ABSTRACT
A prototype online extreme precipitation monitoring system is developed from the TRMM TMPA nearreal-time
precipitation product. The system utilizes estimated equivalent average recurrence interval (ARI)
for up-to-date precipitation accumulations from the past 1, 2, 3, 5, 7, and 10 days to locate locally severe
events. The mapping of precipitation accumulations into ARI is based on local statistics fitted into generalized
extreme value (GEV) distribution functions. Initial evaluation shows that the system captures historic extreme
precipitation events quite well. The system provides additional rarity information for ongoing precipitation
events based on local climatology that could be used by the general public and decision makers for
various hazard management applications. Limitations of the TRMM ARI due to short record length and data
accuracy are assessed through comparison with long-term high-resolution gauge-based rainfall datasets from
the NOAA Climate Prediction Center and the Asian Precipitation–Highly-Resolved Observational Data
Integration Toward Evaluation of Water Resources (APHRODITE) project. TMPA-based extreme climatology
captures extreme distribution patterns from gauge data, but a strong tendency to overestimate from
TMPA over regimes of complex orography exists.