PREDICTING WITHIN-UNIT VARIABILITY OF LANDSLIDE SUSCEPTIBILITY IN THE OREGON COAST RANGE FROM LIDAR TOPOGRAPHY

Material strength of a geologic unit or material is a critical component of landslide susceptibility, hazard, and risk analyses. For regional-scale hazard and risk modeling, mapped geologic units are typically assigned a single value for material strength, regardless of spatial heterogeneities that may exist. In the Eocene Tyee Formation of the Oregon Coast Range (OCR), mapped landslides reveal discrete zones of very high (2–3 slides/km²) and low (~0 slides/km²) landslide density that do not correlate to zones of expected higher precipitation or earthquake shaking. These observed differences likely reflect variable material strength, and therefore landslide susceptibility, across the region. Field observations across the Tyee Formation show significant changes in composition, ranging from massive sandstone to silt-dominated rhythmically bedded sandstone and siltstone. We present a model that predicts spatial heterogeneity of material strength and landslide susceptibility within the OCR based on topographic frequency content and morphological indices derived from high-resolution lidar elevation data. This model was developed using mapped landslide densities and verified by an independent set of field observations of rock mass strength as a proxy for landslide susceptibility. Results from this model offer a spatially continuous estimate of variations in relative material strength across a geologic unit by leveraging the morphology of the landscape, which could significantly improve regional-scale landslide hazard and risk studies where uniform properties would otherwise be assumed.