AI System Tracks Icebergs From Birth to Breakup, Revealing Hidden Climate Impact
Global, Sunday, 8 February 2026.
British Antarctic Survey scientists developed AI that automatically tracks icebergs throughout their entire lifecycle, creating detailed ‘family trees’ by linking fragments back to parent icebergs. This breakthrough fills a major gap in climate research, as melting icebergs release massive freshwater volumes that reshape ocean currents and affect global climate patterns. Previously, only large named icebergs could be tracked manually, leaving thousands of smaller fragments unmonitored. The system analyzes satellite imagery to solve a ‘digital jigsaw puzzle,’ matching geometric shapes of ice fragments. Data feeds into global climate models like NEMO, improving predictions as polar ice loss accelerates. The technology also enhances maritime safety by predicting iceberg movements in polar shipping routes.
Revolutionary Tracking Capabilities Transform Climate Science
The British Antarctic Survey announced on February 5, 2026, the development of an artificial intelligence system that represents a fundamental shift in how scientists monitor iceberg behavior [1][2]. Ben Evans, lead author and machine learning expert at the British Antarctic Survey, emphasized the significance of this breakthrough: “What’s exciting is that this finally gives us the observations we’ve been missing. We’ve gone from tracking a few famous icebergs to building full family trees. For the first time, we can see where each fragment came from, where it goes and why that matters for the climate” [1][2]. The system addresses a critical limitation in climate research, as scientists previously could only track the largest named icebergs by visually following them one by one via satellites, leaving the fate and impact of thousands of smaller fragments largely unknown [2].
Digital Jigsaw Puzzle Solves Complex Ice Fragment Tracking
The AI system operates by analyzing distinctive geometric shapes of icebergs captured in satellite imagery, essentially performing what researchers describe as a digital “jigsaw puzzle” [2]. When an iceberg fractures, the technology pieces together which fragments once belonged together, linking scattered “child” fragments back to their original “parent” iceberg [2]. This capability represents something that has not previously been possible at scale [2]. The system was rigorously tested on real satellite observations of Arctic icebergs calved from Petermann Glacier and other parts of north-west Greenland [1][2]. The technology identifies individual icebergs, assigns each a unique identity, and follows them over time throughout their entire lifecycle [2].
Integration with Global Climate Models Enhances Predictions
The breakthrough provides vital new information that feeds directly into global climate and ocean models, with data specifically incorporated into the NEMO ocean model, which forms part of the UK Earth System Model [1][2]. This integration is particularly crucial as ice loss from polar ice sheets is expected to increase as the climate warms [2]. The system reveals where iceberg meltwater is released into the ocean, information that can reshape understanding of ocean currents, marine ecosystems, and global climate patterns [2][7]. According to British scientists, this AI tool could fill a “major blind spot” in predicting climate change, as icebergs release enormous volumes of freshwater when they melt in open waters [7].
Maritime Safety Applications and Future Implications
Beyond climate research, the technology offers significant practical applications for maritime safety in polar regions [1][2]. Navigation routes could be adjusted using the AI to reduce risk by understanding where hazardous ice originates and how it is likely to move [1]. Ships operating in polar waters can better understand iceberg movements and plan safer routes through treacherous polar regions littered with icebergs [7][9]. The research was funded through multiple sources, including the UK Engineering and Physical Sciences Research Council, The Alan Turing Institute, and the Polar Science for a Sustainable Planet programme at the British Antarctic Survey [1][2]. This comprehensive funding demonstrates the recognized importance of this technological advancement in addressing both scientific and practical challenges posed by changing polar ice conditions.