The Effect of Climate Regime on Satellite Estimates of Precipitation

Current long-term satellite rainfall data sets suffer from a variety of problems due to sampling limitations and other biases, which limits their utility for climate applications. Investigations over the east Pacific have shown substantial differences in climatological satellite rainfall estimates between passive microwave and infrared retrieval techniques. Preliminary comparisons of vertical profile retrievals from the Tropical Rainfall Measuring Mission (TRMM) also indicate that this issue is an important one for producing accurate monthly rainfall estimates over the global tropics for. The lack of in-situ observations over many ocean regions combined with significant seasonal and interannual changes in the type of precipitating systems makes this a difficult problem to address.

We have attempted to investigate the effect of differences in precipitation systems over a number of climate regimes using a combination of satellite and in-situ data sets. These include infrared observations from geostationary satellites, passive microwave estimates from SSM/I and TMI on polar-orbiting platforms, vertical profile retrievals from the TRMM precipitation radar and 3D radar scans taken during various field programs. Comparisons over the tropical east Pacific indicate that a significant portion of the total precipitation is produced by warm rain systems containing little or no ice layer. Although it is not as intense as deep convective systems, the lower cloud tops associated with these warm systems cannot be distinguished from nonprecipitating clouds using satellite infrared-based precipitation estimation techniques. As a result, variations in the amount of precipitation produced by warm versus cold (convective) systems can result in substantial biases in monthly mean satellite retrievals. This can have significant implications for the use of long-term satellite derived precipitation data sets in climate studies.