A methodology for uncertainty analysis of landslides triggered by an earthquake

Abstract

Strong earthquakes lead to slopes instability in mountainous regions and subsequently trigger landslides. Landslides may cause serious damage to infrastructure such as roads and railways, and may also lead to human injuries and death. The severity of the direct and indirect consequences of damage caused by landslides is inevitable and needs a more reliable approach for risk analysis associated with this phenomenon. In this paper, a novel methodology is employed to estimate the probability of permanent displacement of slopes in mountainous regions. This methodology is developed by using Bayesian network and probabilistic modeling of Newmark displacement. As a part of this methodology, the static safety factor is determined considering uncertain effective parameters. An uncertainty analysis is then performed to estimate the permanent displacement of soil block affected by critical and peak ground accelerations. To demonstrate the application of the developed methodology, slope stability in the Alborz Mountains was considered as a case study. The results highlight that in the absence of the earthquake, instability of the slope is unlikely in the Alborz Mountains. However, with occurrence of an earthquake, it is possible to observe the permanent displacement for the slopes of between 0.2 cm and 5 cm. It was also found that displacements of most probable permanent slope (probability of 19%) occurred in the range of between 1 and 1.5 cm.