1. A dynamic assessment method of geological disaster susceptibility and its application based on effective rainfall in Wuhan City
- Author
-
Ming LI, Li ZHANG, Qihua XIONG, Kai YUAN, and Zhonglü HUANG
- Subjects
geological hazard ,prediction model ,dynamic assessment ,machine learning ,Meteorology. Climatology ,QC851-999 - Abstract
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.
- Published
- 2024
- Full Text
- View/download PDF