Stochastic Method for Predicting Risk of Slope Failure Subjected to Unsaturated Infiltration Flow

The increase of rainfall and rise of groundwater level can cause frequent failures of unsaturated soil slopes that result in catastrophic landslides. This paper proposes a stochastic method for predicting the risk of failure of an infinite soil slope subjected to unsaturated infiltration flow. Stochastic models for shear stress and strength at an arbitrary plane of an infinite slope are developed. The accuracy of the proposed method is verified with the Monte Carlo simulation method. The merit of the proposed method lies in its analytical form, which can easily facilitate practical applications. Furthermore, a risk-based sensitivity analysis is undertaken in this study to identify the factors that are the most random and that affect the slope failure most significantly. It is found in this study that the developed stochastic models can capture the randomness of all factors that contribute to slope failure. It is also found that the slope inclination angle ($\theta$), soil air entry value (${\psi }_b$), water table rise from matric suction ($y_b$), dry soil unit weight (${\gamma }_d$), and soil-specific constant ($a_{\varphi }$) are important random variables in the accurate prediction of slope failure. The paper concludes that the proposed stochastic method can serve as a tool for geotechnical engineers to predict the risk of slope failure with accuracy and ease.