Early-warning Model for Rainfall-induced Landslide Disaster in Enyang District,Bazhong City
Rainfall is the inducing factor for 90％ of landslide events. In this paper,206 rainfall induced landslide events in Enyang District of Bazhong City were analyzed by the correlation analysis and logistic re-gression method,and the prediction accuracy of rainfall landslide warning model was considered from the rainy and non-rainy seasons,which improved the accuracy of rainfall landslide implicit statistical warning model. The results of the correlation between previous rainfall and landslide events showed that the rainfall in Enyang area is characterized by a short single rainfall time,and the rainfall in the rainy season is not only larger in rainfall but also stronger in intensity than that in the non-rainy season. The rainfall on the day of the first day before the landslide is significantly related to whether the landslide occurred. According to the anal-ysis of rainfall grade data,the probability of landslide on the day increased rapidly from the light rain to the heavy rain,and the probability of landslide increased slowly from the heavy rain to the heavy rain. In the calculation of the probability of the rainfall-induced landslide events in the logistic regression model,the prediction accuracy of the logistic regression analysis in the rainy season and non-rainy season was 80.56％,which was higher than that of 75.00％obtained in the logistic regression analysis in the whole year.
landsliderainfallrainy seasonlogistic regression model
万方数据
维普
