Turing Institute deploys AI to predict Greenland ice sheet tipping point

TL;DR:

• The Alan Turing Institute is part of the five-year GIANT project, a 17-partner international collaboration funded by ARIA to understand how Greenland’s melting glaciers could push the North Atlantic toward a climate tipping point.
• The Turing’s role centres on its DeepSensor AI toolkit, which fuses satellite, ship, and autonomous vehicle data to fill observational gaps and predict the risk of sudden ice loss.
• Researchers will head to Greenland this summer with the RRS Sir David Attenborough, deploying drones, autonomous underwater vehicles, and ice-embedded sensors to gather unprecedented data.

An international team of scientists is heading to Greenland this summer for a two-month expedition that will rely heavily on AI to understand whether the ice sheet’s accelerating melt is pushing the Atlantic Ocean toward a critical tipping point.

The fieldwork is part of GIANT (Greenland Ice sheet to AtlaNtic Tipping points), a five-year project led by the British Antarctic Survey and funded by the Advanced Research + Invention Agency (ARIA). The 17-partner collaboration aims to answer a question with global consequences: how much meltwater is entering the North Atlantic from Greenland’s 200-plus fjord glaciers, and what does that mean for ocean circulation patterns that regulate Europe’s climate?

AI fills the observational gaps

The Alan Turing Institute’s contribution focuses on building a prototype early warning system for rapid glacier change. The challenge is that Greenland’s narrow fjords have so far been impossible to represent in global climate models, and observational data is inevitably patchy.

The Turing’s DeepSensor toolkit uses AI to fuse data from multiple sources, including satellites, surface sensors, ships, and autonomous vehicles, creating a continuous picture from fragmented observations. The system helps determine optimal sensor placement and provides physics-aligned predictions of ice sheet behaviour where direct measurements are unavailable.

This data will feed into the UK Earth System Model (UKESM), which currently omits key processes at glacier-ocean interfaces. By the end of the project, researchers aim to incorporate fjord dynamics into the model, from ice cracking and calving to freshwater export into the ocean.

The expedition

The RRS Sir David Attenborough will transport researchers and equipment to Kangerlussuaq Fjord in south-east Greenland, serving as both a floating laboratory and launch platform. The technology deployed includes uncrewed surface vessels mapping glacier fronts with sonar, drones surveying from above, autonomous underwater vehicles collecting data at glacier faces, and sensors drilled directly into the ice.

A second study site at Petermann Glacier in north-west Greenland will provide contrasting data on how different glacier shapes respond to warming.

The project sits within the Turing’s broader Environmental Forecasting mission, which applies AI to weather, ocean, and sea ice prediction. For the UK specifically, the stakes extend beyond science: disruption to North Atlantic circulation patterns would directly affect European weather systems and accelerate global sea level rise. The early warning system the team is building may eventually give governments the lead time they need to prepare.