Encounter probability analysis of typhoon and plum rain in the Taihu Lake Basin
Encounter probability analysis of typhoon and plum rain in the Taihu Lake Basin
复制标题
DOI：
10.1007/s11431-010-4158-2
复制
发表时间：
2010-11
期刊：
影响因子：
--
通讯作者：
Siyi Hu;Zongzhi Wang;Yintang Wang;Haoyun Wu;Ju-liang Jin;Xiang Feng;Liang Cheng
中科院分区：
文献类型：
--
作者：
Siyi Hu;Zongzhi Wang;Yintang Wang;Haoyun Wu;Ju-liang Jin;Xiang Feng;Liang Cheng
研究太湖流域遭遇台风、梅雨的可能性，对于调整设计暴雨方案、利用洪水资源、做出流域防洪实时调度决策具有重要意义。基于现有台风与梅雨遭遇现象描述的研究成果，分别选取台风影响太湖流域的起始时间、梅雨季节的开始时间和梅雨季节的结束时间作为台风和梅雨的特征指标。根据1954年至2009年的同步数据，辨识了上述3个随机变量的分布函数，并推导了表达台风和梅雨遭遇可能性的公式。鉴于该公式的复杂性和求解的复杂性，在阿基米德函数中三个函数（Gumbel Copula、Clayton Copula和Frank Copula）适应性比较的基础上，利用Gumbel Copula函数构造了台风和梅雨的联合分布函数。针对基于Gumbel Copula函数的联合分布函数公式的特点，选择数值积分和蒙特卡罗模拟两种方法求解相遇概率，最后分析了计算结果的有效性。
It is of great significance to study the encounter possibility of typhoon and plum rain in the Taihu Lake Basin in order to adjust the schemes of design storm and utilize flood resource and make the decision of real-time scheduling for flood control in the Basin. Based upon the existing research findings of description of encounter phenomena of typhoon and plum rain, the initial time of typhoon affecting the Taihu Lake Basin, the start time of plum rain season and the end time of plum rain season were selected as the characteristic indices of typhoon and plum rain, respectively. According to their synchronous data from 1954 to 2009, the distribution functions of the above-mentioned three random variables were identified and a formula expressing the encounter possibility of typhoon and plum rain was derived. In view of the complexity and solution of this formula, joint distribution functions of typhoon and plum rain were constructed by the use of Gumbel Copula function on the basis of the adaptability comparison among three functions (Gumbel Copula, Clayton Copula and Frank Copula) in Archimedean functions. For the characteristics of the formula with the joint distribution functions based on Gumbel Copula function, two methods, i.e., numerical integration and Monte Carlo simulation, were selected to solve the encounter probability, and the validity of the computed results was analyzed finally.