EU Science Hub

Synoptic Pan-European Landslide Susceptibility Assessment: The ELSUS 1000 v1 Map

In order to identify areas in Europe susceptible to landslides in the context of the EU Soil Thematic Strategy and the associated Proposal for a Soil Framework Directive, a harmonised approach encompassing geographically-nested susceptibility assessments (“Tiers”) and, where possible, the use of comparable datasets as input criteria for susceptibility modelling was devised. The first version of the 1 km grid size European Landslide Susceptibility Map (ELSUS 1000 v1), covering the EU and neighbouring countries, is derived from “Tier 1” assessment. The mapping approach employed includes first a climate-physiographic regionalisation of the study area. For each region, a spatial multicriteria evaluation model is established to evaluate landslide susceptibility using commonly available pan-European datasets on slope angle, lithology and land cover, which are considered as the main conditioning factors for all types of landslides at this scale. Factor weights are assigned through pairwise comparisons using analytical hierarchy processes for each region, while region-specific factor class weights are initially established by computing landslide frequency ratios using more than 102,000 landslide locations across Europe. For each model region, a pixel-based susceptibility index is calculated by linear summation of conditioning factor weights and factor class weights. Each index map is then evaluated and classified into five susceptibility levels using true positive ratio breaks derived from receiver operating characteristics curves obtained with the landslide inventory. Finally, the region-specific classified susceptibility maps are spatially combined into the synoptic ELSUS 1000 v1 map. The map is available from the European Soil Data Centre (ESDAC), hosted by the Joint Research Centre, together with ancillary datasets, including a reliability evaluation of the susceptibility map. Further work is in progress to improve the accuracy of the map, mainly by integrating into the assessment a new pan-European lithological dataset and further landslide locations for areas not represented in the current inventory.