Foundations for an Earth Digital Twin

The European Union is developing a ‘digital twin’ of planet Earth that would simulate the atmosphere, ocean, ice and land with extreme detail, providing forecasts of floods, droughts, and fires from days to years in advance. The effort is called Destination Earth and will attempt to model human behavior to enable leaders to see the impacts of weather events and climate change on society and determine the effects of different climate policies.

By studying the planets atmosphere in boxes only one kilometer across, a scale many times finer than existing climate models, Destination Earth can base its forecasts on far more detailed real time data than ever before. The new models resolution will enable to directly render convection and vertical transport of heat critical to the formation of clouds and storms, rather than relying on algorithmic approximation.

The model will also include real-time data charting atmospheric pollution, crop growth, forest fires and other phenomena known to affect weather and climate.

Classical Earth system models are based on both well-known and lesser-known physical laws. With the help of mathematical and numerical models, the state of a system at a future time is calculated from what is known about the state of the system at a present or past time.

However, the price is that the increasingly complex ESMs require immense computational resources. Despite recent developments, even the predictions of the latest models contain uncertainties. For example, they tend to underestimate the frequency and strength of extreme events. Researchers are worried that abrupt changes could occur in certain subsystems of Earth, so-called tipping elements in the climate system, which classical modeling approaches cannot predict accurately. Many key processes, such as type of land use or the availability of water and nutrients, cannot be represented well. 

The challenges of classical ESM methods, but also the ever-increasing amounts of available Earth observations, open up the field for the use of artificial intelligence. The advantage is that these self-learning systems do not require knowledge of the possibly very complex and not even fully known physical laws and relationships. Instead, they are trained on large datasets and learn the underlying systematics themselves. This powerful and flexible concept can be extended to almost any desired complexity. For example, a neural network can be trained to recognize and classify patterns in satellite images, such as ocean eddies, cloud structures or crop quality.