First Assessment of SMOS Data Over Open Ocean: Part II-Sea Surface Salinity

• Published in: IEEE Transactions on Geoscience and Remote Sensing
• Main Authors: Boutin, Jacqueline, Martin, Nicolas, Yin, Xiaobin, Font, Jordi, Reul, Nicolas, Spurgeon, Paul
• Format: Article in Journal/Newspaper
• Language: English
• Published: Ieee-inst Electrical Electronics Engineers Inc 2012
• Subjects: L-band; microwave radiometry; ocean salinity; sea surface; Sea ice
• Online Access: https://archimer.ifremer.fr/doc/00074/18557/16108.pdf; https://doi.org/10.1109/TGRS.2012.2184546; https://archimer.ifremer.fr/doc/00074/18557/

We validate Soil Moisture and Ocean Salinity (SMOS) sea surface salinity (SSS) retrieved during August 2010 from the European Space Agency SMOS processing. Biases appear close to land and ice and between ascending and descending orbits; they are linked to image reconstruction issues and instrument calibration and remain under study. We validate the SMOS SSS in conditions where these biases appear to be small. We compare SMOS and ARGO SSS over four regions far from land and ice using only ascending orbits. Four modelings of the impact of the wind on the sea surface emissivity have been tested. Results suggest that the L-band brightness temperature is not linearly related to the wind speed at high winds as expected in the presence of emissive foam, but that the foam effect is less than previously modeled. Given the large noise on individual SMOS measurements, a precision suitable for oceanographic studies can only be achieved after averaging SMOS SSS. Over selected regions and after mean bias removal, the precision on SSS retrieved from ascending orbits and averaged over 100 km $times$ 100 km and 10 days is between 0.3 and 0.5 pss far from land and sea ice borders. These results have been obtained with forward models not fitted to satellite L-band measurements, and image reconstruction and instrument calibration are expected to improve. Hence, we anticipate that deducing, from SMOS measurements, SSS maps at 200 km $times$ 200 km, 10 days resolution with an accuracy of 0.2 pss at a global scale is not out of reach.