The Arctic Ocean is full of sounds—sea ice cracking, beluga whales calling, and ships moving through newly opened waters.
Now, researchers at MIT Lincoln Laboratory are building a low-cost sensor network that could turn this underwater cacophony into a way of monitoring a rapidly changing Arctic—without requiring people to venture onto the ice constantly.
The challenge is that these sounds can be difficult to separate. When sea ice fractures, the movement releases vibrations that travel through the ice, creating distinctive acoustic and seismic signatures, sometimes called icequakes.
Those signals can overlap with sounds from marine animals or human activity. The researchers’ 2024 measurements even captured marine-mammal songs. Their goal is to understand how these different signals travel through ice, identify their characteristic signatures, and eventually distinguish one source from another.
“We’re interested in things that make sound underneath the ice. For example, our OIC 2024 data contained marine-mammal songs. We need a better understanding of how such signals propagate through ice and how to distinguish these signals from other sources,” Ben Evans, one of the researchers and associate technical staff at MIT Lincoln Laboratory, said.
Listening through the ice
Evans and his colleague David Whelihan have been working at the Navy’s Operation Ice Camp (OIC) since 2022, developing a network of inexpensive, commercially available sensors that could continuously monitor the Arctic.
During OIC 2024, they deployed integrated sensor nodes. This March, they returned with a higher-fidelity geophone, an instrument capable of detecting vibrations in the sea ice. However, getting equipment onto Arctic ice proved almost as difficult as designing it.
Back-to-back blizzards delayed the team’s arrival by a week. Temperatures repeatedly dropped to −25°F, while winds reached 25–30 mph, with gusts up to 40 mph. There were days when “we had to put on hats and sometimes goggles just to go between tents, whereas at previous OIC events we walked around with long johns and pants,” Evans added.
Amid the challenging conditions, the team eventually transported the sensors by sled to a remote location. However, after deploying only about one-quarter of their planned equipment, worsening conditions forced them to turn back before drifting snow erased their tracks.
They managed to retrieve the sensors about a week later during a brief period of clear weather.
The weather also complicated a second experiment involving a communications modem developed by Norwegian defense startup Havguard. Unlike conventional radio signals, which seawater quickly weakens, the system uses magnetic fields to communicate through ice.
After testing the equipment in Vermont, researchers took it to an Arctic lagoon near Utqiaġvik, Alaska. They drilled a 2-by-3-foot hole through 3.6 feet of ice and lowered a remotely operated vehicle carrying the modem beneath it.
Using a sonar system, a Doppler velocity logger and drone imagery, they tracked the vehicle while measuring the modem’s performance. The prototype achieved through-ice data transmission of about 1.2 kilobytes per second—a modest rate, but one the researchers consider encouraging.
Fewer boots on Arctic ice will be better
The two technologies could eventually work together. Sensors could detect and record ice vibrations, animal calls and other activity, while through-ice communications could allow researchers to retrieve data without repeatedly traveling to dangerous locations.
This is important as shrinking sea ice opens Arctic waters to increasing military and commercial activity. Distinctive acoustic signatures from cracking ice could eventually help scientists monitor environmental changes, support coastal communities, and improve understanding of activity across the region.
By combining these signatures with other measurements and machine-learning techniques, researchers hope to distinguish icequakes from marine-mammal vocalizations and eventually improve automated monitoring of ice-fracturing events—potentially turning the Arctic’s vibrations into an early-warning signal for changes in the ice.
The researchers also hope to use machine learning to distinguish icequakes from marine-mammal vocalizations. Before OIC 2028, they plan to develop and test air-droppable versions of their sensors and improve the modem’s packaging and integration with their sensor network.
The prototype still needs further development, and the current data rate is limited, but the broader objective is clear, which is to gather more information from the Arctic while putting fewer people in harm’s way.
“The through-line in all this work is minimizing boots on the ice. Especially this year, we learned that the weather is in charge of our access to the Arctic,” Dave Whelihan, one of the researchers and an undersea systems engineer at Lincoln Lab, said.
“We need ways to easily get sensors where we want them and to retrieve the data they collect, even in these extremely challenging conditions,” Whelihan concluded.