Scientists have pinpointed the key factors that allow pockets of warm seawater to flow beneath the Antarctic ice shelves, melting the ice from below and destabilizing glaciers inland.
The research, led by the University of Cambridge, used a remarkably detailed model simulation to show how seasonal changes in wind strength and sea ice, together with contours on the seafloor, influence the movements of deep ocean water around Antarctic ice shelves.
"With our high-resolution model, we can see that seasonality is crucial for ocean-ice shelf interactions, and show that it must be considered when predicting the fate of Antarctic ice shelves," said lead author Josh Lanham from Cambridge’s Department of Earth Sciences, which was published in the journal Nature Communications Earth & Environment.
Ice shelves don’t directly contribute to sea level rise when they melt, because they are floating and displacing seawater, but they do protect vulnerable ice inland by preventing it from flowing out to sea.
The Antarctic Ice Sheet contains sufficient freshwater to raise sea levels by about 58 metres. Although it is anticipated that ice loss from Antarctica will be a major factor in sea level rise by 2100, the extent of its contribution in future climate projections is unknown.
A major uncertainty arises from the complex ocean processes that deliver heat to the ice sheets—a primary cause of ice melt—something that is both difficult to directly observe and replicate within model simulations.
Scientists think that Antarctic melting is primarily caused by warm, salty water penetrating beneath the floating ice shelves and melting them from beneath. Known as ‘circumpolar deep water’, this warm water can detach in pockets and flow under the ice shelves.
However, the reasons behind these warm water intrusions—along with their regional variations, and their changing frequency and duration—remain unknown, adding uncertainty to forecasts of ice loss and sea level rise driven by climate change
The new research indicates that model forecasts, including those used by the IPCC, do not fully account for the small-scale and seasonal processes that trigger warm water intrusions because they rely on ocean data averaged over large regions and throughout the year.
“Previous models haven’t been able to resolve important ocean processes occurring on a scale of just a few kilometres, such as the formation of turbulent, swirling eddies and features on the seabed, and yet our model shows just how much impact they have on ice shelves,” said Cambridge Earth Science’s Professor Ali Mashayek, senior author of the study.
The new model examined currents in the Southern Ocean in detail, splitting the Earth’s surface into grid cells as small as 1 km by 1 km. “This type of work is made possible because of the assimilation of a wide range of data and advances in computational power,” said Mashayek.
Into the model, the researchers incorporated ocean observations made by satellites, as well as data collected from the ocean by robotic instruments that drift with the currents. They also used ocean temperature and salinity data acquired by seals tagged with sensors as part of an international decade-long program. Because seals can swim long distances and dive several kilometres deep, they can gather observational data from areas that would usually be inaccessible. Finally, the team tested their model by checking it could replicate real observations made by the seals at specific locations near the shelf.
The researchers found that during certain wind regimes, cold surface currents descend to the continental shelf, blocking warm water from reaching the ice shelf. “That blob of cold water effectively insulates the ice shelf from heat offshore,” said Lanham. In contrast, changes in wind can weaken this barrier, allowing warm water in.
Sea ice growth, which varies by location, was also key. When seawater freezes, it expels salt, making the surface ocean denser and causing it to sink. This dense water can sit on the shelf and block heat offshore, they found.
Sea floor contours can further accentuate the influence of seasonal changes in wind and sea ice, said Lanham. “Features like canyons can funnel the warm water so it has a short cut to the shelf, whereas areas of raised topography can have a blocking effect,” he explained.
The researchers say their findings indicate that global climate models might struggle to represent these detailed ocean interactions, and call for models that can fully capture the seasonality of warm water intrusions. “To predict the fate of Antarctic ice shelves, we need to consider not only the atmospheric heat content and the oceanic transfer of anthropogenic heat, but also the high-frequency variability of winds and the complex interactions of winds, bathymetry, and oceanic currents,” said Mashayek.
The researchers now plan to test their model further with observational data. Recent surveys of the waters surrounding the Antarctic Peninsula in the Bellingshausen Sea, Antarctica, could provide those much-needed observations. Lanham will work on this new data, together with researchers from Cambridge’s Department of Applied Mathematics and Theoretical Physics and the British Antarctic Survey, later this year.
Reference: Joshua Lanham, Matthew Mazloff, Alberto Naveira Garabato, Martin Siegert, and Ali Mashayek, Nature Communications Earth & Environment (2025), Seasonal regimes of warm Circumpolar Deep Water intrusion toward Antarctic ice shelves.
Feature image: The Bellingshausen ice shelf, photo credit: Laura Cimoli