雅砻江流域非一致性洪水分析

Analysis of Non-stationary Flood in Yalong River Basin

雅砻江干流下游河段大规模梯级水电站的逐步开发改变了下垫面条件,同时也改变了天然洪水的分布特性,不再满足传统洪水频率分析中对其一致性的要求。通过定性和定量分析,得出雅砻江锦屏一级水电站年最大1日洪量和年最大3日洪量均在1989年发生变异。在此基础上,针对年最大1日洪量及其发生日期两个变量进行了研究。首先以变异年份1989年将原序列划分为两个子序列,并构造了P-Ⅲ混合分布模型;其次应用方向数据统计理论,将Von Mises分布作为其发生日期的分布;最后通过Clayton Copula函数求取了两变量的联合超越概率分布、条件超越概率密度等。分析结果可为该电站合理确定设计洪水及汛期防洪调度提供相关参考。

The gradual development of large-scale cascade hydropower stations in the lower reaches of Yalong River mainstream has changed the underlying surface conditions and the distribution characteristics of natural flood, which makes the flood unable to satisfy the requirements of stationary in traditional flood frequency analysis. Through the qualitative and quantitative analysis,the annual maximum 1-day flood volume and the annual maximum 3-day flood volume of Jinping I Hydropower Station have a mutation in the year of 1989. On this basis, two variables of annual maximum 1-day flood volume and its occurrence date are studied. Firstly, the P-Ⅲ mixed distribution model is constructed with two subsequences before and after the mutation year of1989. Secondly, the Von Mises distribution is used as the distribution of annual maximum 1-day flood volume occurrence date by applying the directional data statistics theory. Finally, the exceeding joint probability distribution and exceeding conditional probability density of the two variables are obtained by Clayton Copula. The results can provide some references for reasonable flood design and flood control of the station.