The ice in the Arctic is in trouble, and a group of geoengineers wants to do something about it. Their initiative, called Real Ice, figured out a way to rebuild Arctic sea ice, on a very small scale, using little more than augers and water pumps. 

Their first attempts proved a success, according to the group’s 2024 to 25 research report, released in June. They were able to regrow approximately 30 centimeters of ice in addition to the ice that formed naturally without any help, or for a total of about 50 centimeters overall. 

What makes the results especially noteworthy, according to Real Ice, is that Cambridge Bay, Canada, the region of the Arctic where the scientists did their work, was losing about 6 centimeters of ice per decade on average. That means the additional 30 centimeters effectively reversed 50 years of ice thinning in the pockets where they were working. 

The ability to regrow Arctic sea ice, if it can be done at scale, would be an important step in trying to balance out the ecosystem-threatening effects of climate change that’s resulting from the burning of fossil fuels. The National Oceanic and Atmospheric Administration says that the Arctic has lost approximately 12% of its sea ice per decade since around 1980. Scientists are worried that all the sea ice could be gone as soon as the 2030s.

That would have huge impacts on wildlife habitats and human communities in the region — and globally as the change contributes to the warming of ocean waters.

Real Ice worked by drilling holes in the existing sea ice and then sending autonomous underwater vehicles equipped with water pumps down under the ice. Once in position, the pumps shot water up on top of the ice, flooding the vicinity. In the frigid temperatures, which bottomed out below minus 60 degrees Celsius, the water would almost immediately freeze, adding new layers.

The team pumped 30,000 cubic meters of water from below the ice to make the thicker ice.   

There was enough extra ice created that satellites could capture images of it, even after most of the rest of the ice in Cambridge Bay had melted.   

This was a small-scale experiment. Real Ice occupied a 1 square kilometer section of Cambridge Bay, a location selected for its stable ice and intensely cold winter. The group had nearby support from the Canadian High Arctic Research Station and the Ekaluktutiak Hunters and Trappers Organization.

The group’s long-term goal is to deploy these systems in more areas to reinforce thinning ice to help it better withstand increasingly warm summers and slow the rate at which Arctic sea ice is thinning. 

One wrinkle in the Real Ice team’s results was that the flooded regions had less snow packed on top at the end of the winter. Snow is an important element for protecting ice thanks to its high albedo, which is a measure of how much solar light and energy is reflected by any given material. Snow has one of the highest albedos, and it’s why full moons on snowy nights are so much brighter. 

However, the researchers noted that the ice they made appeared to have a higher albedo on its own, judging by aerial drone observations and satellite imagery. Real Ice CEO Andrea Ceccolini told The Guardian that one possible explanation was trapped air bubbles from the water being pumped out and frozen so quickly, which would cause the ice to be less clear than what freezes without this human intervention. 

Real Ice is now conducting its 2026 field research, and hoping to answer some of those questions.

The group did not respond to requests for additional comment.