Application of nonstationary generalized logistic models for analyzing the annual maximum rainfall data in Korea

Recently, the various approaches for the nonstationary frequency analysis have been studied since the effect of climate change was widely recognized for hydrologic data. Most nonstationary studies proposed the nonstationary general extreme value (GEV) and generalized Pareto models for the annual maximum and POT (peak-over-threshold) data, respectively. However, various alternatives is needed to analyze the nonstationary hydrologic data because of the complicated influence of climate change. This study proposed the nonstationary generalized logistic models containing time-dependent location and scale parameters. These models contain only or both nonstationary location and scale parameters that change linearly over time. The parameters are estimated using the method of maximum likelihood based on the Newton-Raphson method. In addition, the proposed models apply to the annual maximum rainfall data of Korea in order to evaluate the applicability of the proposed models.