Algorithm for Extraction of Surface Emissivity in the Context of VIRTIS on Venus Express

Abstract

Venus nightside multispectral images aquired by VIRTIS contain information of surface emissivity at wavelengths close to 1 micron. This information is relayed by thermal emissions of the surface escaping to space through the NIR spectral ‘windows' in the atmosphere. On its way through the atmosphere the thermal radiation is modified by scattering and absorption by clouds. Variations in the optical thickness of the clouds modulate the spatial distribution of upwelling radiation. Multiple reflections between surface and clouds generally wash out image contrast from surface emissivity. We present an algorithm to extract surface emissivities by separating atmospheric influences from the images inside the spectral windows at 1.02, 1.10 and 1.18 micron. The necessary processing steps detailed are: 1) Removal of scattered sunlight 2) Binning of appropriate images inside the window spectral ranges 3) Correction of limb darkening 4) Removal of contrast due to attenuation by clouds 5) Removal of contrast due to surface temperature 6) Correction for multiple cloud-surface-reflections For a first quick guess on general trends of emissivity with respect to geological circumstances this algorithm is applied with several simplifications and ad-hoc assumptions. This simple model assumes the atmosphere of Venus to be one horizontally homogenous layer with spatial variation of transmittance allowing for direct inversion of the data. For an improved estimation of surface emissivity tabulated results from previous forward modeling of radiative transfer are used for steps 2) to 6). With this approach the accuracy of a detailed modeling of the atmosphere of Venus is combined with the speed and traceability of the step-by-step inversion of radiance data using the simplified model.