COUPLED SURFACE-ATMOSPHERE REFLECTANCE (CSAR) MODEL .1. MODEL DESCRIPTION AND INVERSION ON SYNTHETIC DATA

Absorption and scattering processes in the atmosphere affect the transfer of solar radiation along its double path between the Sun, the Earth's surface, and the satellite sensor. These effects must be taken into account if reliable and accurate information on the surface must be retrieved from satellite remote sensing data. One approach consists in characterizing the state of the atmosphere from independent observations and correcting the data with the help of radiation transfer models. This approach requires a detailed and accurate description of the composition of the atmosphere (e.g., aerosol and water vapor profiles in the case of advanced very high resolution radiometer data) at the time of the satellite over-pass and requires significant computer resources. An alternative method is to attempt to simultaneously retrieve surface and atmospheric parameters by inverting a coupled surface-atmosphere model against remote sensing data. This study describes such a coupled model and the results of its inversion against synthetic data, using a nonlinear inversion technique. The results obtained are encouraging in that realistic directional reflectances at the top of the atmosphere can be produced, and the inversion of the model against these synthetic data is capable of estimating surface and atmospheric variables. The accuracy of the retrieval is studied as a function of the amount of noise added to the data. It is shown that some surface or atmospheric parameters are easier to retrieve than others with such a coupled model, and that although it appears to be difficult to accurately and reliably estimate the water vapor amount from channel 2, there is a definite possibility of retrieving the aerosol loading from simulated channel 1 data, if the type of aerosol can be assumed.