基于Copula函数的水库汛期分期设计

Application of Copula Function to Reservoir Seasonal Design Flood

鉴于水库汛期分期设计是充分利用水库洪水资源的重要手段，以红枫水库为例，在实测系列洪水中加入历史特大洪水，采用超定量方法取样，用混合VonMises分布拟合洪水发生时间，建立了基于二维Frank Copula函数的洪水发生时间与量级的联合分布，推求分期设计洪水，进而进行了调洪计算，确定了水库汛期分期方式为汛前期（5月1日～5月31日）、主汛期（6月1日～7月31日）、汛末期（8月1日～9月30日），并在满足防洪标准的前提下，确定了各分期汛限水位分别为1239．1、1236．5、1239．6m。

In view of reservoir flood season staging design is the important way to make full use of the reservoir flood resources. Taking Hongfeng Reservoir as an example, use of the historical flood in observed flood was considered. Mixed Von Mises distribution was used to describe the flood occurrence dates of the Peaks-Over-Threshold （POT） sampling se- ries. Based on two-dimensional Frank Copula function, a joint distribution which considered the flood occurrence dates and magnitudes was established to derive seasonal design flood for flood regulating calculation. The result shows that the flood season of Hongfeng Reservoir may be divided into three sub-seasons, i. e. , the early flood season（from May 1 to May 31）, the main flood season（from June 1 to July 31）, and the late flood season（from August 1 to September 30）. On the premise of meet the flood control standard, the limited water level of reservoir by stage respectively is 1 239. 1, 1 236.5 and 1 239.6 m.