Probabilistic slope analysis considering the cross-correlation among soil properties of mountain slopes in South Korea

Abstract

The uncertainty of soil properties and their cross-correlations play significant roles in probabilistic results, emphasizing the need to consider rational and accurate values. In this study, the cross-correlations among soil properties (cohesion, friction angle, and unit weight of soil) were identified using data obtained from 100 ground surveys conducted on natural mountain slopes, confirming a strong correlation among the key soil parameters in the slope stability analysis. Their cross-correlations were discussed in comparison with previously reported values, and a valid correlation matrix with internal consistency was established. To assess the influence of spatial variability of soil properties and their cross-correlation, a probabilistic analysis of slope stability was conducted, and a sensitivity analysis was conducted to measure the relative contribution of the variability in soil properties to the uncertainty in the factor of safety for slope stability. Consequently, the cross-correlations among soil properties are crucial for deriving reliable results. In particular, although the variability of the unit weight of soil and its correlation with other soil properties have often been ignored, considering these factors and establishing the correct correlation matrix have a significant effect on the results of probabilistic analysis. Additionally, the results of sensitivity analysis indicated that the sensitivity of soil properties to the factor of safety for slope stability was significantly influenced by their cross-correlation matrix.