Recent improvements in a method for remotely sensing precipitation and latent heating distributions based upon satellite-borne, passive microwave radiometer observations are summarized. In applications to synthetic data, estimated rainfall rates at sensor footprint-scale (14 km) are subject to significant random errors, but these errors are substantially reduced by spatial averaging. After spatial-averaging, rain rate and latent heating profile estimates exhibit biases that arise from a lack of specificity in the information contained in the microwave radiance data.
The retrieval method is applied to observations from the Tropical Rainfall Measuring Mission Microwave Radiometer (TMI). Retrieved instantaneous precipitation and heating distributions show general self-consistency and delineate plausible storm structures in an application to TMI observations of a mesoscale convective system over the tropical North Atlantic. Well-known climatological distributions of rainfall are reproduced by global, monthly-mean TMI precipitation estimates from July 2000. Zonal-mean heating profiles in the Tropics from the same period exhibit a primary maximum of heating near 7 km altitude and a secondary peak near 3 km, while at higher latitudes in the Southern Hemisphere, a vertical structure with heating aloft and cooling at lower altitudes is derived.