A Machine-Learning Framework for Forecasting Climate-Induced Shifts in Mosquito Vector Ecology and Arboviral Disease Risk Across the Middle East

Abstract Climate change and urbanization are reshaping the distribution of Aedes mosquitoes and the risk of arboviral diseases across the Middle East. We developed an ensemble machine-learning framework integrating high-resolution CMIP6 climate projections, synthetic entomological data, and socio-environmental covariates to forecast vector suitability, abundance, and population-weighted transmission risk under multiple Shared Socioeconomic Pathways. Models demonstrated high predictive performance and identified temperature, vapor-pressure deficit, and urbanization as dominant drivers. Projections indicate pronounced northward and altitudinal expansion of Aedes habitats, earlier seasonal peaks, and extended transmission windows, with high-emission scenarios substantially increasing populations at risk by mid- to late-century. Coupling suitability with temperature-dependent extrinsic incubation periods and human exposure produced spatially explicit risk maps relevant for surveillance and control. This framework provides interpretable, uncertainty-aware forecasts to support climate-adaptive public-health planning in the Middle East.