A new network is taking shape to bring artificial intelligence and climate science closer together in the Euro-Mediterranean region. EUMACC (Euro-Mediterranean AI for Climate Change) is a network of researchers and practitioners working at the intersection of climate science, AI, risk assessment and regional climate impacts.

The Euro-Mediterranean region is warming faster than most of the planet. Heat extremes, prolonged droughts, wildfires and stronger storms are already affecting water resources, agriculture, energy, coastal areas and ecosystems, and are projected to intensify under all major emissions scenarios. At the same time, AI and machine learning have become structural tools in climate research, used for detection, forecasting, downscaling, uncertainty characterisation and impact assessment.

EUMACC was created to make this growing research community more connected, visible and impactful. Its mission is to:

• strengthen collaboration between institutions and make it easier to form competitive research consortia;
• increase the visibility and reach of members’ work through shared communication;
• co-organise scientific and outreach activities, from thematic webinars to summer schools that train the next generation of researchers.

EUMACC is conceived as a knowledge infrastructure for long-term cooperation. It is organised in five thematic divisions: Advancing AI; Earth observation and remote sensing; Climate and Earth system modelling; Natural hazards, extremes, risk and impacts; and Adaptation and resilience. The divisions offer a light coordination structure that helps members find shared interests, coordinate proposals and events, and may evolve into more structured working groups as the network grows. Joining is completely free and carries no formal commitment.

The first initiative: The EUMACC Summer School

The first EUMACC activity was the Summer School Learning the Climate System: From Processes to Decisions, held from 7 to 10 September 2026 at Villa del Grumello in Como. It was organised by Politecnico di Milano and Fondazione Alessandro Volta, with the patronage of CMCC, and brought together 30 students and early-career researchers from 13 countries, selected from 71 applications.

The programme moved from climate processes and predictability, through AI methods, to impacts, adaptation and climate services. Lectures were given by leading international scientists, including Sonia Seneviratne (ETH Zürich), Davide Faranda (CNRS/LSCE), Markus Reichstein (Max Planck Institute for Biogeochemistry), Mierk Schwabe (DLR), Thomas Brunschwiler (IBM Zürich), Gustau Camps-Valls (University of Valencia), Ning Lin (Princeton University), Amir AghaKouchak (UC Irvine) and Arthur Hrast Essenfelder (European Commission JRC).

Alongside the lectures, participants worked in teams on a project that ran through the whole school: forecasting tropical cyclone intensity with machine learning. Starting from historical cyclone records and ERA5 reanalysis data, they built a dataset, developed and compared models, and tested them on an unseen period. They also evaluated how probabilistic forecasts could support simplified Early Action Protocols for the Philippines, Bangladesh and Honduras.

The organising committee included Andrea Castelletti and Matteo Giuliani (POLIMI and CMCC) and Guido Ascenso and Elena Matta (CMCC). Practical sessions were led by Guido Ascenso, Aurelio Raffa Ugolini (CMCC) and Andrea Ficchì (POLIMI).

Join the network

The Summer School is only the beginning. EUMACC will continue to promote webinars, workshops and training opportunities for the AI-for-climate community. To find out more, follow upcoming events and join the network, visit eumacc.eu.