The influence of different simplified sliding-block models and input parameters on regional predictions of seismic landslides triggered by the Northridge earthquake

Abstract

Regional seismic landslide hazard maps are based on predictions of rigid-sliding-block displacement derived from estimates of earthquake ground shaking, topography, geology, and shear strength. The confidence in these predictions requires comparisons with field observations of landslide occurrence during previous well-documented earthquakes. This paper presents a comparison between observed landslides from the 1994 Northridge, California earthquake and predicted landslides based on sliding-block displacement estimates. Seven empirical displacement models, each of which uses a different combination of ground-motion parameters, are investigated to evaluate which models and associated ground-motion parameters best predict seismic landslides. Using best estimates of ground shaking and shear-strength properties from the Northridge earthquake, sliding displacements are calculated and compared with the locations of observed landslides. Only 20–40% of the observed landslides are captured and the total area of predicted landslides is much larger than observed. The ability to predict landslide occurrence accurately depends less on the displacement model and associated ground-motion parameters, and more on the uncertainty in the model parameters, particularly the assigned shear-strengths. Because current approaches do not take into account the spatial variability of shear strength within individual geologic units, the accuracy of the predictive models is controlled predominantly by the distribution of slope angles within a geologic unit. Assigning overly conservative (low) shear-strength values results in a higher percentage of landslides accurately identified but also results in a large over-estimation of the total landslide area. Making more accurate maps of seismic landslide hazards will require methods to define intra-formational variations in shear strength.