基于Copula函数的非连续序列洪水频率分析  

Multivariate discontinuous flood frequency analysis using Copulas

作　　者：李天元[1] 郭生练[1] 栗飞[1] 徐小伟[1] 

机构地区：[1]武汉大学水资源与水电工程科学国家重点实验室,湖北武汉430072

出　　处：《武汉大学学报（工学版）》2011年第6期681-684,共4页Engineering Journal of Wuhan University

基　　金：国家自然科学基金项目(编号:51079100);国家科技支撑计划(编号:2009BAC56B01)

摘　　要：现行洪水频率分析采用单变量分布曲线,应用Copula函数构造洪水联合分布时又未考虑历史洪水资料.本文将一维非连续序列的极大似然法扩展应用到二维联合分布的参数估计,建立了基于Copula函数的非连续序列洪水频率分析模型,并以长江三峡水库设计洪水为例,验证该模型的可行性和合理性.结果表明,所建模型合理可行,为多变量非连续序列设计洪水分析计算提供了一种新的方法.The current flood frequency analysis based on univariate distribution function and Copula-based multivariate model didn＇t account for historical information and paleo-flood.A Copula-based discontinuous flood frequency analysis model was proposed with parameter estimated by two-dimensional maximum likelihood method.The design flood of the Three Gorges reservoir was selected as case study.The results show that the proposed model is reasonable and feasible,which provides a new approach for multivariate discontinuous flood frequency analysis.

关 键 词：历史资料 设计洪水 COPULA函数 联合分布 极大似然法 

分 类 号：P333[天文地球—水文科学]