Cloud top pressure retrieval from measurements within the O2 -A band with the satellite imaging sensor MERIS

In the near future, global coverage space-borne measurements within the absorption band of oxygen at 761 nm (O₂-A band) with an imaging spectrometer of medium spatial resolution will become available through ESA´s Medium Resolution Imaging Spectrometer (MERIS) on ENVISAT. In a cloudy atmosphere, the radiance ratio between an absorbing channel and a window channel is directly related to the averaged photon path length which is mainly determined by the cloud top pressure (ctp). Artificial neural network training has been used to invert the results of radiative transfer simulations. This paper presents the principles of the retrieval scheme, an error and sensitivity analysis to specify the influence of viewing geometry, surface reflection, cloud optical thickness and cloud type on the performance of the algorithm