| Title |
Analysis of Changes in Rainfall Frequency Under Different Thresholds and Its Synoptic Pattern |
| Authors |
김태정(Kim, Tae-Jeong) ; 권현한(Kwon, Hyun-Han) |
| DOI |
https://doi.org/10.12652/Ksce.2016.36.5.0791 |
| Keywords |
빈도해석;절점기준;포아송-GPD;Bayesian 기법;종관기후 Frequency analysis;Threshold;Poisson-GPD;Bayesian;Synoptic pattern |
| Abstract |
Recently, frequency of extreme rainfall events in South Korea has been substantially increased due to the enhanced climate variability. Korea is prone to flooding due to being surrounded by mountains, along with high rainfall intensity during a short period. In the past three decades, an increase in the frequency of heavy rainfall events has been observed due to enhanced climate variability and climate change. This study aimed to analyze extreme rainfalls informed by their frequency of occurrences using a long-term rainfall data. In this respect, we developed a Poisson-Generalized Pareto Distribution (Poisson-GPD) based rainfall frequency method which allows us to simultaneously explore changes in the amount and exceedance probability of the extreme rainfall events defined by different thresholds. Additionally, this study utilized a Bayesian approach to better estimate both parameters and their uncertainties. We also investigated the synoptic patterns associated with the extreme events considered in this study. The results showed that the Poisson-GPD based design rainfalls were rather larger than those of based on the Gumbel distribution. It seems that the Poisson-GPD model offers a more reasonable explanation in the context of flood safety issue, by explicitly considering the changes in the frequency. Also, this study confirmed that low and high pressure system in the East China Sea and the central North Pacific, respectively, plays crucial roles in the development of the extreme rainfall in South Korea. |