Enhancement of Temperature-humidity Retrieval Algorithms of Satellite MW Data Processing

Additional a priori and/or supplementary information on the state of the atmosphere potentially improves the efficiency of retrieval algorithms in comparison to the usage of mean climatic statistics of the atmosphere. Exploration and evaluation of the efficiency of the application of such additional information is the main point of this work. To achieve this goal, this work investigates the statistical approach expansion by including new types of a priori information about the temperaturehumidity state of the atmosphere. The developed methodology introduces an efficiency measure that allows the estimation of the efficiency of the following a priori information types for atmospheric profile retrieval from satellite MW data processing: 1) covariance matrices of the full vector temperature-humidity state of the atmosphere along a vertical profile, 2) covariance matrices of atmosphere variations along the atmospheres horizontal layer, and 3) physical limits of humidity profile variations. The results of atmospheric profile retrieval, based on the Levenberg-Marquardt algorithm, confirm the possibility of using this approach to restore atmospheric parameters in real time, i.e. for times shorter than the correlation intervals in the atmosphere.