Aerosol optical thickness retrieval from GOME spectra

The GOME instrument, on board the ERS-2 satellite, has been designed in order to collect radiation over the entire wavelength region from 240 to 790 nm, in which several atmospheric species and also aerosols and clouds can be observed. A prototypal processor for the aerosol optical thickness retrieval and aerosol classification starting from GOME data has been developed. This processor has been devised as a tool to be used for the development of an operational GOME data processing chain. The implemented retrieval algorithm is based on a spectral reflectance fitting procedure of the measured radiances by GOME instrument. The maximum likelihood principle has been used in order to define the objective function. The ranking is made choosing the minimum among the least squares residuals computed for different aerosol classes. For each pixel the output of processor gives the aerosol optical thickness, the aerosol classification, a relative retrieval residual and a flag that indicates if the pixel is cloudy. The results of some different GOME real data sets are shown.