Scientists to create a Digital Twin of Antarctica

The project has received further funding of €350K (total of €1.35m) from the European Space Agency (ESA) to deliver a digital twin of Antarctica demonstrator to advance their understanding of the Antarctic Ice Sheet’s supra and sub-glacial hydrology, its evolution, and its role within the broader ice sheet and ocean systems. The research is led by Noel Gourmelen from the School of GeoSciences, while Amos Storkey from Informatics will be guiding the use of AI in the project.

The project ties with the European commission’s plans to build a replica of the Earth in support of decision making as part of its Green Deal action plans and Digital Strategy.

The 4DAntarctica consortium is an international collaboration of scientists from the UK, Denmark, Germany, France, Italy, and Austria, funded by the ESA and led by Edinburgh to collect a dataset relative to the Hydrology of the Antarctic Ice Sheet.

The overall aim of the 4DAntarctica project is to advance our understanding of the Antarctic Ice Sheet’s supra and sub-glacial hydrology, its evolution, and its role within the broader ice sheet and ocean systems by building an unprecedented dataset using the latest satellite sensors and techniques. The Digital Twin of Antarctica demonstrator is an opportunity to pull together all these datasets into a dynamic and interactive reconstruction of the ice sheet. Earth Observations, Artificial Intelligence and numerical simulation, will be combined to improve the products, and to generate projection of the ice sheet evolution under future climate change scenarios. We are looking forward to work with ESA, the EC and various stakeholders in developing this Digital Twin of Antarctica.

Dr Noel Gourmelen

Ice sheets and Meltwater

Ice sheets are a key component of the Earth system, impacting on global sea level, ocean circulation and bio-geochemical processes. Their future evolution is the key uncertainty in projecting Sea Level for the future. Significant quantities of liquid water are being produced and transported at the ice sheet surface, base, and beneath its floating sections, creating feedback between the ice sheet, atmosphere and ocean systems. At the surface, atmospheric warming leads to melt and to the formation of networks of rivers and lakes. This process accounts for about 60% of ice loss from Greenland, and while in Antarctica the impacts of meltwater are proportionally much lower, its volume is largely unknown and projected to rise.

Artificial Intelligence and Machine Learning

A critical part of building a digital twin is working out how to map precisely what is on the ground from the data and measurements collected at each point in time. For example, previous work by the team on Greenland has demonstrated the ability to better infer the actual ice depth directly from satellite measurements by using machine learning methods to make the best use of the information. In Antartica there are many bigger challenges that also require careful inference, including understanding what is going on under the surface. Prof Amos Storkey, from the School of Informatics, will be part of the team and will guide the use of Artificial Intelligence within the project.

Creating a full understanding of Antarctica and its dynamics involves getting the most out of the complex information we have. That is what Machine Learning and Artificial Intelligence is all about: learning how to best piece together the current sources of knowledge to provide greater clarity about what we can know and good measures of what we don’t.

Professor Amos Storkey

Related Links

4DAntarctica

School of Geosciences

Noel Gourmelen

Amos Storkey