基于有效降雨的武汉市地质灾害易发性动态评估方法及其应用  

A dynamic assessment method of geological disaster susceptibility and its application based on effective rainfall in Wuhan City

作　　者：李明 张丽 熊启华 袁凯 黄钟吕 LI Ming;ZHANG Li;XIONG Qihua;YUAN Kai;HUANG Zhonglü(Wuhan Meteorological Observatory,Wuhan 430040;Hubei Geological Environment Monitoring Station,Wuhan 430034;Internal Service Centre of Wuhan Meteorological Bureau,Wuhan 430040)

机构地区：[1]武汉市气象台,武汉430040 [2]湖北省地质环境总站,武汉430034 [3]武汉市气象局机关服务中心,武汉430040

出　　处：《暴雨灾害》2024年第5期572-579,共8页Torrential Rain and Disasters

基　　金：武汉市知识创新专项(2022022101015009);武汉市气象局重点项目(WHZ202201)。

摘　　要：降雨等地质灾害诱(触)发耦合因素是地质灾害发生的充分条件。利用2016—2022年武汉市地质灾害数据、降雨实况网格数据以及地质环境基础等资料,采用聚类分析、多元回归等方法,对武汉市地质灾害分布特点及其与强降雨的关系进行研究,并使用多源数据融合、机器学习等方法对地质灾害易发性进行动态评估。结果表明:武汉市约76.49%的地质灾害由降雨诱发,武汉地质灾害强降雨影响期为13d,在影响期内发生地质灾害的可能性高达87.69%。基于有效降雨量的武汉市地质灾害气象风险预测模型,通过引入最新的地质灾害发生情况、地质环境和雨量信息等数据,可实现地质灾害易发性分区的动态评估,24h平均准确率可达79.51%,在2022年地质灾害预测应用中识别率达90%,具有较好的地质灾害预测能力。The coupling factors of geological hazards such as rainfall are sufficient conditions for the occurrence of geological hazards.In this study,the distribution characteristics of Wuhan's geological disasters and their relationship with heavy rainfall by cluster analysis were investigated using multiple regression and other methods,based on Wuhan's geological disaster data,rainfall live grid data,and geological environment foundation during 2016-2022.Meanwhile,multi-source data fusion,machine learning,and other methods were employed to dynamically assess the vulnerability of geological disasters.The results are as follows.About 76.49%of geological disasters in Wuhan are induced by rainfall.The impact period of heavy rainfall on geological disasters in Wuhan is 13 days,and the possibility of geological disasters during the impact period is as high as 87.69%.A meteorological risk prediction model for geological disasters in Wuhan based on effective rainfall was established.This model can achieve a dynamic assessment of geological disaster-prone zoning by introducing the latest geological disaster occurrence,geological environment,and rainfall information.The 24-hour average accuracy rate can reach 79.51%.In the application of geological disaster prediction in 2022,the recognition rate reaches 90%,which shows good geological disaster prediction capabilities.

关 键 词：地质灾害 预测模型 动态评估 机器学习 

分 类 号：P429[天文地球—大气科学及气象学]