The Tropical Rainfall Measuring Mission (TRMM) Satellite carries two instruments that directly sense precipitation: the TRMM Microwave Imager (TMI) and the Precipitation Radar (PR). This analysis compares the precipitation products created from instantaneous TMI and PR data. The motivation is understanding and improvement of the standard TRMM precipitation products. In standard processing, the TMI rain algorithm converts radiances in 9 channels to rainfall rate by matching them to expected radiances computed from cloud models. The PR rain algorithm converts reflected radiation to rainfall rate using reflectivity-rainfall (ZR) relations. Both types of data are then averaged in space and time into gridded monthly mean rainfall products. Since the TMI swath is three times wider than the PR swath, the monthly means of TMI and PR are derived from mostly different data samples. One way to compare TMI to PR is to use the gridded monthly mean products and assume that both samples are representative of the precipitation that occurred during the month. Another way, used in this work, is to remap the instantaneous TMI and PR data into common coordinate systems to compare measurements from the same precipitation sample.

Joint distributions of TMI and PR rainfall data show that several factors influence the relationship between TMI rainfall rate and PR rainfall rate. Surface type (ocean, land, or coast) affects the background against which the rain signature appears in the TMI and is used by the TMI rain algorithm to choose which database of cloud models to use. Surface type markedly affects the relationship. The PR rain algorithm uses rain type (stratiform, convective, or other) to choose which ZR relationship to use. For rain that PR classifies as stratiform, the mean TMI rainfall rate is higher than the mean PR rainfall rate. In contrast, for rain that PR classifies as convective, the mean TMI rainfall rate is lower than the mean PR rainfall rate. Location, rain intensity, and storm height also affect the relationship. Finally, since the TMI measures radiance in 9 channels with IFOV widths ranging from 4.5 km to 60 km while the PR measurement measures reflected radiation in a 4.4 km x 4.4 km IFOV, the size of the box used for comparison is important. A larger box size yields a higher correlation for rainfall rates above 1 mm/hr.