EUNIS Inland habitats with no or little soil and mostly with sparse vegetation, predicted distribution of habitat suitability - version 1, Nov. 2021

This metadata corresponds to the EUNIS Inland habitats with no or little soil and mostly with sparse vegetation, predicted distribution of habitat suitability dataset. These habitat types are non-coastal habitats on substrates with no or little development of soil, mostly with less than 30% vegetation cover which are dry or only seasonally wet (with the water table at or above ground level for less than half of the year). Habitats which may have a high vegetation cover include crevices of rocks, screes or cliffs and habitats formed by carpets of moss. Includes subterranean non-marine caves and passages including underground waters and disused underground mines and habitats characterised by the presence of permanent snow and surface ice other than marine ice bodies. The modelled suitability for EUNIS inland habitats with no or little soil and mostly with sparse vegetation, is an indication of where conditions are favourable for the habitat type based on sample plot data (Braun-Blanquet database) and the Maxent software package. The modelled suitability map may be used as a proxy for the geographical distribution of the habitat type. However, note that it is not representing the actual distribution of the habitat type. As predictors for the suitabilty modelling not only Climate and Soil parameters have been taken into account, but also so-called RS-EVB's, Remote Sensing-enabled Essential Biodiversity Variables like Landuse, Vegetation height, Phenology, LAI(Leave Area Index) and Population density. Because the EBV's are restricted by the extent of the Remote Sensing data (EEA38 countries and the United Kingdom) the modelling result does also not go beyond this boundary. The dataset is provided both in Geodatabase and Geopackage formats. The Training map files show the modelled suitable distribution, omitting the 10% of occurrence records in the least suitable environment under the assumption that they are not representative of the overall suitable habitat distribution. The 10 percentile training presence is an arbitrary threshold which omits all regions with habitat suitability lower than the suitability values for the lowest 10% of occurrence records.