Joint-Risk Evaluation of Extreme Precipitation Using Copulas

Effective identification of extreme precipitation characteristics can help urban storm management. In this study, maximum consecutive 5-day precipitation (RX5D), annual count of days when precipitation is higher than 30 mm (R30), and annual total extreme precipitation (R95p) are selected for reflecting extreme precipitation characterise. Copula method is developed for investigating the dependence relationships between these indices. Take Xiamen Island as a case study, where is a coastal city with intense precipitation in rainy season. Results show that R95p will significantly increase with an average rate of 4.3 mm/year, making 91.5% of contribution to the total precipitation increment. In the 10-year return period level, the R95p is larger than 785 mm accompanied with RX5D larger than 311 mm and R30 larger than 15.6 days. Such results can help managers decide the flood control infrastructure design parameters.