چکیده (انگلیسی):

Sea Surface Salinity, SSS, retrieved from measurements of Soil Moisture and Ocean Salinity, SMOS, satellite mission using the European Space Agency version 5 processors displays systematic errors, in particular in the vicinity of large land masses and in the open ocean as a function of latitude and season. The main goal of this paper is to quantify the contribution of systematic errors close to and away from land by examining the statistics of SMOS minus interpolated fields of in situ SSS before and after removal of climatological mean fields. Using interpolated fields of in situ SSS as reference, we find that the precision of monthly SMOS SSS anomalies, SSSANO, relative to a SMOS SSS monthly climatology computed over four years between 45°N and 45°S is 30% better than the one of SSS when considering pixels further than 1000 km from the coast. This improvement reaches 40% when all pixels are included. At 1600 km from the coast, the standard deviation of the difference between SMOS SSS and in situ interpolated fields is 0.15. These very encouraging results show that it is possible to correct part of the systematic errors in SMOS SSS from the SMOS SSS themselves, pending improvements in SMOS reconstructed brightness temperatures resulting from improved calibration and image reconstruction processes. The largest SSS anomalies are observed in the tropical Indian Ocean as related to Indian Ocean Dipole and in the tropical Pacific Ocean where latitudinal migration of the SSS minimum in the South Pacific Convergence Zone is related to El Niño Southern Oscillation phases and where strong SSS anomalies are observed in the Intertropical Convergence Zone during positive phases of ENSO in early 2010 and in late 2014, early 2015.