A Method to Retrieve Rainfall Rate Over Land from Trmm Microwave Imager Observations

Over tropical land regions, rain rate maxima in mesoscale convective systems revealed by the Precipitation Radar (PR) flown on the Tropical Rainfall Measuring Mission (TRMM) satellite are found to correspond to thunderstorms, i.e., Cbs. These Cbs are reflected as minima in the 85 GHz brightness temperature, T85, observed by the TRMM Microwave Imager (TMI) radiometer. Because the magnitude of TMI observations do not discriminate satisfactorily convective and stratiform rain, we developed here a different TMI discrimination method. In this method, two types of Cbs, strong and weak, are inferred from the Laplacian of T85 at minima. Then, to retrieve rain rate, where T85 is less than 270 K, a weak (background) rain rate is deduced using T85 observations. Furthermore, over a circular area of 10 km radius centered at the location of each T85 minimum, an additional Cb component of rain rate is added to the background rain rate. This Cb component of rain rate is estimated with the help of (T19-T37) and T85 observations. Initially, our algorithm is calibrated with the PR rain rate measurements from 20 MCS rain events. After calibration, this method is applied to TMI data taken from several tropical land regions. With the help of the PR observations, we show that the spatial distribution and intensity of rain rate over land estimated from our algorithm are better than those given by the current TMI-Version-5 Algorithm. For this reason, our algorithm may be used to improve the current state of rain retrievals on land.