Slope Reliability Analysis and Risk Assessment: A Modern Computational Perspective

Slope failures or landslides are a major geo-hazard worldwide. Quantitative slope risk assessment and control have been used as an effective method for mitigating landslide hazards. A key task of quantitative slope risk assessment is to evaluate probability and consequences of slope failure, which relies on the understanding of site conditions, slope stability modeling and analysis, uncertainty quantification and propagation, etc. It can be rationally accomplished under a probabilistic framework through slope reliability analysis and risk assessment. Slope reliability analysis and risk assessment have an advantage of incorporating various uncertainties inevitably encountered in slope engineering, such as soil stratification uncertainty, inherent spatial variability, transformation and model uncertainties, multiple failure modes, etc. With the advance of modern computational technologies and powers, slope reliability analysis and risk assessment has advanced rapidly in recent years. This study presents some recent developments of advanced computer-based methods for slope reliability and risk assessment. Major topics include efficient slope reliability analysis and risk assessment methods based on surrogate models and advanced simulation methods, and effects of the spatial variability of soil properties on the failure modes, probability of failure and risk of slopes. The presented methods can provide versatile and promising tools for slope reliability analysis and risk assessment in spatially variable soils.

Keywords: slope, reliability analysis, risk assessment, spatial variability, modern computational methods, uncertainty.