Natural Hazards and Earth System Sciences (Apr 2022)

Insights from the topographic characteristics of a large global catalog of rainfall-induced landslide event inventories

  • R. Emberson,
  • R. Emberson,
  • R. Emberson,
  • D. B. Kirschbaum,
  • P. Amatya,
  • P. Amatya,
  • P. Amatya,
  • H. Tanyas,
  • O. Marc

DOI
https://doi.org/10.5194/nhess-22-1129-2022
Journal volume & issue
Vol. 22
pp. 1129 – 1149

Abstract

Read online

Landslides are a key hazard in high-relief areas around the world and pose a risk to populations and infrastructure. It is important to understand where landslides are likely to occur in the landscape to inform local analyses of exposure and potential impacts. Large triggering events such as earthquakes or major rain storms often cause hundreds or thousands of landslides, and mapping the landslide populations generated by these events can provide extensive datasets of landslide locations. Previous work has explored the characteristic locations of landslides triggered by seismic shaking, but rainfall-induced landslides are likely to occur in different parts of a given landscape when compared to seismically induced failures. Here we show measurements of a range of topographic parameters associated with rainfall-induced landslides inventories, including a number of previously unpublished inventories which we also present here. We find that the average upstream angle and compound topographic index are strong predictors of landslide scar location, while the local relief and topographic position index provide a stronger sense of where landslide material may end up (and thus where hazard may be highest). By providing a large compilation of inventory data for open use by the landslide community, we suggest that this work could be useful for other regional and global landslide modeling studies and local calibration of landslide susceptibility assessment, as well as hazard mitigation studies.