AI climate models ready for European science
Emulators and learned downscaling adapted and retrained for European use, with every model tested against observations before anyone relies on it.
Developing an AI platform for European climate modelling, designed for deployment on European HPC and AI Factory infrastructures and aligned with initiatives such as DestinE.
Emulators and learned downscaling adapted and retrained for European use, with every model tested against observations before anyone relies on it.
A modular system that handles the whole AI model lifecycle on European supercomputing and AI Factories, designed from the start to work alongside Destination Earth.
A demonstrated pre-operational prototype by 2030, proven in four settings that matter to science, industry and public services.
Four use cases will show what the platform can do, from open questions in climate science to decisions taken in shipping, energy and public health.
Explores the physical robustness and generalisability of the AI climate models. It also enables experiments that are impractical with conventional models, such as very large ensemble projections or long, high-resolution simulations of rare extremes.
Assesses how future changes in winds, sea ice and extreme weather could reshape shipping routes and safety at sea, providing data for climate-resilient maritime planning.
Evaluates the effect of chronic climate stressors such as heat, drought and storms on energy generation and grids over the coming decades. This enables site-specific forecasts that support resilient infrastructure design and risk mitigation.
Maps evolving health risks in future climates, including the risk of infectious disease outbreaks, to enable early warning and adaptation in the public health sector.
The CLAIMA Consortium met online on 14 Sep 2026 to open four years of work on AI for climate. This new Horizon Europe project is dedicated to advancing AI-driven climate science across Europe.