Abstract: The depth integration algorithm simplifies the three-dimensional model of landslide sliding along the surface to a two-dimensional model, enhancing solving efficiency by diminishing the number of unknowns in the governing equation. Material point method (MPM) has the advantages of both grid based method and meshless method, and can avoid mesh distortion when simulating large landslide deformation. A numerical model for landslides has been developed using the depth-integral coupled material point method, with a detailed description of the algorithm's specific process provided. Utilizing the influence domain material point method (IDMPM), two benchmark tests were conducted on typical landslide scenarios: one with a non-inclined, smooth bottom, and the other with an inclined, non-smooth slope. The depth integral coupled material point method model demonstrates high accuracy in predicting key slip parameters, including remote distance, velocity, and depth. In contrast to the conventional material point method, the depth integral coupled material point method model significantly enhances operational efficiency. Furthermore, the research outcomes offer a robust theoretical foundation and timely support for analyzing and predicting the extent of damage in landslide geological disasters, conducting hazard assessments, and facilitating emergency response.

Key words: depth integral, material point method, landslide, large deformation, depth integral coupled material point method