Climate-AI-infection-REsponse (CLAIRE)

Important results from the project

UMU and LU, together with researchers from Heidelberg, has developed state-of-the-art algorithms to examine and identify relevant climate variables for predicting infectious diseases, as well as the spread of mosquitoes and ticks across Europe and Sweden. Using climate scenarios, these models can create projections for the spread of diseases and mosquito/tick populations. SMHI identified relevant climate indices, extracted, quality-assured, and bias-corrected climate data until 2100. An sftp-server solution has been set up for data delivery.

Expected long term effects

The impact of these algorithms is twofold. First, they can be used as an early warning system to help decision-makers better prepare for future outbreaks. Second, they can create scenarios for disease outbreaks using climate scenarios, aiding informed policy decisions. The algorithms can be adapted to other diseases or spread processes, enhancing machine learning methodologies. Relevant climate variables, including temperature and precipitation, have been identified and delivered with high spatial and temporal resolution.

Approach and implementation

The algorithms were developed with researchers from UMU and Heidelberg University. The main developers were two postdocs at LU. To facilitate communication and knowledge transfer with other CLAIRE members, several trips to UMU and Heidelberg were made, along with general meetings with other participants. SMHI´s work focused on four areas: a) two-way knowledge transfer between partners, b) defining climate datasets and variables for impact modeling, c) generating, bias-correcting, and post-processing climate data, and d) establishing a data pipeline to partners.