A Non-Linear, Time-Variant Approach to Simulate the Rainfall-Induced Slope Failure of an Unsaturated Soil Slope: A Case Study in Sapa, Vietnam

Abstract

In this study, the Transient Rainfall Infiltration and Grid-Based Regional Slope-Stability Analysis (TRIGRS), v2.1 program, and module SLOPE/W in the Geostudio package were adopted for assessing rainfall-induced slope failure. TRIGRS was developed by the United States Geological Survey to determine the time-varying groundwater table at the regional scale under rainfall infiltration. The program employs partial differential equations represented by one-dimensional vertical flow in homogeneous materials for unsaturated conditions. With the application of a simple runoff routing scheme combined with the mass balance between rainfall, infiltration, and runoff over the study area, the distribution of the transient pore-water pressures within the entire landscape was simulated considering both the surface and subsurface flow. Additionally, compared to the traditional two-dimensional approach, the topographical conditions were also considered during the groundwater simulation. For conducting the slope stability analysis, a typical cross-section was constructed based on the site description. The predicted water-tables at the observed time of failure of the typical section were extracted and used in SLOPE/W to conduct the time-dependent modelling of rainfall-induced slope failures. In this study, the non-linear method was employed for simulating unsaturated soil shear strength, and the stability of the slope was evaluated using Bishop’s simplified method. We applied the approach to the landslide event that occurred on August 5, 2019, in Sapa district, Lao Cai province, Vietnam. The event resulted in severe damage and blocked the road for days. The predicted results on the stability of the slope as the factor of safety were compared with the actual slope failure during the event. The results showed that, by inputting accurate data, the applied approach could provide valuable evidence about the time of the slope failure.