论文部分内容阅读
传统的洪水单变量频率分布形式不能反映在变化环境下洪水序列的真实分布情况,且不适合构建洪水多变量联合分布进行洪水特征的多变量联合分析。以大清河南支沙河上游王快水库入库年洪峰序列和年最大6日洪量序列为基本数据,基于各序列变异点诊断结果,运用混合分布法确定各序列的理论分布,并以此为边缘分布,采用Copula函数法构建其二维联合分布,对两变量重现期及特定条件下的洪峰和洪量条件频率进行了分析,计算了两变量联合分布设计值。结果表明,非一致性洪水单变量重现期介于二维联合重现期与二维同现重现期之间;当峰量具有较高相关性时,发生超过某一频率洪峰设计值的洪峰,会有较大可能发生超过同频率洪量设计值的洪量。基于两变量联合分布得到的洪峰、洪量设计值比单变量同频率法得到的设计值偏大,从工程设计角度偏于安全,对防洪控制是有利的。
The traditional flood univariate frequency distribution can not reflect the true distribution of flood sequences in a changing environment, and is not suitable for constructing a multivariate joint analysis of flood characteristics with flood multivariate joint distribution. Taking the annual flood peak sequence and the maximum 6th flood amount sequence of the Wangkui reservoir in the upper reaches of the Zhousha River in Henan Province as the basic data, the theoretical distribution of each sequence was determined based on the diagnostic results of each sequence variation point, The Copula function method is used to construct the two-dimensional joint distribution. The flood peak and flood condition frequency under the two-variable period and the specific conditions are analyzed, and the joint design of the two variables is calculated. The results show that the non-consistent flood single-variable return period lies between the two-dimensional joint recurrence period and the two-dimensional co-occurrence period. When the peak quantities have a high correlation, the flood peak values exceeding a certain frequency occur Hongfeng, there will be a greater probability of occurrence of floods with the same frequency flood design value. The flood peak value based on the joint distribution of two variables is larger than the design value obtained by the univariate frequency method, which is safe from the perspective of engineering design and is good for flood control.