The European Union, as part of a huge investment into digital infrastructure and supercomputing, has announced plans to build a ‘digital twin’ of the Earth. Destination Earth, or DestinE, will pull in an unprecedented amount of real-time data from climate, atmospheric, meteorological and behavioural sensors to construct a high-precision model of the planet.
By rendering the Earth’s atmosphere to a one-kilometre scale, DestinE’s supercomputing capacity will go far beyond today’s modelling infrastructure. Europe’s current climate forecasting model, led by ECMWF (the European Centre for Medium-Range Weather Forecasts) runs at the comparatively coarse scale of nine kilometres. ‘DestinE aims to develop a very high precision model of the Earth to anticipate, monitor, better understand and react to the climate change challenges ahead of us,’ says Johannes Bahrke, European Commission spokesperson for the digital economy. It’s hoped that this digital twin will give policymakers the computing power necessary to gauge how climate change will impact society and at the same time to visualise how the decisions we make as a society could alter the trajectory of climate change.
We can think of DestinE as the supercomputer behind the European Green Deal, which aims to make Europe the first climate-neutral continent by 2050. ‘This cannot happen without its twin objective: the digital transition. Europe needs ambitious projects to be up to this commitment,’ says Bahrke. Building a tool that can accurately anticipate possible futures will be a crystal ball for the Green Deal – building the computational power behind it, however, has been a key challenge.
An intercontinental race is currently underway to develop ‘exascale’ computing systems – the holy grail of computing power, capable of performing one billion calculations per second. The USA and China have traditionally been front-runners, but the EU has now entered the race. Building the necessary computational power to run DestinE is a key challenge for the EU, but with DestinE's announcement comes signs of progress. On 18 September 2020, it announced an €8 billion investment to build the next generation of supercomputers, assembled under the European High Performance Computing Joint Undertaking. Until then, it will make do with recent acquisitions of three advanced (but not quite exascale) systems totalling €830 million, which should be operational by 2021. When active, they will multiply the EU’s computing power by a factor of eight.
For climate scientists, the EU’s investment offers a way to make use of underutilised climate data. ‘We already collect huge amounts of data, but most is thrown away because existing climate and computational models aren’t powerful enough to digest it,’ says Björn Stevens, director of the Max Planck Institute for Meteorology. ‘There are many sources of data that we don’t even try to assimilate. The challenge is to bring this into models, and that’s where DestinE comes in. It can open up a new dimension of trying to find information in data that we already collect.’
The first step towards launch is to develop the core digital modelling platform, which is hoped to be operational by 2023. By 2027, the EU aims to have developed the full digital twin.
Once it’s up and running, DesintE should have a wide range of uses. For meteorologists, it could be used to anticipate natural disasters before they happen and with higher precision and granularity; climate scientists could use it to peer into the future of our climate; biogeographers could continually monitor the health of the planet by pulling in real-time biodiversity and land-use data.
It also offers new opportunities to integrate disciplines. Currently, climate scientists extract information from models and pass it to experts in agriculture, economics or sociology, who separately seek to understand the data’s relationship with human behaviour. DestinE will be capable of integrating these fields into a single platform – flood, drought and heatwave models can be paired with models of migration, economics and environmental pollution.
This integration of human and climate data is key to guiding the EU’s Green Deal. For years, governments have sought a way to assess the impact and efficiency of environmental policies. The European Commission says that DestinE will provide a solution. Proposed legislative measures at the EU and national levels will soon be simulated using DestinE; their effectiveness will be evaluated against alternative approaches, steering the commission’s path to net zero carbon.
Many challenges still remain. Even with the requisite computing power, digesting the resulting data into usable information won’t be easy. A new type of modelling code may have to be developed to deal with the scales of atmospheric, climate, behavioural and environmental data that DestinE will process. A Japanese team running a 1 km scale meteorological system took half a year to extract useful information from data collected over just a few days. Plus, the US is hot on the tail with their ‘Aurora’ exascale computing system. The US Department of Energy (DOE) have also made significant progress in a similarly ambitious project to DestinE called E3SM, which is aiming to advance the US’ own computing power through the goal of producing an exascale Earth system model.
DestinE differentiates itself from E3SM through its planned usability for the public. ‘Everyday’ public users, business and land owners will be able to access DestinE via the cloud, running the interface from desktop or tablet. They can use the interface to model their own climate and environmental interventions: A farmer might use DestinE to decide whether and where to install wind turbines, or to understand how they might feasibly transition to biofuels; schoolchildren might use it to model the air quality effects of their class cycling to school, rather than driving; many might simply use it as a gamification tool to help their understanding of climate change. Democratising the tools for climate modelling, Stevens thinks, will be key to achieving the scale up of climate-friendly practices that are too easily dismissed as ‘drops in the ocean’.
For Stevens, democratising climate modelling in this way is key to achieving the scale-up of climate-friendly practices that are too easily dismissed as ‘drops in the ocean’. ‘To me, the Anthropocene is the time period where individuals can influence the Earth at global scale,’ he says. ‘DestinE can give people the tools to see how local actions can scale to global effects.’