基于二维极值分布的亳州地区强降水概率特征模拟

Probability Characteristics of Heavy Precipitation Based on Bivariate Gumbel-Logistic Model at Bozhou Area

在全球气候变化的影响下,极端天气气候事件频繁发生。该文结合近年来极端天气气候变化研究中应用极值统计理论的新成果,采用亳州4个站建站以来汛期的实测降雨资料,探讨了二维Gumbel分布模型,在研究其经验频率和理论频率计算方法的基础上,根据随机变量的边缘分布和条件概率分布,分析了汛期最大强降水雨量及最大过程雨量的联合分布,建立了其联合重现期和条件重现期的分布模型,拟合效果很理想。

Frequent climate extremes often occur under changing climate. Using the long-term observed precipitation data, the two-dimension Gumbel model is applied at Bozhou station based on the new theory of extremes statistics. The empirical frequency and theoretical frequency of precipitation extremes are analyzed, and then the joint distribution of the heaviest precipitation and its corresponding heaviest total precipitation is investigated. The models of joint return period distribution and Conditions return period distribution are built, which is useful for the studies of precipitation extremes.