Thursday, 15 January 2009: 12:00 PM
The Development of an Improved Satellite Rainfall Dataset for Climate Applications
Room 126BC (Phoenix Convention Center)
Existing long-term satellite rainfall datasets such as GPCP and CMAP rely on accurate rainfall estimates from the current constellation of passive microwave radiometers to calibrate or adjust more frequent infrared observations from geostationary satellites over oceans. As a result, these merged products are susceptible to biases in the microwave retrievals resulting from errors in the instrument calibrations, changes in the radiometer constellation over time, and regime-dependent errors in the retrieval algorithm. To deal with inconsistent calibrations between radiometers, a Level 1C brightness temperature dataset has been developed in which existing radiometer data is intercalibrated to a common standard and stored in a common HDF format. An intercalibration working group within the Precipitation Measurement Missions (PMM) science team is currently working to develop suitable calibration adjustments for this intercalibrated Level 1C dataset. To address issues related to biases in the retrieval algorithm arising from sensor differences and/or regime-dependent changes, the existing operational ocean rainfall retrieval algorithm for TRMM and AMSR-E has been completely revised. This major new release, referred to as GPROF 2008, employs an apriori database created from a combined TRMM radar and radiometer analysis of precipitation, eliminating the need for sensor specific empirically tuned coefficients. Initial results from a beta version of GPROF 2008 indicate significant improvements in the detection of light rain, consistency between retrievals from sensors with different characteristics (i.e. TMI and SSM/I), and better agreement with estimates from the TRMM radar. While the intercalibration efforts as well as the implementation and testing of GPROF 2008 for the current radiometer constellation, which includes TRMM TMI, SSM/I, SSM/IS, AMSR-E, and WindSat, are ongoing, results from these activities are presented showing the potential for reducing and better characterizing climate errors/biases in this next generation climate rainfall dataset.
Supplementary URL: