An Early Warning System for sand fly-borne diseases in the Iberian Peninsula

Climate-sensitive vector-borne diseases are increasingly influenced by environmental and climatic variability, posing growing challenges for public health preparedness under climate change. Within the Planet4Health project, an Early Warning System (EWS) is being developed to support anticipatory decision-making for climate-sensitive diseases by integrating climate, environmental, and epidemiological information into operational risk products.

This contribution presents an EWS focused on the sand fly vector (Phlebotomus spp.) and leishmaniasis over the Iberian Peninsula, using machine learning–based modelling approaches. The system integrates high-resolution climate data, climate-derived indicators (e.g. temperature, humidity, and precipitation-related indices), land and environmental variables, and vector presence information to model conditions favourable for sand fly activity and disease transmission. The modelling strategy prioritises interpretable machine learning techniques to ensure transparency and usability for public health and veterinary stakeholders.

The EWS operates across multiple temporal scales, addressing short-term and seasonal forecasts, while also incorporating climate projections to assess potential future changes in  environmental suitability for sand flies and associated disease risk. Machine learning models are trained and evaluated using historical climate and entomological data, capturing non-linear relationships between environmental drivers and vector presence while explicitly accounting for uncertainty. Model outputs are translated into spatially explicit risk maps and alert-oriented indicators designed to support operational surveillance and decision-making.

Results from the Iberian sand fly–leishmaniosis case study demonstrate that the EWS successfully reproduces known spatial patterns of vector suitability and seasonal dynamics across the Peninsula, as well as interannual variability linked to climatic anomalies. The modular and data-driven design of the system supports adaptation of the framework to other regions and climate-sensitive diseases, in line with the broader objectives of Planet4Health.

 

 

Funding: The PLANET4HEALTH consortium is funded by the European Commission grant 101136652. The five Horizon Europe projects, GO GREEN NEXT, MOSAIC, PLANET4HEALTH, SPRINGS, and TULIP, form the Planetary Health Cluster. The data for EDENext was obtained from the Palebludata website (https://www.palebludata.com). The data for Vectornet was obtained from the ECDC.