GEOMORPHIC FINGERPRINTS OF LARGE LANDSLIDES AT A WATERSHED-SCALE: A CASE STUDY FROM THE CARSON RANGE, WESTERN NEVADA

The automated detection and extraction of large (> 1 km²) landslides from coarse (> 10 m) DEMs remain challenging several decades after remote sensing techniques and digital terrain analyses were first integrated into geomorphic surveys. Landslide deposits and scarps are notoriously difficult for GIS algorithms to distinguish in large spatial datasets due to their heterogeneous geometries and surface features resulting from variable deposit ages, land cover, lithologies, and depositional environments. The characteristic features of individual deposits are modulated by local geologic and environmental conditions, further reducing the capacity of automated workflows to identify landslides over large spatial scales. In order to resolve these issues, new strategies and frameworks must be developed to identify characteristic landslide signatures within large topographic datasets.

We explore a new framework in which the geomorphic signatures of landslides of different ages are assessed at a watershed-scale using publicly-available terrain analysis software. Leveraging a recently-compiled database of previously-mapped landslide deposits throughout Nevada, we extract 10 m-resolution topographic data from individual catchments in the Carson Range and test them for systematic, morphologic variations with age and/or presence of landslide deposits.

We use the TopoToolbox package for MATLAB (Schwanghart and Scherler, 2014) to calculate fluvial and hillslope-based metrics (including surface roughness, anisotropic coefficient of variation, hilltop curvature, channel steepness index, and channel chi index) for A) 18 large landslides and landslide complexes mapped in the Carson Range, B) their respective drainage basins, and C) neighboring drainage basins with no mapped landslides. For each catchment, we quantify how the extent and relative age of landslide deposits scales with each basin-wide metric and assess how local geologic and environmental conditions modulate these relationships. In the future, we aim to use the results of this study to inform and validate prognostic models that predict the extent of landsliding in individual watersheds.