RegionGrow3D: A Deterministic Analysis for Characterizing Discrete Three-Dimensional Coseismic or Precipitation-Induced Landsliding on a Regional Scale

Abstract:

Characterizing landslide hazards on a regional scale presents an effective means of mitigating potential losses of life, property and infrastructure, especially during significant seismic or rainfall events. Regional-scale landslide susceptibility is typically characterized through deterministic or probabilistic mapping, which relies heavily on identifying the size and location of potential slope failures. Deterministic methods of identifying potential landslides provide means of informing susceptibility mapping by considering physical factors, such as hydrology, seismicity, and the geotechnical properties of soil. Common deterministic landslide susceptibility approaches often use one-dimensional slope stability assessments that (1) do not account for potential landslide volume, (2) do not account for sliding direction, (3) do not account for representative site-specific geotechnical shear strength parameters, and (4) have a systematic bias towards mapping only steep slopes as susceptible to failure. Using a 3D limit-equilibrium (LE) based slope stability approach, the RegionGrow3D (RG3D) Method presents a method of characterizing the area, volume and location of potential landslides by producing discrete landslide polygons at a state of incipient failure on a regional scale given antecedent moisture conditions, as well as disturbance events, including rainfall intensity and inertial loading from seismicity. Digital elevation models (DEMs), paired with geotechnical parameters, hydrological parameters, and seismic intensity measures are used to identify and grow 3D, generalized landslide shapes, informing susceptibility analyses that can be used to assess hazard and risk to transportation infrastructure on a regional scale. In this study, the utility of the RG3D Method in susceptibility mapping is demonstrated through case studies in the Oregon Coast Range, USA. The RGM provides an efficient, deterministic, LE-based approach to assessing landslide susceptibility on a regional scale, which also provides a framework for future potential analyses, as it can be modified to incorporate other spatial input parameters.