Representative Profile Model (RPM): A new physically-based model for assessing the hazards of colluvial landslides at local scale

The physically-based models for regional landslide hazard assessment typically use straight and homogeneous slope geometry and an infinite slope assumption. They assume that the sliding surface and saturation line are parallel to the surface, neglecting the variations in topography and soil thickness across different sections of the slope. This simplification can result in substantial inaccuracies in the regional landslide hazard assessment. To address these limitations, this study proposes a novel, spatially-distributed and physically-based model known as the Representative Profile Model (RPM). RPM distinguishes itself by using slope units rather than grid units, as the primary units of assessment. It efficiently integrates soil thickness and groundwater level information to automatically generate a detailed representative profile for each slope unit. These profiles include a ground surface line, a sliding surface, and a saturation line. This means that RPM can well take into account the effects of topographic relief and spatially uneven distribution of soil thickness for quantifying regional slope stability. Moreover, RPM combines the residual thrust method with the Monte Carlo method. This integration allows for the calculation of failure probabilities for each slope unit, thereby enabling comprehensive and complex susceptibility and hazard assessments at a local scale. A local scale assessment of landslide susceptibility and hazard in Tiefeng Township, Wanzhou District, Chongqing was carried out, with the RPM model. Subsequently, a comparative analysis was conducted with the TRIGRS model, which is based on grid units. The superior performance of RPM was clearly demonstrated by our findings.