Abstract
Probabilistic analysis to the drought events is a crucial scientific process that provides foundational data for developing water resource strategies to ensure water supply for municipal, industrial, and agricultural purposes. Drought analysis requires consideration of two variables, duration and severity, making it more complex than flood frequency analysis, which typically involves univariate analysis. In bivariate analysis for drought events, the derivation of a joint probability distribution using the best fitted probability distributions to the selected variables was very difficult or not possible mathematically. Therefore, in recent studies, a Copula function has been applied to resolve this limitation. While recent research has focused on applying Copula functions, the comparative studies presenting results from univariate analysis, bivariate analysis using specific distributions, and bivariate analysis using Copula functions have remained relatively scarce. Therefore, this study tried to focus the comparison of the results from techniques used in drought frequency analysis and suggest the advantage of a Copula function. The selected sites in this study are Hongcheon and Jeongseon in South Korea, which experienced severe drought damages in 2009. Also, the 6 rainfall data sets (historical data and the future data by SSP1-2.6 and SSP5-8.5 climate change scenarios) from two rainfall gauges were used to perform the various types of drought frequency analysis. Especially, the fundamental theory to consider relationship between the return period and the exceedance probability in the bivariate analysis was described to suggested that Copula functions can effectively enhance drought frequency analysis.