Synchronous atmospheric correction and SST retrieval by AATSR data

The water vapor is the main absorb substance of the atmosphere in thermal infrared bands, but the distribution of water vapor amount in the ground varied greatly, so one radio sounding couldn’t represent the water vapor distribution of different pixel in one image. AATSR (Advanced Along-Track Scanning Radiometer) sensor could provide radiance measurements of two channels at two observation angles. So it is possible to retrieve the atmosphere parameters and SST (Sea Surface temperature) synchronously. In order to make the system of equations determined, the two atmosphere parameters that could express whole atmosphere profiles must be confirmed. As the retrieval accuracy is requested, the surface reflected atmospheric downward radiance effect couldn’t be ignored. A climatologically data set of 1761 different radio soundings, the SATIGR database, have been used to develop a new atmospheric correction method, the two atmosphere parameters have been proved that could represent the transmittance, atmospheric upward radiance and downward radiance of different wave band and different view angle, they are the transmittance and atmospheric upward radiance of 11μm at the nadir view angle. Introducing the appropriate sea surface BRDF model, the emissivity and the reflected downward radiance can be calculated precisely. So the SST, wind speed, and two atmosphere parameters are retrieved by Broyden iteration method, when the initial values are found using LUT (look up table) method. Simulations showed :(1) the physics-based algorithm described in this paper could provide more accurate calculation of atmospheric effect. Compared with split window method, the retrieval error of SST is little. (2) The new downwelling atmospheric radiance correct method described in this paper could provide more precise calculation of atmospheric effects. The field test using ATSR-2 (Along-Track Scanning Radiometer- 2) DATA proved this new algorithm can improve the retrieval accuracy significantly