Destination Earth (DestinE), the European Commission’s flagship initiative to build highly accurate digital twins of the Earth system, has officially moved into its third phase.

DestinE, is being implemented by three entrusted entities, the European Centre for Medium-Range Weather Forecasts (ECMWF), the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT) and the European Space Agency (ESA) under the leadership of the European Commission’s Directorate-General for Communications Networks, Content and Technology (DG CNECT).

DestinE’s digital twins for climate change adaptation and weather-induced extreme events provide flexible simulation frameworks that can generate tailored, physically consistent Earth system information, including “what-if” storylines, at the scales where the impacts of extreme events and climate change are observed.

For example, the storyline simulations produced with the climate digital twin enable the replay of record-breaking heatwaves, or devastating floods – and simulate how they would unfold in a different climate, such as in a +2°C warmer world.

The digital twins complement existing European weather and climate prediction capabilities, supporting national and European authorities in better understanding, preparing for and adapting to extremes events and climate change.
In phase two the digital twins - and the Digital Twin Engine framework that underpins them - have been matured and operationalised, and workflows, evaluation, uncertainty quantification, data governance and user access have all been consolidated. Following the successful completion of the first two phases over the past four years, phase three will run until June 2028.

Irina Sandu, Director of Destination Earth at ECMWF, said getting to this stage has taken a lot of work by many partners:
“Working with our partners ESA and EUMETSAT, and in collaboration with more than 100 partner organisations, each bringing their own expertise, has enabled us to operationalise digital twin technology for the Earth system within just a few years. Together, we are developing a unique simulation framework to produce tailored climate, extreme event and impact-sector information, helping users understand possible evolutions and explore the “what if” questions about our changing planet. This kind of capability can only be built through close collaboration across Europe’s scientific, technical and operational communities.

Artificial intelligence activities have also substantially expanded in phase two, including through the development of machine-learning components for different parts of the Earth system (land, ocean, sea ice, waves and hydrology), and AI-based solutions that enhance interactivity with digital twin data.

Director-General of ECMWF, Florian Pappenberger, said:
“Destination Earth’s significant progress throughout its first two phases reflects the shared strength of ECMWF, ESA and EUMETSAT. With phase three, we are turning that progress into lasting capability that will help the European Meteorological and Hydrological Services and other public institutions to better prepare for climate risks. At the same time, developments in AI throughout DestinE are helping to keep Europe at the forefront of creating trustworthy AI for environmental intelligence, which are helping to support preparedness, resilience and innovation across Europe and beyond.”

In phase three, ECMWF and its partners will operate and further evolve the digital twins together with the Digital Twin Engine that enables the running of their workflows on EuroHPC systems and provides tailored access to their data. The focus will be on turning these capabilities into a more integrated and sustained service, linking the twins more closely to enable the simulation of physically consistent storylines across scales, connecting large-scale climate conditions with high-resolution information on extremes events.

Work in phase three will also build on the AI and machine-learning developments from the first two phases. The focus will be on coupling machine-learning components for different parts of the Earth system to build a European AI Earth-system model, and on transforming the wealth of digital twin outputs into high-quality, AI-ready datasets that can feed into Europe’s AI Factories and enable public institutions, SMEs and start-ups to develop new AI-based applications for weather, climate and climate-sensitive impact sectors.