Retrieval of minor constituents in a cloudy atmosphere with remote sensing millimetre wave measurements

The presence of clouds in remote-sensing measurements is often the cause of information loss about the minor atmospheric constituent profiles that are the goal of the measurement. Millimetre- and submillimetre-wave sounding can be used to abate the problems of cloud contamination. A quantitative assessment of this possibility is provided with the systematic simulation of the retrieval of minor atmospheric constituents in the presence of ice cirrus clouds of different density and particle size. The simulations are made for the particular case of the MARSCHALS instrument which is a heterodyne limb-sounding spectrometer that measures the atmospheric thermal emission in the millimetre- and submillimetre-wave regions operating on board a stratospheric aircraft. Four types of measuring conditions are identified: clouds that do not modify the observation, clouds that are adequately described with an atmospheric continuum absorption model, clouds for which an error is introduced when the limited vertical resolution of the sounding does not resolve the sharp discontinuity of the cloud, and clouds which require the modelling of scattering effects for a correct retrieval. When a rigorous modelling of the cloud is performed, the retrieval of minor atmospheric constituent is marginally affected by the presence of the cloud. We find that in the worst case the cloud causes a loss, with respect to clear-sky conditions, of two degrees of freedom from about 90 and of 7% of information content relative to climatological information. Copyright (c) 2007 Royal Meteorological Society.