Identifying Extreme Precipitation from Satellite Precipitation Data Sets

This study examines statistical distributions of locally defined extreme precipitation from Global Precipitation Climatology Program (GPCP) and Tropical Rainfall Measuring Mission (TRMM). Due to short satellite data record, we use Generalized Extreme Value (GEV) distributions to identify locally rare events that are in greater than 5-year return period for real time global precipitation monitoring using TRMM data. We further compare statistics defined from GEV distributions and more commonly used percentage threshold and standard deviation as a function of climate regimes. The relationship between increased occurrences of extreme precipitation and mean precipitation will also be discussed.