Assessing annual probability of rainfall-induced slope failure through a mechanics-based model

The annual failure probability of a slope is an important input for the risk assessment of rainfall-induced landslides. Currently, the annual probability of slope failure under rainfall conditions is usually estimated through statistical analyses of historical landslide events in a region, which cannot consider the local conditions such as soil properties and the slope geometry. In this study, a mechanics-based method is suggested to evaluate the annual failure probability of a slope under rainfall, through which the local conditions of a slope can be explicitly considered. In the suggested method, the uncertainty of soil properties is analyzed based on reliability methods, the uncertainty of annual rainfall events is modeled through a bivariate distribution of intensity and duration, and the most critical annual rainfall is simulated through sampling from the above bivariate distribution. The suggested method is used to study the annual failure probability of a slope assuming it is in Singapore. Through the suggested method, the contributions from different types of rainfalls to the annual failure probability of the slope can also be identified. The suggested method can provide a useful tool to evaluate the annual failure probability of slopes under rainfall conditions, which may increase the capability of the profession to assess the rainfall-induced landslide risk.

Automated summary

This study suggests a mechanics-based method to evaluate the annual failure probability of a slope under rainfall conditions, which can explicitly consider the local conditions of the slope. This method may increase the capability of the profession to assess the rainfall-induced landslide risk by identifying the contributions from different types of rainfalls to the annual failure probability of the slope.