A global landslide non-susceptibility map

At variance with conventional landslide susceptibility assessment, non-susceptibility analysis aims at selecting locations in which the likelihood of landslide occurrence is null or negligible. The advantage of this approach is that it does not require estimating different degrees of likelihood outside of the locations of negligible susceptibility. Thus, it entails the use of simplified classification methods. In this work, we tested and validated the existing non-susceptibility model with 18 global and regional landslide datasets, as a prior for the global application. The existing model was applied previously in Italy and the Mediterranean region, and defined by a nonlinear relief vs. slope threshold curve, below which landslide susceptibility is negligible. Then, we applied a similar analysis, and proposed a global map, using relief and slope obtained from global elevation data at about 90-m resolution. The global map classifies 82.9% of the landmasses with negligible landslide susceptibility. The non susceptible areas are broadly consistent with the very-low susceptibility class in existing global and continental landslide susceptibility maps and a national non-susceptibility map in the conterminous United States. Quantitative analyses revealed that population and settlements are denser within non-susceptible area than elsewhere, which makes the map of potential interest for non-exposure analysis, land planning and disaster responses at a global scale. (c) 2021 Elsevier B.V. All rights reserved.

Automated summary

This study proposes a non-susceptibility analysis method for selecting locations with negligible landslide occurrence likelihood, simplifying classification methods and applying it globally. The global map, which classifies 82.9% of landmasses with negligible landslide susceptibility, can be used for non-exposure analysis, land planning, and disaster response. Population and settlements are denser within non-susceptible areas, making the map potentially valuable for global-scale applications.