Generally, the methods of predicting landslides can be divided into two types - statistical model andnumerical model. Compared with the statistical model, the numerical model can provide more detail andprecise result, but is difficult to employ on basin-scale because of time-consuming calculation. This paperproposed a novel method, which was based on numerical model and multiple regressions as well as usingthe slope unit as the slope-stability analysis target, to predict the landslides on a basin scale. This methodused a new warning indicator, critical water content (Wcr), which is derived from numerical model andhad a clear physical meaning. The new method also had great performance on calculation to predict theoccurring time and the locations of landslides. The heavy rainfall disaster occurring in the Shizugawabasin in 2012, located in Uji, Kyoto, was simulated by the new method. The results showed that the newmethod can not only predict the landslides but also estimate the runoff of the slopes on a basin scale.
展开▼
