基于Copula函数的不同历时降雨量组合概率研究  

Research of Combined Probability at Different Duration Precipitation Based on Copula Function

作　　者：师义成 王文圣[1] 吴莹[1] 

机构地区：[1]四川大学水利水电学院,四川成都610065

出　　处：《水电能源科学》2016年第4期1-5,共5页Water Resources and Power

基　　金：国家自然科学基金项目(51179110)

摘　　要：为查清越西县不同历时降雨量组合概率情况，根据越西雨量站1954-2012年降水资料，基于三维不对称Gumbel Copula函数构建了年最大1日降水量（R1d）、年最大3日降水量（R3d）、年最大7日降水量（R7d）的三维联合分布函数，计算了不同历时降水量的组合概率、条件概率和相应的重现期。结果表明，R1d、R3d、R7d三变量发生极端降水事件的联合概率最大，同现概率最小，单变量发生的概率介于二者之间；单变量的重现期（T）介于三变量联合重现期（丁a）与同现重现期（丁0）之间；通过组合概率分布图和等值线图可得到给定条件下的两个不同组合事件的条件概率；三维不对称Gumbel Copula函数能够很好地描述不同历时降水变量之间的关系；可为城市防洪排涝及县域水资源开发利用提供参考依据。To identify the different durations of rainfall combined probability situation in Yuexi County,three-dimensional joint distribution function of the maximum 1-day precipitation（R1d）,the maximum 3-days precipitation（R3d）and the maximum 7-days precipitation（R7d）was built based on three-dimensional asymmetric Gumbel Copulas function theory using the precipitation data（1954-2012）from Yuexi precipitation station.And then the combined probability,conditional probability and return period at different duration precipitation were calculated.The results indicate that the three variables＇（R1d,R3 dand R7d）joint probability was the largest,the co-concurrence probability was the smallest and the probability of single variable＇s occurrence was between them.Meanwhile,the return period of the single variable（T）ranged between the joint return period（Ta）and co-occurrence return period（T0）of three-dimensional variables.By combined probability distribution and contour map,the conditional probability of two different combination events could be obtained under the given conditions.Therefore,three-dimensional asymmetric Gumbel Copulas function can well describe the relationship between the different duration precipitations.It can provide a reference for urban flood control and drainage as well as water resources development in the county.

关 键 词：三维不对称Copula函数 不同历时降雨量 联合分布 组合概率 

分 类 号：TV122[水利工程—水文学及水资源]