基于Copula的北江最大与最小径流联合分布研究  

Copula-based Joint Probability Behaviors of Hydrological Extremes in the Beijiang River

作　　者：李剑锋[1,2] 刘春玲 张强[1,2] 

机构地区：[1]中山大学水资源与环境系,广东广州510275 [2]中山大学 华南地区水循环与水安全广东省普通高校重点实验室,广东广州510275 [3]珠江水利委员会珠江水利科学研究院,广东广州510610

出　　处：《人民珠江》2013年第1期22-27,共6页Pearl River

基　　金：科技部水专项(201001066);新疆自治区科技攻关项目(200931105);国家自然科学基金面上项目(41071020)

摘　　要：依据北江11个水文站的日流量资料,计算1、3、7天最大径流量和最小径流量。采用Kolmogorov-Smirnov法确定最适合的边缘分布函数,根据Akaike Information Criterion准则确定最适合的Copula函数,估计不同边缘重现期下不同径流量组合的联合重现期。结果表明:联合重现期特具有明显的空间分布特征;不同边缘重现期的联合重现期特征相似;北江西部不容易出现3天和7天最大径流量都较大的洪水现象,但相对容易发生不同历时最小径流量都非常小的干旱水文事件;石角站更容易发生不同历时最大径流量较大,或者不同历时最小径流量都较小的水文事件。Daily runoff data in 11 hydrological stations in the Beijiang River was studied.Maximum(minimum) volumes of runoff in 1,3,and 7 days were calculated based on this dataset.Kolomogorov-Smirnov test was applied to determine the best fitting marginal distribution functions,and Akaike Information Criterion was used to confirm the best fitting Copula functions.And then joint return periods of different combinations of volumes of runoff in various days were estimated in various marginal return periods.The results show that remarkably spatial features in joint return periods were detected.The characters of joint return periods in various marginal return periods were similar.The frequency of hydrological events that maximum volumes of runoff in both 3 and 7 days in the west parts of the Beijiang River are high was relatively low.But in this area,the hydrological drought events that the minimum volumes of runoff in various days are very low occurred in a relatively high frequency.The events that maximum/minimum volumes of runoff in various days are relatively high/low occurred easily in the Shijiao Station.

关 键 词：最大径流量 最小径流量 COPULA 联合重现期 北江 

分 类 号：P338.2[天文地球—水文科学]