2024 Volume 80 Issue 16 Article ID: 23-16091
In this study, we examined a statistical method for easily predicting the debris flow distance. First, a decision tree was constructed using the presence and absence of debris flow as the response variable and the surrounding topography as the explanatory variable at 99 sites in Yamaguchi Prefecture. As a result, the model was able to classify the presence and absence of debris flow with an accuracy rate of 0.850. The best estimation method for the distance of debris flow using this model was applied to estimate the distance of debris flow in the debris flow simulation. Similarly, when the presence and absence of debris flow in the disaster records of 67 sites in Yamaguchi Prefecture was used as the response variable, the model was able to classify the presence and absence of debris flow in the disaster records with an accuracy of 0.808, indicating good estimation on the distance of the debris flow. Our method is considered to estimate the distance of debris flow to the relatively high-risk area compared to the results of debris flow simulation and the debris flow hazard area.