Prediction of seismic displacement of dry mountain slopes composed of a soft thin uniform layer

• We calculate amplification factors and seismic displacements of dry mountain slopes.
• We perform 2D nonlinear analyses to obtain equivalent acceleration time histories.
• We show that shallow mountain slopes should not be modeled as a rigid block.
• A new flexible sliding block model provide reliable estimate of displacement.
• Success rates of the model to predict the landslide hazard categories are evaluated.

In this paper, the earthquake-induced permanent seismic displacement of dry mountain slopes is calculated from a series of two-dimensional dynamic nonlinear finite difference analyses. The mountain slopes considered are composed of a thin, soft, uniform soil layer underlain by an inclined bedrock parallel to the slope. The material properties of the soil, thickness of the soil layer, and slope inclination angles are varied. Equivalent acceleration time histories are calculated at potential sliding surfaces to derive amplification factors, and a Newmark sliding block analysis is used to calculate the seismic displacements. The calculated seismic displacements of the mountain slopes are compared with those predicted by empirical displacement models. The results show that mountain slopes composed of soft soil layers with a shear wave velocity less than or equal to 200 m/s cannot be modeled as a rigid block because the displacement under strong ground motions will be greatly overestimated. The displacement prediction is significantly enhanced if the slope is modeled as a flexible sliding mass and the amplification characteristics are accounted for. A new flexible sliding block model, which uses multiple ground motion parameters, is shown to provide a reliable estimate of the sliding displacement. The success rate of the model to predict the landslide hazard category ranges from 52 to 88.3%.