• Physics 18, 156
Experiments on freezing saltwater have teased apart flow dynamics inside ice pores, offering a possible boost to climate models’ predictive power.
Tapio Haaja/unsplash.com
Arctic sea ice plays a major role in regulating our planet’s climate through its reflection of the Sun’s rays, but today’s climate models vary wildly in their predictions of sea-ice decline. Researchers have now captured a physical process that’s been missing in these models: how the material structure of sea ice evolves as it ages [1]. Their laboratory experiments show that salt expelled from young ice decreases its porosity over time, which, in turn, affects fluid flow, heat transfer, and other processes that influence the melting of the ice.
The predictive power of a global climate model depends on how accurately it represents small-scale processes that can affect the larger atmosphere and ocean. Changes in sea ice represent one such small-scale input. Floating patches of ice reflect sunlight, so less of the Sun’s energy gets absorbed by the ocean. But in a vicious cycle, melting ice exposes more ocean, which in turn absorbs more heat, causing more melting. The rate of melting depends on several factors that include salinity and porosity. But climate models lack a mechanistic description of how these processes couple with each other. “That’s why it’s so important to understand in detail how ice interacts with saline water flows,” says Feng Wang of Tsinghua University in China.
Sea ice is known to lose its salt content as it freezes, resulting in denser, saltier water occupying small pockets (or pores) in the young ice—a process known as brine rejection. Laboratory studies have explored the effects of brine rejection on the ice structure and the formation of pores, but these earlier experiments have typically run for less than a week. Wang and his colleagues designed an experiment that persisted for nearly a month and focused on changes in salinity. “This unprecedented duration allows us to resolve the salt-diffusion timescale that controls the slow evolution of ice porosity,” says Wang.
The researchers started with a 24 × 12 × 6-cm3 rectangular tank filled with a saline solution whose salinity was close to that of seawater. The tank featured thermally controlled side walls: The left wall was maintained at a below-freezing temperature, while the right wall was held at a constant above-freezing temperature. This configuration immediately established a convective current that transported warm fluid toward the cold wall, where an ice layer developed. Cameras and salinity probes continuously monitored the freezing process.
By correlating synchronized measurements of ice morphology, salinity, and porosity, the researchers reconstructed the complete lifecycle of the ice layer. After the relatively fast phase of ice growth—which took place in the first three days—Wang and his colleagues expected the system to stabilize and “just sit there.” But over the following two weeks, the team observed that the ice layer changed shape while maintaining the same average thickness. The ice also became noticeably more transparent—less porous—as the salinity of the surrounding liquid increased, leading the researchers to suspect that the ice evolution depends on desalination.
To better understand this evolution, the researchers first modeled the motion of a single brine-filled pore within the ice. Because of a temperature gradient in the ice, the pore is colder on one side than the other. The brine freezes on the colder side, increasing the salinity near the freezing front. Salt then diffuses toward the warmer side, causing melting there. This drives the pore to migrate slowly toward the warmer region, eventually reaching the ice edge where the pore’s brine is dumped into the surrounding water. Pore migration via diffusion, the researchers say, is the main mechanism behind sea-ice aging. To further investigate the final morphology of the ice, the team performed computer simulations, which showed that the ice in the tank eventually behaves as a dense ice layer, free of brine-filled pores.
Diego Perissutti, a physicist who specializes in multiphase fluid flow at the Technical University of Vienna, says that the work “fills a research gap” in understanding how salinity affects the long-term evolution of ice microstructures. He is impressed by the simplicity of the model and its capability to “capture the key features of the system with good accuracy,” which could be useful for large-scale climate models. Less-porous ice should be slower to melt, Perissutti says, but it’s too soon to say what implications the results might have for sea-ice and climate predictions.
While sea-ice porosity may decline on long timescales, the natural environment is influenced by many additional processes, Wang says. He adds that future work should focus on numerical models capable of modeling the coupled dynamics between flow inside porous ice and the convective fluid beneath it.
–Rachel Berkowitz
Rachel Berkowitz is a Corresponding Editor for Physics Magazine based in Vancouver, Canada.
References
- Y. Du et al., “Sea ice aging by diffusion-driven desalination,” Phys. Rev. Lett. 135, 104201 (2025).
