Prediction of landslide runout based on influencing factor analysis

Abstract

Landslide runout predictions are essential for producing hazard maps and siting mitigation structures. Data from 55 catastrophic historical landslides in the mountainous area of China are used to determine how topography, material, triggering factors, volume, and initial posture of the slide mass influence runout. Plots and regressions between slide mobility, expressed in terms of the equivalent coefficient of friction (ratio of the fall height to the runout distance), and the slide volume show the necessity to exclude data from zones where motion is blocked by steep orthogonal slopes (“T-type topography”) in defining meaningful empirical relationships. Landslide material exerts strong control on runout, with colluvium showing the strongest dependence on slide volume. Rainfall-induced rockslides have higher mobility than gravity-induced rockslides, but rockslides triggered by great earthquakes may have lower mobility than gravity-induced rockslides. The extra sum of squares analysis results demonstrate that slide volume is the dominant predictor for estimating runout, with posture also influencing the runout distance of non-seismic rockslides. These results provide the basis for our new, multiple-factor predictive method for runout distance. The predicted runout distance in the test case was about 115% of the actual distance. This method may enhance the accuracy of runout distance estimates for landslides in the mountainous area of China.