基于聚类与统计分级的河北省气象灾害综合区划研究  

Study on Regionalization of Meteorological Disasters in Hebei Province Based on Cluster and Statistics Classification

作　　者：王瑛[1,2] 祁京 王姝苓 李璨 俞海洋 WANG Ying;QI Jing;WANG Shuling;LI Can;YU Haiyang(Key Laboratory of Environmental Change and Natural Disaster of Ministry of Education,Beijing Normal University,Beijing 100875,China;Academy of Disaster Reduction and Emergency Management,Ministry of Emergency Management and Ministry of Education,Beijing 100875,China;Hebei Xiong’an New Area Meteorological Bureau,Hebei 071100,China;China Meteorological Administration Xiong’an Atmospheric Boundary Layer Key Laboratory,Hebei 071100,China)

机构地区：[1]北京师范大学环境演变与自然灾害教育部重点实验室,北京100875 [2]应急管理部教育部减灾与应急管理研究院,北京100875 [3]河北雄安新区气象局,河北雄安071700 [4]中国气象局雄安大气边界层重点开放实验室,河北雄安071100

出　　处：《灾害学》2023年第2期85-88,113,共5页Journal of Catastrophology

基　　金：河北雄安新区气象局“雄安新区重大气象灾害风险评估及对策研究”项目。

摘　　要：传统的气象灾害区划方法过于依赖专家经验。因此,该文探索基于聚类与统计分级方法,由计算机独立完成区划的技术。以河北省暴雨洪涝、冰雹、大风等9种气象灾害为例,通过K均值聚类的方法实现了全自动的区域划分,将河北省气象灾害综合区划分为6个区,并对比了不同单灾种危险性分级对最终区划的影响,基于分位数法得到的区划方案最优,其各个区域的分布与历史灾情分布较为一致。The traditional regionalization depends too much on experience from experts.Therefore,based on clustering and statistical classification methods,This paper explore the technology of zoning independently completed by computer.Taking nine meteorological disasters including rainstorm,flood,hail and gale in Hebei province as examples,This paper realize automatic regional division by k-means clustering method,and divide Hebei province into 6 regions.This paper also compare the influence of different hazardous classification of single disaster on the final zoning,and the results show that the zoning scheme based on quantile method is the best,and the regionalization are basically consistent with the historical disaster distribution.

关 键 词：气象灾害 灾害区划 聚类 分级 河北省 

分 类 号：P429[天文地球—大气科学及气象学] X43[环境科学与工程—灾害防治] X915.5