A joint European initiative has set in motion the creation of the most advanced digital replica of Earth to date
Destination Earth (DestinE) is an EU-funded initiative to create a digital twin of the planet. It will leverage vast quantities of data, processed by the very latest Earth system models and developments in high-performance computing (HPC), to monitor and predict the interaction between environmental change and human activities with the aim of supporting future sustainable development.
To achieve this, the European Commission has joined forces with the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT), the European Centre for Medium-Range Weather Forecasts (ECMWF) and the European Space Agency (ESA). Contribution agreements were signed by each agency in December 2021, covering the first 30 months of a 7- to 10-year program. The initiative was officially launched on March 30, 2022.
DestinE will underpin effective decision making by helping policy makers monitor and simulate the Earth system and its evolution in view of human interventions. It will also potentially enable them to anticipate environmental disasters and resultant socioeconomic crises, and enable the development and testing of scenarios for more sustainable development where Earth system change information is crucial. Furthermore, the initiative will play a crucial role in helping Europe to become the world’s first climate-neutral continent by 2050 (Green Deal) by developing novel information pathways for the development of efficient climate adaptation strategies.
The DestinE initiative has been divided into three major components: the Data Lake, the Core Service Platform and the Digital Twin Engine supporting the first two digital twins of the Earth system. Advances in artificial intelligence (AI), machine learning, cloud computing, high-speed connectivity networks, big data handling and supercomputers, combined with the accumulated knowledge from numerical weather prediction (NWP) and the EU’s Copernicus Earth observation program, will all play a part in making the concept of a digital twin of Earth a reality.
The first phase of the initiative will focus on development and will run until June 14, 2024. It is being backed by €150m (US$152m) in funding from the Digital Europe Programme, with further support available from Horizon Europe – the EU’s key funding program for climate change research and innovation (which has a budget of €95.5bn [US$97bn]) – for research and innovation opportunities to support the development of DestinE. The Recovery and Resilience Fund is also available for the activities of member states linked with DestinE.
The Destination Earth Coordination Group will look to involve member states and associated countries. Independent scientific and technological expertise is being provided by a strategic advisory board. Public workshops are also being held to ensure that the system responds to real user needs.
High-quality information, digital services, models, scenarios, forecasts and visualizations will be provided first to public sector users and then gradually to scientific communities, the private sector and the general public.
Data Lake
DestinE will use a dedicated Data Lake, implemented by EUMETSAT, that will provide storage space and seamless access to various data spaces. It will be built on existing scientific and operational data sets, such as the Copernicus Data and Information Access Services (DIAS), and will be complemented by other non-spatial sources, such as sensor-based environmental data and socioeconomic data. Any new data originating from the digital twins will also be integrated, creating a coherent and self-standing data space. It will provide access to the data needed for the digital twins and the Core Service Platform operations and will host user data, shared with the DestinE user community while supporting near-data processing to maximize performance and service scalability.
Lothar Wolf, the Destination Earth program manager at EUMETSAT, describes the Data Lake as having three pillars. The first, he says, will make data accessible from other foreign data spaces. These data spaces, specified by the European Commission, include soil moisture deficit data, climate change data, socioeconomic data, seismic data, environmental data, atmospheric data, renewable energy data, marine data and economic data. This may come from satellites, citizen science, sensors and IoT systems, or even users’ own data.
For the second pillar, the DestinE Data Lake is required to store everything that has been defined as ‘Destination Earth data and information’. This formal definition, Wolf says, means that “you can focus rather on those data sets that are really necessary for further processing by the users or applications within the context of Destination Earth and that have passed the related quality controls and conform to the applied data policies, etc.”
Of course, modeling is not the only output where this can happen. “Every AI or machine learning or whatever other type of higher-level application that might come via the Core Service Platform or other applications that generate data on the basis of the output of the digital twins, their own data, and [be]classified as Destination Earth data and information, will also be stored by the Data Lake. You end up with this unique super melting pot of data that either stores data or makes it accessible,” says Wolf.
The third pillar will support data approximate computing to implement big data distributed workflows. “This is new,” enthuses Wolf. “It is the first time we will be able to support the so-called ‘what-if’ scenarios, make use of analysis-ready data, data cubes and fully distributed and cloud-native computation in such a wide context.”
Fed by real-world observations, the digital twins will enable users to understand extreme weather and climate-related events from the past and present, as well as anticipate what might happen in the decades ahead
Digital Twin Engine
ECMWF is creating the software and data environment needed to power digital twins of the Earth system. This environment is referred to as the Digital Twin Engine (DTE).
According to Dr Nils P Wedi, the digital technology lead for Destination Earth at ECMWF, and former head of the agency’s Earth system modeling section, the Digital Twin Engine will facilitate a radical improvement in the realism of global Earth system models.
“We see the DTE as providing a step change in the environment within which Earth system models operate and thus improving their quality and timeliness of information, particularly in the climate context. This includes a step change in resolution where we’re really thinking of one kilometer globally, or maybe four-kilometer grid-spacing, but not more than that. It will very much go from global to local scales.”
“We’re not starting from scratch, especially in NWP where we are already dealing with time-critical forecasting. We are building on components that have stood the test of time, particularly on the time-critical operational side of things, but [we]will now apply these to the pressing problem of climate change,” he says.
Of course, digital twins are nothing new and are already in use across several industries, such as the automotive and construction sectors. DestinE’s DTE will provide the foundation for creating digital replicas of the highly complex Earth system, and extend to the different domains of Earth science, such as extreme natural disasters, climate change adaptation, oceans and biodiversity. Ultimately, the vision is to integrate these digital replicas to form a comprehensive digital twin of the complete Earth system. This will enable users to revisit the past, understand the present and predict the future using a simulated and continuously evolving ‘living’ replica.
The initial focus of the DTE will be on the effects of climate change and extreme weather events, their socioeconomic impact and possible adaptation and mitigation strategies. Two initial twins will be created, both of which will integrate data on oceans, atmosphere, land, hydrology and sea ice.
Extreme Natural Disasters Digital Twin (weather-induced and geophysical extremes)
The natural disasters digital twin will help anticipate the occurrence and impacts of extreme natural events with increased precision. Decision makers will be able to assess how trustworthy their risk management strategies are likely to be. For example, when considering flood or drought management, authorities will be able to develop more effective local emergency management plans, prepare for disruptions to hydropower and effectively deploy drought-resilient crops.
“We’re trying to connect to the people that have to make decisions about the local climate, using applications that more readily integrate and more directly translate information from both weather and climate,” says Wedi. “We can provide a stream of ‘digital observation’ information from the digital twin, people can digest it as it happens, and can derive from it what they want to see and what they want to learn and apply it to their own particular problem.”
Climate Change Adaptation Digital Twin
Designed to develop trustworthy and reliable adaptation strategies to mitigate the effects of climate change, the Climate Change Adaptation Digital Twin will aim to reliably predict the impact of climate change at regional and national levels. In smart agriculture, for example, the twin can build on data analysis on projected rainfall and drought, and the correlation with possible plant diseases or infestations, and provide agriculture authorities with guidance on the quality of the underlying predictions, helping them to assess the trustworthiness of their measures.
“They would have this continuous stream of background information that they can use in their system to back up, compare and analyze what is happening on the ground against what the digital twins are providing,” adds Wedi.
“They could actually also give us their information or measurements and we could try and feed it in,” he continues. “We will provide machine learning tools that allow us to connect their data with our data. That’s part of the DTE – we want to provide the tools that help people to understand our data in light of their experience.”
However, there should be no requirement to share data, and private firms that wish to keep their own data space private while still using DestinE components will be free to do so.
“We’re not providing the system only for humans,” Wedi adds. “We are also thinking about a system that provides this flow of information to machines. Then, machines such as smartphones and other devices could just pick it up. I think this project will form the groundwork that’s required to make that ultimately happen and I think that’s where Destination Earth can make a difference.”
Core Service Platform
At the heart of DestinE will be a secure, user-friendly cloud-based digital modeling and open simulation platform, implemented by ESA. It will open DestinE to a wide range of stakeholders ranging from experts, scientists and policy makers to individual users.
Users will have cloud-based access to DestinE models, algorithms, applications and natural and socioeconomic data to exploit and test their own models. The overall system and its components will be user-friendly and flexible to adapt to a wide spectrum of needs and scenarios.
“It can be thought of as the entry point to the Destination Earth platform,” says Wedi. “It’s where the user registration happens, where you may get your local sandbox and where you can access what I would call the deeper services.”
The platform will also enable users to integrate their own data and develop their own applications. “DestinE is not only about the capacities and the abilities of running these high-resolution models on supercomputers; it’s also how you bring this information to the end user and how they can interact with the data and tools provided,” adds Wedi.
Future focus
All three entrusted entities have already begun publishing and processing tenders for acquiring various components to be integrated into DestinE. Conclusion of these tenders is set to run from the autumn through until the beginning of 2023.
For EUMETSAT, the first tender will procure the necessary engineering support services. “This is for the services that will help us on technical engineering questions that might arise during the execution of the main domain deployment,” says Wolf.
The agency’s second tender is for the end-to-end Data Lake service with all components, including infrastructure and software services. The final tender is for independent service validation and quality and KPI routine monitoring.
One service that ECMWF is looking to procure, as defined by one of the tenders, is the ability to demonstrate a visualization and rendering service – potentially including virtual reality and augmented reality – that will provide the user with a visual representation of digital twin data.
“We have asked for this in a service form, so that it could be applied and reapplied in a cloud platform,” says Wedi. “The idea is that you will be able to really fuse the digital world and the real world and bring in, say, observations on the fly, and be able to zoom in and (virtually) move around, like you do in computer games today. This is what we envisaged so hopefully we get some good responses that would allow us to at least demo some of this on the digital twin data that we’re going to produce.”
Concluding in 2027, the second phase will involve further enhancements to the DestinE system and anticipated integration of additional digital twins and related services.
“By the end of phase one, we want to end up with a prototype of an operational platform, digital services for some aspects, also including other aspects that are a bit more explorative. Phase two will surely concentrate more on the operational aspects, making it more robust and widening its user base,” says Wedi.
By 2030, the initiative aims to produce a ‘full’ digital replica of the Earth.
This article originally appeared in the September 2022 issue of Meteorological Technology International. To view the magazine in full, click here.
