摘要
福建省是我国台风灾害比较严重的地区之一,随着社会经济的飞速发展,台风灾害给人们带来的损失和潜在风险显著提升。由于受极值数据的影响,台风灾害损失数据不符合传统正态分布特征,文章尝试采用基于极值理论的POT模型与对数正态分布所组成的复合模型拟合福建省历年因台风灾害造成的经济损失,分析台风巨灾损失的厚尾特征。分析方法中选取的POT模型能最大限度的使用极值数据,并且缺失的数据对POT模型影响比较小,而复合的对数正态分布能改善POT模型选取阈值具有的主观性,因此复合模型能充分体现在极值数据影响下的优势。研究结果为今后的气象灾害损失额度的分析、巨灾类金融产品开发提供科学依据,具有实际意义。
Fujian province is one of the areas with severe typhoon disasters in China.With the rapid development of social economy,the losses and potential risks brought by typhoon disasters to people are significantly increased.Due to the influence of extreme value data,typhoon disaster loss data does not conform to the traditional normal distribution.In this paper,we try to use the composite model of POT model and lognormal distribution based on extreme value theory to fit the economic losses caused by typhoon disasters in Fujian province over the years,analyze the tail characteristics of typhoon catastrophe losses.The POT model selected in the analysis method can maximize the use of extreme value data,and the missing data has little influence on POT model.Because the composite lognormal distribution can improve the subjectivity of POT model selection threshold,the composite model can give full play to its advantages under the influence of extreme value data.The research results provide scientific basis for the analysis of meteorological disaster loss and the development of catastrophe financial products in the future.
作者
罗金炎
徐飞
李燕
吴嘉颖
沈煜
LUO Jinyan;XU Fei;LI Yan;WU Jiaying;SHEN Yu(School of Mathematics and Data Science,Minjiang University,Fuzhou 350108,China;New Huadu Business School,Minjiang University,Fuzhou 350108,China)
出处
《沈阳师范大学学报(自然科学版)》
CAS
2020年第5期417-421,共5页
Journal of Shenyang Normal University:Natural Science Edition
基金
福建省科技厅自然科学基金资助项目(2017J01769)。
关键词
POT模型
极值理论
对数正态分布
台风灾害损失分布
POT model
extreme value theory
lognormal distribution
typhoon disaster loss distribution