基于Copula函数的桃林口水库防洪风险分析

为了对桃林口水库进行防洪风险分析,采用阿基米德Gumbel-Hougaard Copula连接函数,构建了入库洪峰与洪量的两变量联合分布模型,由随机抽样模拟洪峰和洪量系列,然后基于洪水类别生成洪水过程线。同时综合考虑了水库泄洪能力与水位库容关系水力不确定性因素,采用Monte Carlo法计算桃林口水库在现行常规调度条件下不同风险因子组合对应的水库调度风险定量关系。结果表明,桃林口水库防洪风险系统最主要的控制因素是水文风险,水力风险影响次之。降低汛限水位及采取合理的控泄方式可极大地减轻下游的防洪压力,而该调度方式下水库本身的风险率很小。研究成果为当地水库管理与防汛部门制定科学、合理的防洪规划,为未来合理利用洪水资源探索汛限水位动态化管理提供一定的参考。

Copula function based risk analysis for flood control of Taolinkou Reservoir

In order to carry out the risk analysis for the flood control of Taolinkou Reservoir， the joint distribution model for the two variables， i.e. reservoir inflow flood peak and flood volume， is established with Archimedes Gumbel-Hougaard Copula function， with which the series of flood peak and flood volume are simulated by stochastically sampling， and then the flood hydrograph is generated based on the type of flood. Meanwhile， the hydraulic uncertainties of the flood discharge capacity and the relation between water level and reservoir capacity are comprehensively considered， and the quantitative relationships of the reservoir operation corresponding to the combinations of different risk factors under the current conventional operation of Taolinkou Reservoir are calculated with Monte Carlo Method. The result shows that the most important controlling factor of the flood control risk system for Taolinkou Reservoir is the hydrological risk， while the hydraulic risk has certain impact on it. Lowering the flood limited water level and adopting reasonable mode of discharge control can greatly alleviate the pressure from the downstream flood control， and that the risk rate of the reservoir itself is very small under this operation mode. The study result provides certain reference for the local reservoir management and flood control departments to make scientific and reasonable flood control planning and explore the dynamic management of the flood limited water level for the reasonable utilization of the flood resources in the days to come.