Retrieval of Precipitation Profiles from Multiresolution, Multifrequency Active and Passive Microwave Observations

Abstract

In this study, a technique for estimating vertical profiles of precipitation from multifrequency, multiresolution active and passive microwave observations is investigated. The technique is applicable to the Tropical Rainfall Measuring Mission (TRMM) observations, and it is based on models that simulate high-resolution brightness temperatures as functions of observed reflectivity profiles and a parameter related to the raindrop size distribution. The modeled high-resolution brightness temperatures are used to determine normalized brightness temperature polarizations at the microwave radiometer resolution. An optimal estimation procedure is employed to minimize the differences between the simulated and observed normalized polarizations by adjusting the drop size distribution parameter. The impact of other unknowns that are not independent variables in the optimal estimation, but affect the retrievals, is minimized through statistical parameterizations derived from cloud model simulations. The retrieval technique is investigated using TRMM observations collected during the Kwajalein Experiment (KWAJEX). These observations cover an area extending from 5° to 12°N latitude and from 166° to 172°E longitude from July to September 1999 and are coincident with various ground-based observations, facilitating a detailed analysis of the retrieved precipitation. Using the method developed in this study, precipitation estimates consistent with both the passive and active TRMM observations are obtained. Various parameters characterizing these estimates, that is, the rain rate, precipitation water content, drop size distribution intercept, and the mass- weighted mean drop diameter, are in good qualitative agreement with independent experimental and theoretical estimates. Combined rain estimates are, in general, higher than the official TRMM precipitation radar (PR)-only estimates for the area and the period considered in the study. Ground-based precipitation estimates, derived from an analysis of rain gauge and ground radar data, are in better agreement with the combined estimates than with the TRMM PR-only estimates, which suggests that information useful for improving the radar-only estimates is contained in the brightness temperature data.