Direct Validation Analysis of TRMM Precipitation Retrievals Based on TRMM Field Experiment Measurements

We describe several intercomparisons between Level 2 TRMM precipitation estimates and validation datasets from the TRMM-LBA and TRMM-KWAJEX field programs, and various independent validation data sources. Other intercomparison studies already conducted and involving relatively high quality ground validation measurements are showing biases below 10%. These results testify to the basic soundness of the physical principles used to design the TRMM facility algorithms. Less is known about the precision errors and specific weaknesses in the three standard TRMM Level 2 algorithms which give rise to random errors. The standard Level 2 algorithms consist of a radiometer-only method (algorithm 2A-12), a radar-only method (algorithm 2A-25), and a combined radiometer-radar method (algorithm 2B-31). This study seeks to use the high quality ground validation measurements collected during the TRMM field experiments to help determine the cause of differences between the radiometer-only, radar-only, and combined Level 2 algorithms, and just as important, set realistic uncertainty limits on the TRMM algorithm products.