Estimating the Return Period for Statistical Probable Maximum Precipitation

Abstract

This study proposes a simple alternative to statistically estimate the Probable Maximum Precipitation (PMP). The PMP is the upper limit of physical rainfall depth and is contradictory to the concept of probability. However, since the upper limit of physical rainfall depth cannot be identified clearly from observed data, it was judged that making a probabilistic estimation of PMP might be reasonable. Hershfield's km, calculated from the daily rainfall depth recorded at 80 stations of the Automated Synoptic Observing System (ASOS) with more than 10 years of observations, was fitted to the Generalized Extreme Value (GEV) distribution with a shape parameter given as a linear function of the average of the annual maximum of the time series at each observation site. Data from the oldest 15 sites in Korea (1961-2017, 57 years) were used to verify the methodology suggested in this study. Hershfield's statistical PMPs and the hydro-meteorological PMPs were somewhat different; however, the return period for the PMPs in terms of daily rainfall depth was estimated to be approximately 20,000,000 years at all sites studied. Keywords: Frequency Analysis, Hershfield, Probable Maximum Precipitation, Return Period