Errors on the retrieved sea surface salinity from microwave radiometry due to inaccuracies in the ancillary data

Abstract

Up to now, no space mission has attempted to measure the Sea Surface Salinity (SSS) due to technological challenges. The European Space Agency's SMOS (Soil Moisture and Ocean Salinity) mission, to be launched in 2007, aims at the generation of global SSS maps with a spatial and temporal resolution adequate for climatic and ocean general circulation studies. The sensor aboard SMOS is an L-band interferometric radiometer with full polarimetric capability called MIRAS (Microwave Imaging Radiometer by Aperture Synthesis). At low frequencies the brightness temperature measured by a radiometer presents maximum sensitivity to surface salinity, and optimal conditions are found at L-band (1400 MHz-1427 MHz), which it is a band protected for passive measurements. Even though the sensitivity is maximum at this frequency, it is not very high: 0.5 K/psu for sea temperature of 20/spl deg/C, decreasing to 0.25 K/psu for 0/spl deg/C. Furthermore, other variables influence the brightness temperature, as the sea roughness (mainly caused by the wind speed), sea foam, and also sea surface temperature. When SMOS will be in orbit it will be necessary to have accurate measurements of the sea roughness and surface temperature as much coincident as possible -both in time and space- to SMOS acquisitions, because errors on these ancillary data produce large errors on the retrieved salinity. This paper presents a study of the effect on the retrieved salinity of inaccuracies in the wind speed and wave height values. It compares the retrieved salinities when using different sources of these ancillary data: satellite measurements and meteorological models. It also investigates the possibility of allowing the algorithm to retrieve an effective wind speed, in addition to the sea surface salinity.