Controls on the size and geometry of landslides: Insights from discrete element numerical simulations

Oded Katz, Julia K. Morgan, Einat Aharonov, Brandon Dugan

Research output: Contribution to journalArticlepeer-review


We use numerical two-dimensional discrete element method (DEM) to study the controls on the size and geometry of an individual landslide. This method allows us to analyze the temporal and spatial evolution of slope failure and landsliding from the intact, pre-failure slope to the restabilized, post-failure slope. The size of our simulated landslides reflects the amount of disintegrated material available for sliding, which is controlled by the peak strength of the slope material and to a lesser degree by the pre-failure slope angle. The observed size of natural landslides, therefore, reflects mainly the local natural material strengths. We hypothesize that the observed global characteristic landslide size is a result of the limited thickness (generally up to a few meters) of disintegrated and weathered material that exists on hillslopes, poised and ready for sliding. The primary controls on the landslide geometry, quantified as the ratio of their thickness to length t/. l, prove to be the residual friction angle of the slope material and the original slope angle. The t/. l is observed to be unchanged for a given initial slope angle and independent of the slope material peak strength or the slope height. We suggest that the narrow range of observed natural t/l is a result of narrow ranges of mechanical properties and initial slope geometries involved in natural slope failure.

Original languageAmerican English
Pages (from-to)104-113
Number of pages10
StatePublished - 1 Sep 2014


  • Discrete element method
  • Geometry
  • Landslide
  • Numerical simulations
  • Size
  • Slope failure

All Science Journal Classification (ASJC) codes

  • Earth-Surface Processes


Dive into the research topics of 'Controls on the size and geometry of landslides: Insights from discrete element numerical simulations'. Together they form a unique fingerprint.

Cite this