Quantifying annual occurrence probability of rainfall-induced landslide at a specific slope

Abstract

Although many countries and regions (e.g., Australia, the Netherlands, Hong Kong) have adopted risk assessment and management methods for landslide risk mitigation, a conventional method based on factor of safety (FS), instead of annual failure probability, is widely used in existing practice of slope engineering. It remains challenging for geotechnical practitioners to assess quantitatively annual failure probability and landslide risk of a specific slope, hindering implementation of landside risk assessment and mitigation in practice. This study proposes an innovative yet practical method to assess quantitatively annual failure probability of a site-specific slope caused by rainfall infiltration. The proposed method integrates slope stability analysis currently used in geotechnical practice with rainfall infiltration into the slope and annual exceedance probability of rainfall to obtain explicit relationships for estimation of the annual failure probability. The quantified annual failure probability comprehensively reflects possible rainfall patterns to the slope, and model settings and soil properties used in slope seepage and stability analyses. It can be implemented conveniently in geotechnical practice through a repeated execution of conventional slope seepage and stability analyses. The proposed method is illustrated by a real slope example in Hong Kong for calculating the annual failure probability of the slope before and after slope retrofitting.