Abstract
Tick-borne encephalitis (TBE) is the most common viral tick-borne disease in Europe causing thousands of human infections every year. Available risk maps in Europe are solely based on human incidences, but often underestimate areas with TBE virus circulation as shown by several autochthonous cases detected outside known risk areas. A dataset of more than 1300 georeferenced TBE virus detections in ticks and mammals except for humans was compiled and used to estimate the probability of TBE virus presence in Europe. For this, a random forests model was implemented using temperature- and precipitation-dependent bioclimatic variables of the WorldClim dataset, altitude, as well as land cover of the ESA GlobCover dataset. The highest probabilities of TBE virus presence were identified in Central Europe, in the south of the Nordic countries, and in the Baltic countries. The model performance was evaluated by an out-of-bag error (OOB) of 0.174 and a high area under the curve value (AUC) of 0.905. The TBE virus presence maps may subsequently be used to estimate the risk of TBE virus infections in humans and can support decision-makers to identify TBE risk areas and to encourage people to take appropriate actions against tick bites and TBE virus infections.