A Self-Consistency Approach to Improve Microwave Rainfall Rate Estimation from Space

Abstract

A multichannel statistical approach is used to retrieve rainfall rates from brightness temperatures (TB's) observed by passive microwave radiometers flown on a high-altitude NASA aircraft. Brightness temperature statistics are based upon data generated by a cloud radiative model. This model simulates variabilities in the underlying geophysical parameters of interest, and computes their associated TB's in each of the available channels. By further imposing the requirement that the observed TB's agree with the TB values corresponding to the retrieved parameters through the cloud radiative transfer model, the results can be made to agree quite well with coincident radar-derived rainfall rates. Some information regarding the cloud vertical structure is also obtained by such an added requirement. The applicability of this technique to satellite retrievals is also investigated. Data which might be observed by satellite-borne radiometers, including the effects of nonuniformly filled footprints, are simulated by the cloud radiative model for this purpose. Results from statistics generated using different hydrometeor vertical profiles in the cloud radiative model are examined. It is found that errors in the retrieved rainfall rates, and retrieval biases, decrease with increasing agreement between simulated TB's and those corresponding to the retrieved geophysical parameters.