基于Copula方法的极值洪水频率与风险分析  

Extreme Value Flood Frequency and Risk Analysis Based on Copula Method

作　　者：欧阳资生[1] 王同辉[2] 

机构地区：[1]湖南商学院金融学院,湖南长沙410205 [2]湖南师范大学数学学院,湖南长沙410081

出　　处：《统计与信息论坛》2012年第1期71-76,共6页Journal of Statistics and Information

基　　金：教育部人文社会科学项目<基于极值统计的洪水频率分析模型及其在洪灾保险中的应用研究>(09YJA910003)

摘　　要：洪灾风险分析中的一项重要内容是关于洪水发生概率(频率)的估计。以湖南省四大水系中的四个站点50多年的年最高水位数据为基础,结合常用于灾害分析中的分布模型估计出每个站点的洪水频率分布,再利用Copula函数模型得到两两水系间水位协同变化的联合分布函数,进而估计每两条河流同时发生洪水灾害的概率。An important part of flood risk analysis is the flood frequency estimation. In this paper, annual maximum water levels at four water gauging stations, which represent the four major river systems in the Hunan Province with long history data （more than 50 years） for flood frequency analysis, were chosen to estimate the distribution of annual maximum series of each of these gauging stations. Then the joint distribution function of each of the two rivers is obtained with the Copula function model. For the above reasons, the probability of flooding disasters that each of the two rivers happened at the same time in the Hunan Province is estimated.

关 键 词：洪水频率分析 年最大水位Copula函数 极值分布 

分 类 号：F222.3[经济管理—国民经济] P333[天文地球—水文科学]