A novel parameter inversion method for an improved DEM simulation of a river damming process by a large-scale landslide

Abstract

The Discrete Element Method (DEM) is one of the most common numerical methods employed for the study of landslide disasters. Herein, this study proposes a method addressing the challenges associated with the selection of contact parameters for DEM blocks (or particles) when simulating landslide hazards. In 2018, two landslides occurred successively at the same slope of Baige, in Tibet, China. To efficiently simulate the entire process of landslide dynamics, a Graphics Processing Unit (GPU) program for blocky DEM, named CoSim-DEM, is developed. Furthermore, a parameter inversion method for the DEM simulation considering the deformation and failure process of landslide is provided and used to obtain the contact parameters for DEM blocks based on the failure process of the first landslide. The failure process of the second landslide is also simulated with the parameters obtained. The geomorphic characteristics of the landslide dams of the two times lanslides obtained by CoSim-DEM are validated by the field investigations. Based on the numerical results, the landslide damming process is divided into five stages and the entire path of the landside is divided into three zones. The results of this study provide a method to construct a database of parameters for geomaterials by using an inversion method based on the previous landslide cases. The method will be important for the rapid assessment and study of landslide hazards in the future.