Rainfall retrievals using data from the Tropical Rainfall Measuring Mission (TRMM)

The launch of the Tropical Rainfall Measuring Mission (TRMM) in November 1997 heralded a new era in rainfall measurement from space: Instruments such as the TRMM Microwave Imager (TMI) providing better spatial resolutions, while the Precipitation Radar (PR) supplying hitherto unavailable calibration/validation data over large areas of the tropics. This paper investigates the currently available statistical/empirical techniques used for the retrievals of rainfall estimates from the Special Sensor Microwave/Imager, and presents results from these algorithms applied to the TMI data. Data from coincident overpasses between the TMI and SSM/I sensors are used to generate rainfall estimates: The results suggest negligible bias between the sensors, despite differences in look-angles and field of views. Comparison is then made between both the TMI and SSM/I estimates and the rainfall retrievals from the PR: These results show extremely good correlations between the passive microwave algorithms and PR retrievals for over water estimates. Over land areas the correlations are less good, but still are very high with very low bias.

Comparison is then made between the passive microwave data and surface data to assess the errors between the satellite and surface observations. A technique is suggested where the diurnal information, obtainable from the TMI rainfall estimates, may be used to augment the rainfall estimates of the SSM/I to correct for biases in the SSM/I sampling habits.