r.slope.stability 2.0 Training data

This dataset includes a GRASS location containing the required spatial input and a start script to perform different types of simulations with r.slope.stability. One single landslide affecting a geologically complex computer-generated slope is defined through the parameters provided in the start script as well as a digital terrain model (dem in m above sea level) and five raster maps representing the bottom of each lithological layer (depth1-depth5, each in m below the terrain surface, the numbers at the end corresponding to the number of the respective layer). The lithology raster map defines the soil class for those areas where no layers are defined, and the gwdepth raster map represents the depth of the groundwater table. Since all raster maps can be created using simple r.mapcalc operations, they are - for demonstration purposes - automatically produced at the beginning of the start script, overwriting the maps already available in the GRASS location.

This dataset includes a GRASS location with the required spatial input as well as a start script, which can be used to perform different types of r.slope.stability simulations within the GRASS location. Note that the area of interest is reduced through the g.region command in order to save computational time. The GRASS location contains a digital terrain model (dem in m above sea level) as well as raster maps of the lithological classes (lithology with five classes, and lithology_simp with one class), layer depths (depth1 is the soil depth, depth2 is the lower boundary of the area of interest, constantly set to 100 m), the distance from the epicentre of a potentially landslide-triggering earthquake (distance in km), the critical displacement (newmarkref in cm), and the observed landslides (landslides where 1=landslide and 0=no landslide). The digital terrain model was modified from existing data and the observed landslides dataset was derived from the IFFI Inventory and is distributed under a CC BY SA 4.0 license. All other datasets as well as the parameters provided in the start script are fictitious.