Testing Trends and Return Value Estimation for Extreme Precipitation

This paper deals with usefulness of generalized extreme value distribution based. Annual maxima of daily rainfall data dating from 1901 to 2006 are modeled for ten locations in South Korea. The generalized extreme value distribution (GEVD) is fitted to data from each location. The location parameter of the GEVD is formulated as a function of time to describe the extremes of rainfall and to predict their future behaviors. We quantified the change in extreme rainfall for each location and provided return levels for the years 2016, 2026, 2056, 2106. We also derived estimates of return levels for extreme rainfall. This approach is applied to extreme daily rainfall data for Seoul Korea. The return level of heavy rainfall is calculated from the GEVD models.