Impact of TRMM Data on a Low-Latency, High-Resolution Precipitation Algorithm for Flash-Flood Forecasting

Abstract

ata from the Tropical Rainfall Measuring Mission (TRMM) have made great contributions to hydrometeorology from both a science and an operations standpoint. However, direct application of TRMM data to short-fuse hydrologic forecasting has been challenging because of the data refresh and latency issues inherent in an instrument in low Earth orbit (LEO). To evaluate their potential impact on low-latency satellite rainfall estimates, rain rates from both the TRMM Microwave Imager (TMI) and precipitation radar (PR) were ingested into a multisensor framework that calibrates high-refresh, low-latency IR brightness temperature data from geostationary platforms against the more accurate but low-refresh, higher-latency rainfall rates available from microwave (MW) instruments on board LEO platforms. The TRMM data were used in two ways: to bias adjust the other MW data sources to match the distribution of the TMI rain rates, and directly alongside the MW rain rates in the calibration dataset. The results showed a significant reduction in false alarms and also a significant reduction in bias for those pixels for which rainfall was correctly detected. The MW bias adjustment was found to have much greater impact than the direct use of the TMI and PR rain rates in the calibration data, but this is not surprising since the latter represented perhaps only 10% of the calibration dataset.