不同重现期标准双变量设计洪水计算方法

Calculation method for bivariate design flood of different return periods

为了确定合理的双变量设计洪水组合,采用C测度、联合和同现重现期3种不同双变量重现期对基于Copula函数的洪水峰量组合进行频率分析,选取极大似然法、条件最可能组合法和双变量同频率法3种方法,研究了不同重现期标准下双变量设计洪水对应的计算方法,从不同方法计算得到的设计洪水组合中选取最优设计洪水值组合,并以新安江水库流域为例进行了计算。结果表明:采用C测度重现期作为双变量频率分析的标准,可确定合理的危险事件的范围,有效规避以同现重现期为标准带来的安全隐患,解决以联合重现期为标准造成的性价比低的问题。

To determine a reasonable bivariate design flood synthesis, three typical bivariate return period standards including C-measure return period,joint return period, and co-occurrence return period were selected for the frequency analysis of flood peak based on Copula function. The maximum likelihood method,conditional most likely combination,and bivariate equivalent frequency combination were chosen as the representative calculation methods. The optimum design flood synthesis was then obtained from the results calculated by the different methods. The calculation method was applied to the Xin＇an River Basin. The result shows that C-measure return period can be a standard in bivariate frequency analysis as it defines rational dangerous regions. It can also avoid the potential safety hazard of co-occurrence return period and is more economic than joint return period.