Numerical models of translational landslide rupture surface growth

We analyze the initiation and enlargement of the rupture surface of translational landslides as a fracture phenomenon using a two-dimensional boundary-element method. Both processes are governed largely by the stress field and the pre-existing planes of weakness in a slope. Near the ground surface, the most compressive stress becomes either parallel or perpendicular to the slope, depending on the topography and regional stresses. The shear stress available to drive slope-parallel sliding in a uniform slope thus is small, and therefore pre-existing weaknesses are required in many cases for sliding. Stresses in a uniform slope favor the initiation of sliding near the slope base. Sliding can progress upslope from there in retrogressive fashion. Most slopes are not uniform and notches in a slope will concentrate stresses and generally promote sliding there. As the region of sliding at depth enlarges, the stress concentration near the edge of the area of slip will tend to rise. Stress concentrations can become sufficient to open fractures above and below a basal slide plane, in keeping with observations. If one tip of a slide plane intersects the ground surface, then stresses near the other tip can increase markedly, as can slip. Our analyses show that slope-parallel sliding along a plane at depth will cause downslope extension in the upslope half of a slide mass and shortening in the downslope half, consistent with observations. Displacement profiles that could be interpreted as rotational can result from sliding along such a plane, however careful analysis of surface deformation can be used to understand sliding at depth.