Inverse problems in stellar occultation

Abstract

Remote sensing plays a key role in monitoring the atmosphere, which becomes increasingly important for the Earth's changing environment. The stellar occultation technique is a novel method to monitor vertical distribution of ozone and other trace gases from the troposphere to the upper mesosphere. GOMOS (Global Ozone Monitoring by Occultation of Stars) flying on board the European Space Agency's ENVISAT satellite is the first operational stellar occultation instrument.

Satellite measurements are indirect, and therefore inversion methods play a key role in the retrieval of atmospheric parameters. This dissertation is dedicated to inversion methods for stellar occultation measurements and to the optimization of retrievals.

Optimization of retrievals by inclusion of a priori information about smoothness of atmospheric profiles is considered. Two methods are developed. One of them –"the target resolution method"– develops the classical Tikhonov regularization. The second method includes smoothness a priori information in the form of Bayesian optimal estimation. The methodology for creating a priori information about smoothness of atmospheric profiles is developed.

This dissertation considers optimal selection of measurements based on information theory, aiming at optimal design of future instruments, as well as at improved efficiency of the current data processing. Two optimization problems, both taking the information content of the measurements as a criterion, are defined and discussed. The selecting procedures were developed, compared with each other and existing methodologies and applied to selection of the most informative channels for GOMOS measurements in the UV-Visible wavelength range.

The dissertation presents a feasibility study for retrieval of temperature and density profiles from pointing measurements by stellar occultation instruments. This study introduces extra geophysical parameters that can be obtained from GOMOS measurements.

The inversion methods developed are applied to GOMOS measurements. However, the methods are formulated in a general form that allows their application beyond the GOMOS mission.