Enhancing rainfall frequency analysis through bivariate nonstationary modeling in South Korea

Abstract

This study conducted a bivariate nonstationary frequency analysis utilizing rainfall events to capture the multidimensional nature of rainfall phenomena and rainfall pattern variability in South Korea. Extreme events were identified by the peaks over threshold (POT) method which enhanced the accuracy of estimation. The nonstationary model, incorporating a nonlinear regression using time as a covariate instead of the scale parameter in the generalized Pareto distribution (GPD), provided a more stable interannual variability of rainfall representation under a dynamic climate compared to stationary models. The ability of the bivariate POT method threshold $\left(T_{and}\right)$ to enhance our understanding of climate change by extracting events with high values in both variables was confirmed. Furthermore, bivariate analysis and nonstationarity significantly influenced the estimation of the return period, indicating that the proposed framework facilitates robust adjustment to nonstationary rainfall patterns, ensuring the dependable utilization of current design frequencies in the context of climate change.