Use of a Maximum Entropy Method as a Regularization Technique during the Retrieval of Trace Gas Profiles from Limb Sounding Measurements

Abstract

Abstract The retrieval of trace gas profiles from radiance measurements of limb sounding instruments represents an inverse problem: vertical profiles of mixing ratios have to be extracted from sequences of horizontally measured radiances recorded by a spectrometer. Typically, these retrievals are plagued by random noise and systematic errors, necessitating the use of regularization techniques during inversion calculations. In the following, the use of selected maximum entropy operators as a regularization tool is discussed and their performance with conventional optimal estimation and Tikhonov-type regularization techniques is compared. The main gain with the proposed maximum entropy operators is that no a priori knowledge is needed; a reasonable initial guess profile is fully sufficient. The approach is verified by using simulated data of the Michelson Interferometer for Passive Atmospheric Sounding (MIPAS) instrument, an infrared Fourier transform spectrometer flown on the European Envisat mission.