Change-point detection of long-duration extreme precipitation and the effect on hydrologic design: a case study of south Taiwan

摘要

An increase in the global temperature has intensified the hydrologic cycle, which affects the temporal patterns of precipitation. This study analyzed a long-term annual dataset measuring maximum precipitation in south Taiwan, and identified the change point of the time series using the cumulative sum technique. The result reveals a clear change point of the annual maximum rainfall for 24-h durations in 2004 at most observations. The average 24-h-duration precipitation depth in the study area increased by 27 and 36% for 20-year and 100-year extreme events compared with and without data after 2004, respectively. The long-duration precipitation depth demonstrates a significant positive trend following the change point. Furthermore, this study assesses the changes of hydrologic design while precipitation data are updated annually. The designed 20- and 100-year storm will decrease abruptly when the observed data are subsequently updated until 2004. Because of climate change, this issue is worthy of attention in hydrologic designs.