基于Copula函数的降水事件危险性评估 ——以子牙河流域典型区为例  

Hazard assessment of precipitation events based on Copula function——Take the typical area of Ziya River Basin as an example

作　　者：赵铁松 王丽荣 郭恩亮 张琪[3,4] ZHAO Tiesong;WANG Lirong;GUO Enliang;ZHANG Qi(Hebei Meteorological Disaster Prevention Center,Shijiazhuang 050021,China;College of Geographical Science,Inner Mongolia Normal University,Hohhot 010022,China;College of Applied Meteorology,Nanjing University of Information Science and Technology,Nanjing 210044,China;Jiangsu Key Laboratory of Agricultural Meteorology,Nanjing 210044,China)

机构地区：[1]河北省气象灾害防御中心,河北石家庄050021 [2]内蒙古师范大学地理科学学院,内蒙古呼和浩特010020 [3]南京信息工程大学应用气象学院,江苏南京210044 [4]江苏省农业气象重点实验室,江苏南京210044

出　　处：《自然灾害学报》2020年第6期199-208,共10页Journal of Natural Disasters

基　　金：国家自然科学基金项目(41375038);河北省科技厅重点研发项目(19275408D)。

摘　　要：降水事件危险性研究作为世界范围内的关注重点,也是河北省的热点问题。本文利用子牙河流域山区河北段16个站点1981-2018年日降水数据识别降水事件,综合考虑降水事件发生的概率和强度构建危险性评估模型。降水事件强度综合考虑持续天数和过程累积雨量两个要素,基于Copula函数确定不同持续天数和累积雨量的降水事件发生的概率。结果表明:1981-2018年期间区域平均降水事件为177次,且研究区降水事件的连续降水天数和累积降水量存在显著相关关系,适合采用Copula函数建立两者的联合分布;连续降水天数的边缘分布拟合用EXP函数和GEV函数效果较好,对于累积降水量的拟合用GEV函数和GAM函数表现较好;在最优Copula函数选择方面,Symmetrised Joe-Clayton函数模拟效果最好;研究区降水事件危险性存在明显差异,总体表现为南部地区危险性高于北部地区。以子牙河流域山区河北段为例开展降水事件危险性研究对于河北省灾害预防、减轻灾害损失具有重要意义。As the focus of worldwide attention,research on the risk of precipitation events is also a hot issue in Hebei Province.In this paper,daily precipitation data from 1981 to 2018 of 16 stations where are distributed in the mountainous area of Ziya River Basin,Hebei Province,are used to identify precipitation events,and a hazard assessment model is constructed by comprehensively considering the probability and intensity of precipitation events.The intensity of precipitation events comprehensively considers the two factors:continuous days and accumulated rainfall of the process,and determines the probability of occurrence of precipitation events with different continuous days and accumulated rainfall based on the Copula function.The results show that:during the period from 1981 to 2018,the regional average precipitation events were 177,and there was a significant correlation between the number of consecutive precipitation days and cumulative precipitation in the study area,it is suitable to use the Copula function to establish the joint distribution of the bivariate data.The EXP function and the GEV function are better to fit the marginal distribution of continuous precipitation days,and the GEV function and the GAM function are better to fit the marginal distribution of the cumulative precipitation.In terms of optimal Copula function selection,the Symmetrised Joe-Clayton function has the best performance.There is a clear difference in the hazard level of precipitation events in the study area,the overall performance is that the hazard level in the southern is higher than that in the northern region.Taking the mountainous area of Ziya River Basin in Hebei Province as an example,it is of great significance for disaster prevention and disaster loss reduction in Hebei Province to assess the hazard of precipitation events.

关 键 词：降水事件 边缘分布 COPULA 危险性 子牙河流域 

分 类 号：X43[环境科学与工程—灾害防治]