Inverse radiation problem to retrieve hydrometeors from satellite microwave radiances

This study presents a new hybrid method that combines regression analysis with genetic algorithms for the retrieval of hydrometeors (cloud liquid water, ice and rain) in the atmosphere, from satellite microwave radiances. A three layered atmosphere model (divided into 30 sub-layers) is used to generate simulated profiles of hydrometeors. The equation governing the transfer of radiation is solved using the finite volume method to obtain radiances (brightness temperatures) in the microwave region. This is known as the forward problem and is solved repeatedly to create a database with which regression equations are developed for the monochromatic microwave radiances, for six typical frequencies ranging from 6.6 to 85 GHz. The regression is done using nonlinear parameter estimation techniques. The inverse problem of retrieving the hydrometeors characteristics from microwave radiances is accomplished by posing the parameter estimation problem as an optimization problem, wherein, minimization of the sum of squares of residuals between the estimated and known radiances, for the above mentioned six typical frequencies, is done. In this study, genetic algorithms have been used for solving the minimization problem.