Forecasting human tick-borne encephalitis risk across Europe through quantitative models integrating environmental and exposure drivers

Tick-borne encephalitis (TBE) poses a growing public health challenge across Europe, driven by ecological, climatic and socio-demographic changes. We present a quantitative spatio-temporal predictive framework to forecast the annual probability of TBE across Europe at regional (NUTS3) and municipal administrative levels. The model leverages a boosted regression tree algorithm trained on standardized human TBE case data provided by the European Centre for Disease Prevention and Control and national public health institutions. Our modelling framework integrates drivers related to both the hazard of TBEV circulation in the environment and human exposure to tick bites, namely the habitat suitability of vertebrate hosts, precipitation, forest cover, autumnal cooling rate, forest road density, and population density. Results exhibit strong predictive performance for the years 2017-2025 (regional AUC ≈ 0.84; municipal AUC ≈ 0.82), demonstrating robust discrimination between regions with and without reported TBE cases. Predicted risk maps highlight high probability areas in central-eastern Europe, the Baltic states, and Nordic coastal regions, while temporal trend analysis reveals statistically significant increases in predicted TBE risk extending into north-western and south-western European areas. These results reflect the combined effects of environmental suitability and human activity patterns, underscoring their value for early risk assessment. By integrating ecological hazard and human exposure drivers into a dynamic modelling framework, this work provides actionable early annual risk estimates of human TBE risk before the start of the tick questing season, supporting surveillance and prevention activities by public health authorities under global environmental change.