基于Copula函数的广州市短历时暴雨与潮位组合风险分析

Risk analysis of combinations of short duration rainstorm and tidal level in Guangzhou based on Copula function

广州是内涝频发的城市,短历时强降雨与外江洪水、潮水顶托是诱发内涝的主要因素,也是内涝风险研究的重要内容。运用Archimedean Copula函数构建了广州市年最大1小时降雨量与年最大潮位、年最大潮位与相应时段1小时降雨量以及年最大1小时降雨量与相应时段潮位三种联合分布,建立了广州市雨潮组合风险概率模型,得到了雨潮组合的条件风险概率、同现风险概率、治涝风险概率及重现期。分析表明基于Copula函数的雨潮联合分布拟合较好,组合风险分析可靠,分析结果可为广州市城区内涝防治提供科学依据。

Guangzhou is a waterlogging-prone city, and short duration rainstorm, floods and tides are three key factors of its urban waterlogging. Using Archimedean Copula functions, this study calculated three joint distributions for annual maxima of one-hour rainfall and tidal level, annual maximum tidal level and the corresponding one-hour rainfall, and annual maximum one-hour rainfall and the corresponding tidal level. Then, for the different combinations of rainfall and tide in Guangzhou, we developed a risk probability model and calculated various risk probabilities and return periods, including conditional risk probability, simultaneous risk probability, and waterlog prevention risk probability. Results show that the Copula function gives a good fitting to the joint distribution of rainfall and tide and the combined risk analysis is reliable. This study would provide useful information for risk analysis of waterlogging in Guangzhou.