Assessing rainfall-triggered landslides considering small-scale depth to bedrock uncertainty: A random field approach

Depth to bedrock (DTB) is a critical factor for rainfall-induced slope failures. However, the influence of uncertainties in these measurements, particularly at a small-scale, has not been fully understood. Numerical modeling was conducted to assess the impact of a variable bedrock topography on the stability of a real-world unsaturated slope. The simulations included a three-dimensional pore-water pressure estimation, derived from the numerical solution of the Richards equation, coupled with a slope stability assessment using numerical limit analysis. The study explored the potential of incorporating random fields (RFs) into an established DTB model to improve the understanding of rainfall-triggered landslides. The proposed methodology was applied to the analysis of a small watershed within the Papagaio River basin in Brazil, an area historically subjected to landslides triggered by rainfall events. Our main findings reveal that small variations in DTB can significantly impact the safety factor and probability of failure estimations. Furthermore, they influence the shape, location and failure volume associated to predicted landslides. The incorporation of RFs effectively addresses small- scale uncertainty in DTB, controls bedrock morphology, and enhances the assessment of probabilistic numerical modeling for landslide susceptibility. This study highlights the importance of accurate and comprehensive DTB characterization for assessing rainfall-induced landslides at local slope scale.