Time-varying reliability analysis of unsaturated reservoir bank slope under water level drop considering multi-parameter spatial variability

The reservoir water level drop and the spatial variability of soil parameters are important factors affecting slope stability, and it is important to evaluate the reliability of reservoir slopes under the action of these two factors. However, most of the existing reliability analyses only consider the spatial variability of a single shear strength parameter or hydraulic parameter, and only analyze the static reliability at a certain moment, ignoring the influence of multi-parameter spatial variability and time-varying factors. For this reason, a time-varying reliability analysis method for reservoir slopes that considers both factors is proposed. The Karhunen-Loève expansion method is used to discretize the random field of soil parameters, and the Slice Inverse Regression (SIR) method is used to reduce the dimension of random variables. Based on the reduced variables, the Gaussian Process Regression (GPR) surrogate model is constructed, and then the Monte Carlo Simulation (MCS) method is used to evaluate the slope failure probability. Finally, the effectiveness of the proposed method was verified by taking an unsaturated reservoir bank slope as an example, and the reliability variation law of the slope under different water level plunge conditions was explored, and sensitivity analysis of key soil parameters was conducted. The proposed method can accurately and efficiently describe the time-varying law of slope failure probability (Pf), which provides an effective way for the time-varying reliability problem of unsaturated slope considering multi-parameter spatial variability. The faster the reservoir water plummets, the faster the slope safety factor (FS) decreases and the failure probability increases. The spatial variability and correlation of multiple soil parameters have an impact on the calculation results of slope reliability.