerian Basin coincide with those identified by gliders and described by altimetry and MODIS images.

Area D - Board 378

Monitoring sea-surface salinity (SSS) is crucial to investigate the water cycle, the ocean dynamics and the marine biogeochemistry from the global to the regional scale. In the past, SSS has suffered from poor observational coverage, being uniquely sampled via pointwise in-situ observations, also impacting the accurate assessment of its surface patterns. In order to overcome the sparseness of the SSS observations, several efforts have been made. In the framework of the Copernicus Marine Environment Monitoring Service (CMEMS) a daily, 1/4° nominal resolution, mesoscale resolving SSS multi-year product (L4) was developed by CNR (National Research Council of Italy). Such product relies on a multi-dimensional (multivariate) optimal interpolation (OI) algorithm combining SMOS salinity retrievals and in-situ salinity measurements with satellite-derived sea surface temperature (SST). This product will be distributed by CMEMS operationally in near real time starting from fall 2018. However, there are no planned missions to secure continuity in the satellite SSS measurements following SMOS or SMAP missions, which could potentially reduce both the accuracy and the effective resolution of this product in the future. The Copernicus Imaging Microwave Radiometer (CIMR) satellite mission is a high priority task within the European Copernicus Expansion programme. At present, the mission is in preparatory phase and, if selected, the expected launch year will be 2025. CIMR is supposed to provide global mesoscale-to-submesoscale resolving observations of sea-surface temperature, sea-surface salinity and sea-ice concentration.
CNR is contributing to the CIMR preparatory phase evaluating the impact of the CIMR SSS observations within the CMEMS global L4 SSS processing chain. Our study takes advantage of an Observing System Simulation Experiment based on the outputs of the CMEMS-MERCATOR global operational model. A one year long time series of synthetic in-situ and satellite CIMR SSS observations was extracted from the surface layer of the MERCATOR numerical simulations (year 2016). The L4 SSS processing chain was adapted to ingest the new synthetic SSS observations and we could successfully test the OI algorithm via combination with the MERCATOR-derived L4 global SST. The outputs of the processing chain, i.e. the OI L4 SSS, will be compared with the original MERCATOR model outputs. We aim at demonstrating that the combined use of the in-situ and the CIMR synthetic observations will improve the derivation of the global L4 SSS compared to a processing where only in-situ observations are ingested.

Area D - Board 379

The ESA Soil Moisture and Ocean Salinity (SMOS) mission provides two-dimensional interferometric radiometer measurements of L-band (1.4 GHz) brightness temperature from a satellite in polar orbit. At this frequency, the atmosphere is almost transparent and surface emission is strongly related to soil moisture over continental surfaces, salinity and the state of the ocean surface, and to the thickness of sea ice.
At the European Centre for Medium-Range Weather Forecasts (ECMWF) a data monitoring system for the SMOS near real time products was developed to provide quality checks and feedback to ESA. The brightness temperature measured by SMOS is compared to the simulated one generated using the CMEM (Community Microwave Emission Modelling Platform) radiative transfer model developed at ECMWF [de Rosnay et al., 2009] in collaboration with other scientists. With the current configuration, CMEM does not take the sea state (waves, including whitecapping and foam streaks) into account over the ocean.
The RTTOV (Radiative Transfer for TIROS Operational Vertical sounder) radiative transfer model has an ocean emissivity model component called FASTEM (FAST microwave Emissivity Model), which was developed by the Met Office (U.K.) and was extended to include L-band measurements [Liu et al., 2011]. FASTEM has the capability to model the effect of ocean surface roughness and foam on the L-band emissivity.
In this study RTTOV/FASTEM is used to simulate L-band brightness temperature over the ocean starting from ECMWF background model fields. The simulated brightness temperature is then transferred to the SMOS antenna reference frame and compared against SMOS observations. The results are analysed taking into account different aspects of the observations such as the different polarisations, different viewing angles, different geographical areas, RFI contamination. Results from RTTOV/FASTEM are also compared to the CMEM ones to evaluate the impact of the surface roughness modelling capability on the simulated brightness temperature.

Area D - Board 380

Mode waters (MWs) play a major role in the modulation of the Earth climate as they transport a large volume of heat, CO2 and other properties across basins at seasonal to longer time-scales. In the context of anthropogenic global warming, unlocking the understanding of the MWs is critical as climate prediction models from CMIP5 still struggle in their representation. MWs are formed in the deep winter mixed layers, where they acquire their homogeneous properties like salinity, temperature, CO2, heat, nutrients as well as low potential vorticity. They are then subducted in the pycnocline in conjunction with the strong tilted density polar fronts and the separation of the western boundary current. MWs subsequently flow and export the characteristics acquired at the surface at subsurface or intermediate depth into the subtropical gyre and the equatorial region.

MWs in the South Pacific Ocean are of particular interest because of their likely 2-way interaction with the El Niño Southern Oscillation (ENSO). ENSO is the first mode of variability at the interannual time scale with global societal implications.
First, the South Pacific Ocean, where MWs are formed, shows strong interannual variability in sea surface salinity (SSS). The Western Pacific Ocean fresh and warm pool as well as the South Pacific Convergence Zone are very well correlated with ENSO leading to large anomalies that are advected south by Ekman currents into the MWs formation sites. Variability of SSS have implication for the volume and anomalies exported as MWs.
Second, these salinity and volume anomalies exported by MWs could in turn lead to the alteration of the waters entering the western Pacific equatorial region which would affect ENSO events. Studies indeed show that equatorial Pacific subsurface temperature and salinity can modulate ENSO through vertical salinity stratification (formation of barrier layers).

9 years of SSS observation from the SMOS (Soil Moisture Ocean Salinity) mission provides an unprecedented dataset to study the interannual variability of SSS in the MWs formation sites of the South Pacific Ocean. MWs pathways are furthermore identified using modelling and observations in order to estimate the characteristics of the ENSO-related interannual anomalies reaching the equatorial region of the Western Pacific Ocean.

Area D - Board 381

In the framework of the ESA Climate Change Initiative for Sea Surface Salinity and using the ESA SMOS Pilot Mission Exploitation Platform (Pi-MEP), we will assess different passive L-band microwave satellite-based salinity products in the North Atlantic over the full period (2010-2019). In particularly we will assess the seasonal and interannual variability of the satellite products and compare them with available in situ (Argo, TSG, hydrographic section) and model data. We will pay particular attention to the signature in SSS around the subpolar gyre with a focus on capturing the 2013-2017 subpolar gyre cold anomaly. This is one region where extracting reliable satellite-based L3 SSS has so far been very difficult. We aim to highlight the potential strength and weakness of the different SSS products of the different satellites (SMOS, Aquarius, SMAP) and the different processing chain products. We will especially assess the products dedicated to the high latitudes.
Our focus on the subpolar gyre is due to a record-breaking 0-700m cold anomaly (with a strong surface expression) beginning in winter 2013-14 and continuing through to 2017 (Josey et al., 2017). This striking feature has the potential to be an early warning signal of climate-change induced AMOC collapse (Rahmstorf et al, 2016). Heat budget analyses indicate that it is most likely a response to extreme winter surface heat loss (Duchez et al., 2016). However, the debate remains ongoing with changes in ocean circulation being the other leading factor.
Preliminary analysis of in situ observation based ocean salinity fields indicate that the cold anomaly is accompanied by ocean freshening. However, the sparse nature of the observations severely limits our ability to accurately determine the temporal and spatial variability of surface salinity. Combination of assessed space based SSS with ocean surface currents products (GlobCurrent, OSCAR) and in situ measurements will help to understand the processes responsible for generating the salinity.

Area D - Board 382

The Copernicus Imaging Microwave Radiometer (CIMR) is a high priority candidate mission for Copernicus expansion designed to responds directly to the integrated European Union Arctic policy. The mission will provide products to serve communities addressing climate change and safeguarding the Arctic, environment sustainable development in and around the Arctic, international cooperation on Arctic issues and operational sea ice services and global sea surface temperature capability.

Surface characterization from satellites is required to understand, monitor and predict the general circulation of the ocean and atmosphere, the interaction between the surface and the atmosphere, as well as the water and energy cycles. With more than 70% cloud coverage at any time, visible and infrared satellite observations only provide limited information. Regardless of the cloud cover, passive microwaves offer ocean and sea ice information such as Sea Surface Temperature (SST), Sea Surface Salinity (SSS), Ocean Wind Speed (OWS) and High Ocean Wind Speed (HOWS), Sea Ice Concentration (SIC) and thin Sea Ice Thickness (SIT). These products are particularly important for the Arctic environment. Up to now all these oceanic/sea ice parameters estimated from passive microwave observations are limited in spatial resolution and/or accuracy.

CIMR will observe at 1.4, 6.9, 10.65 18.7 and 36.5 GHz (L, C, X , Ku, and Ka bands) at 55° incidence angle, in a polar orbit with low-noise polarized receivers, and a foldable mesh antenna of ~7 m diameter.

Thanks to simultaneous measurements between 1.4 and 36.5 GHz, CIMR combines the merits of the AMSR and the SMOS/SMAP/Aquarius type missions. It will provide global, simultaneous and consistent retrievals of SST, SSS, OWS, HOWS, SIC and SIT, twice daily and regardless of the cloud cover. CIMR will estimate SSS with an instantaneous theoretical precision of better than 0.3 psu, and a spatial resolution of 55 km, and the SST with a precision of 0.2 K and a spatial resolution of 15 km, with coincident measurements of OWS, HOWS, SIC, and thin SIT.

Twice daily accurate instantaneous estimate of SSS and SST will lead to an improved monitoring of the sea surface density. The coincident measurements of SSS, SST, and OWS with sea ice parameters will improve the studies of polar regions. The consistency between high resolution SST and SIC measurements will also be a significant step forward for all applications addressing ocean/sea ice interactions in the marginal ice zone, with rigorous collocation and identical spatial/temporal sampling of the products, as compared to a multi-platform approach.

With no guarantee of successors to AMSR2, SMOS and SMAP, the CIMR mission will provide continuity in the observations of low frequency microwave measurements, with improved spatial and radiometric characteristics, for an all-weather observations capability of key ocean and sea-ice surface parameters.

Area D - Board 383