This paper describes recent results of using Tropical Rainfall Measuring Mission (TRMM) (launched in November 1997) information as the key calibration tool in a merged analysis on a 1* x 1* latitude/longitude monthly scale based on multiple satellite sources and raingauge analyses. The TRMM-based product will be compared with the community-based Global Precipitation Climatology Project (GPCP) results.
The long-term GPCP analysis is compared to the new TRMM-based analysis which uses the most accurate TRMM information to calibrate the estimates from SSM/I and geosynchronous IR observations and merges those estimates together with the TRMM and gauge information to produce accurate rainfall estimates with the increased sampling provided by the combined satellite information. The comparison with TRMM results on a month-to-month basis should clarify the strengths and weaknesses of the long-term GPCP product in the tropics and point to how to improve the monitoring analysis.
The procedure used to produce the GPCP data set is a stepwise approach which first combines the satellite low-orbit microwave and geosynchronous IR observations into a "multi-satellite" product and then merges that result with the raingauge analysis. The Version 1 GPCP products and techniques are described in Huffman et al. (1997). A key feature of the merger technique is a calibration of the geosynchronous IR estimates by the less frequent, but superior estimates based on the low-orbit microwave (SSM/I) data.
Preliminary results produced with the still-stabilizing TRMM algorithms indicate that TRMM shows tighter spatial gradients in tropical rain maxima with higher peaks in the center of the maxima. Over the sub-tropics the preliminary TRMM analyses show lower values in the rain minima. Global tropical and regional values will be compared. These variations will be examined carefully and validated where possible from surface-based observations.
The TRMM analyses will be used to evaluate the evolution of the 1998 ENSO variations, again in comparison with the GPCP aanalyses. Examples of finer time resolution precipitation products based on TRMM and other data will also be examined.