Low-dimensional parameterization of the Kummerow´s microwave radiative transfer model for the atmosphere with precipitation

Proposes a method of radiative transfer model inversion that can be used in rainfall profile retrieval algorithms. The method extracts a minimal number of components of the hydrometeor profile to explain a maximum of measurable variations in observed brightness temperatures. The inverse operators to retrieve these informative profile components are stable and can be calculated and stored in the computer memory. Since the initial model is strongly nonlinear, the structure of the inverse operator depends on the initial guess of the hydrometeor profile. This involves a piecewise linearization of the initial model and creation of a database of inverse operators for a number of quasi-linear models