Tiny saltwater channels have a big influence on sea ice.
Sea ice typically includes pockets or channels of brine that allow salt water to flow vertically through the ice. When those channels align neatly, they need to make up only about 5% of the ice volume before the water can flow. But in more disordered, granular ice, salt water starts to flow only when the brine channels take up more space—roughly 10% of the ice volume, according to a new study published in Scientific Reports.
“If we’re trying to find predictive models about how these ice cores are responding under climate change, it’s going to be necessary to take into account these structural and microstructural conditions.”
This higher threshold could slow the drainage of surface melt ponds, as well as the transport of nutrients to microbial communities inside the ice.
“If we’re trying to find predictive models about how these ice cores are responding under climate change, it’s going to be necessary to take into account these structural and microstructural conditions,” said Stephen Ackley, a sea ice researcher at the University of Texas at San Antonio who was not involved in the study.
Disorderly Constructs
As seawater freezes, it forms a mixture of ice crystals and brine. In calm conditions, the ice slowly grows into long, parallel crystals separated by orderly brine channels. This columnar sea ice is common in the Arctic, and its properties have been widely used in sea ice models.
But in choppy waves or when the ice’s snow-covered surface floods and refreezes, new ice can’t grow into these ordered columns. Instead, it forms small, randomly oriented grains separated by more complex pores containing brine and gases. Called granular ice, this form is more common in Antarctica but is becoming increasingly prevalent in the Arctic as temperatures rise and ice cover thins.
“It’s the sequel we’ve been waiting decades for.”
In 1998, University of Utah mathematician Kenneth Golden established the first estimate of the point at which the brine channels are connected enough to allow water to flow in columnar ice, called the percolation threshold. The new work, also led by Golden, extends a similar analysis to granular sea ice.
“It’s the sequel we’ve been waiting decades for,” said Don Perovich, a sea ice researcher at Dartmouth who was not involved in the new work.
To quantify the percolation threshold for granular ice, Golden and his colleagues collected sea ice samples during two expeditions off the eastern coast of Antarctica in 2007 and 2012. They measured how quickly water moved through the brine channels in the ice. After the 2012 expedition, they also mapped the arrangement of ice crystals within the ice blocks to correlate those permeability measurements with the microscale structure of the ice.
The finding that in granular ice, about twice as much of the ice volume needs to be brine for water to flow compared to columnar ice suggests that brine channels within granular ice are much less interconnected.
With the higher threshold, “you have to reassess all these models, anything that relies on fluid flow through sea ice,” if granular ice is present, said Golden. Granular ice will require warmer or saltier conditions to leave enough brine in the ice structure to meet the percolation threshold and allow water to flow vertically.
For example, the new value could influence models of how meltwater ponds behave atop an underlying ice sheet. If meltwater ponds form above a base of granular sea ice, those ponds will require warmer temperatures before they start draining than melt ponds on columnar ice will.
If these melt ponds remain on the surface longer waiting for those warmer temperatures, they could lower the albedo, or reflectivity, of the ice sheet. That could cause the ice sheet to absorb more heat, leading to a feedback loop that could accelerate melting.
The higher percolation threshold could also affect algae that lives within the ice. Ice algae make up an important food source for krill and crustaceans, which in turn become food for fish, penguins, and whales. Algae rely on water flowing through the ice to deliver nutrients. Because granular ice requires warmer temperatures for that flow to start, it could affect the depth at which algae can live inside the ice, Golden said.
Percolation Consideration
Still, experts say more data are needed to establish percolation thresholds across both Arctic and Antarctic ice. The size of the grains in granular ice can vary substantially at different temperatures, under different formation conditions, and between the poles. Larger grains could lower the percolation threshold, allowing water to flow even when the ice contains much less than 10% brine by volume, said Sønke Maus, a scientist studying ice microstructure at the Norwegian University of Science and Technology who was not involved in the study.
“The data that we have at the moment for the granular sea ice is sparse,” Maus said. “You need a big campaign to collect such data.”
Golden said that in future work he also plans to develop models to compute the electromagnetic properties of both columnar and granular sea ice. Knowing these properties can help scientists determine the thickness and age of an ice sheet from satellite data.
—Skyler Ware (@skylerdware), Science Writer
Citation: Ware, S. (2026), Changes in sea ice microstructure could affect climate models, Eos, 107, https://doi.org/10.1029/2026EO260164. Published on 20 May 2026.
Text © 2026. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.
