At the bottom of the planet is the Southern Ocean, its waters cold and roiling and sheathed with ice many months of the year.

The edge of the ice cover, which melts during summer and forms again in winter, is called the marginal ice zone, and it is incredibly difficult to study. Large icebreaking ships, which have traditionally been used for research in the region, cannot consistently observe small-scale ocean activity. "It's a blind spot of knowledge in our climate system," said Sebastiaan Swart, an oceanographer at the University of Gothenburg in Sweden.

What is known about the marginal ice zone is that it is an important storage system for carbon and heat emitted by humans. The global ocean as a whole stores more than 90% of Earth's excess heat, and the Southern Ocean is the portal through which much of this heat is transferred from the atmosphere. This makes ignorance of the region particularly worrisome.

But Swart and Louise Biddle, a researcher also at Gothenburg, found a way around this methodological roadblock. To do so, they turned to organic instruments that can gather consistent information from under the ice: southern elephant seals.

Seals there have been monitored for decades. Small sensors and trackers that are attached to their bodies and the tops of their heads, like tiny hats, transmit information from dives — depth, lateral distance, water temperature, salinity — that gets filed into open-access databases. A typical southern elephant seal spends around 90% of its time underwater foraging for fish and squid, only surfacing for a couple of minutes between expeditions to catch its breath.

Because of the frequency of these dives, seal data can reveal small eddies and flows in the water. These water fluxes result from many of the same forces, including winds and heat gradients, that create large currents like the Gulf Stream, but are far smaller and called submesoscale flows. Some are only the length of a football field and last no more than a day.

As tiny as they are, submesoscale flows have a direct effect on what Swart calls the "window between the atmosphere and the whole ocean."

This window is known as the mixed layer, a sliver of water on the surface whose depth and stratification determines how much heat and carbon are absorbed by the ocean; the deeper and more well-mixed the layer, the wider the window opens and the easier it is for the ocean to absorb heat and carbon from the atmosphere.

Without the technology to peer under the ice cover, no one knew what kind of submesoscale flows were occurring in the marginal ice zone. Then the two researchers realized "that the seals had been going under the sea ice for years and years," Swart said. "And because they do that, they were collecting the right kind of observations for us to look at the upper ocean under sea ice." The open-access seal data sets could potentially illustrate what kind of submesoscale flows occur under the ice, and whether they occur at all.

What Biddle and Swart found, surprisingly, was that submesoscale flows are nearly as active under the ice as they are in the open ocean, and that they are strongest in the midwinter, when the ice is thickest.

"If these submesoscales are to change in the future, they actually will really change how much heat and carbon is stored in the atmosphere or in the ocean," Swart said. "And so they're really, really important, cumulatively, to the habitable planet."