Modelling of rainfall-induced landslide: a threshold-based approach

Landslides during the rainy season are a usual phenomenon resulting in loss of human, animal, and property and obstructing the transportation facilities of the area. Modelling is an established tool for the reliable, efficient, and feasible study of landslide mechanisms. Modelling is done by simulating the material properties and boundary conditions for an existing failed slope. In the present study, an area subjected to the recurrence of landslides during the rainy season has been selected to study the mechanism that triggers landslides. After extensive fieldwork and computation of hydro-mechanical parameters through a series of laboratory experiments, a semi-similar material physical model test was conducted to study the sliding mechanism and its crucial factor. Furthermore, numerical modelling has also been done using GeoStudio to analyze the seepage and slope stability parameters to verify and compare the results to assess the suitability of this study. An artificial rainfall simulator is designed and developed to simulate the quantity and intensity of rain to study the failure mechanism and the threshold level of rainfall that causes the triggering of landslides. The physical model slope failed at 80 mm threshold rainfall depth at a 30 mm/h fixed intensity. Numerical analysis results show that the slope remains stable with a safety factor of 1.23 before the rainfall. Still, after simulating the rainfall condition for the physical model test, the slope fails at the same rainfall intensity and depth with an approximately similar failure plane with a safety factor of 0.626, indicating that rainfall is the major affecting parameter that generally initiates the landslide. This study also proves the suitability and feasibility of numerical modelling to analyze different slopes, providing scientific guidance for monitoring and early warning so that preventive measures can be taken to reduce its effect.