As glaciers melt, huge chunks of ice break free and splash into the sea, generating tsunami-size waves and leaving behind a powerful wake as they drift away. This process, called calving, is important for researchers to understand. But the front of a glacier is a dangerous place for data collection.

To solve this problem, a team of researchers from the ����Ӱ�Ӵ�ý and collaborating institutions used a fiber-optic cable to capture calving dynamics across the fjord of the Eqalorutsit Kangilliit Sermiat glacier in South Greenland. This allowed them to document — without getting too close — one of the key processes that is accelerating the rate of glacial mass loss and in turn, threatening the stability of ice sheets, with consequences for global ocean currents and local ecosystems.

“We took the fiber to a glacier, and we measured this crazy calving multiplier effect that we never could have seen with simpler technology,” said co-author, a UW assistant professor in Earth and space sciences. “It’s the kind of thing we’ve just never been able to quantify before.”

in Nature on Aug. 13.

The Greenland ice sheet — a frozen cap about three times bigger than Texas ­­­— is shrinking. Scientists have documented its retreat years as they scramble to understand the consequences of continued mass loss. If the Greenland ice sheet were to melt, it would release enough water to raise global sea levels by about 25 feet, inundating coastlines and displacing millions of people.

Researchers also speculate that ice loss is, a global current system that controls the climate and nutrient distribution by circulating water between northern and southern regions.

“Our whole Earth system depends, at least in part, on these ice sheets,” said lead author, a postdoctoral researcher in Earth and space sciences. “It’s a fragile system, and if you disturb it even just a little bit, it could collapse. We need to understand the turning points, and this requires deep, process-based knowledge of glacial mass loss.”

For the researchers, that meant taking a field trip to South Greenland — where the Greenland ice sheet meets the Atlantic Ocean — to deploy a fiber-optic cable. In the past decade, researchers have been exploring how these cables can be used for remote data collection through technology called Distributed Acoustic Sensing, or DAS, that records ground motion based on cable strain. Before this study, no one had attempted to record glacial calving with a submarine DAS cable.

“We didn’t know if this was going to work,” said Lipovsky. “But now we have data to support something that was just an idea before.”

Researchers dropped a 10-kilometer cable from the back of their boat near the mouth of the glacier. They connected it to a small receiver and collected ground motion data and temperature readings along the length of the cable for three weeks.

The backscatter pattern from photons passing through the cable gave researchers a window beneath the surface. They were able to make nuanced observations about the enormous chunks of ice speeding past their boat. Some of which, said Lipovsky, were the size of a stadium and humming along at 15 to 20 miles per hour.

Glaciers are huge, and most of their mass sits below the surface of the water. Mass loss proceeds faster underwater, eating away at the base and creating an unstable overhang. During a calving event, the overhanging portion breaks off and splashes into the sea. Gradual calving chips away at the glacier, but every so often, a large event occurs. During the experiment, the researchers witnessed a large event every few hours.

“Icebergs are breaking off and exciting all sorts of waves,” said.

Following the initial impact, surface waves — called calving-induced tsunamis — surged through the fjord. This stirs the upper water column, which is stratified. Seawater is warmer and heavier than glacial melt and thus settles at the bottom. But long after the splash, when the surface had stilled, researchers observed other waves, called internal gravity waves, propagating between density layers.

Although they were not visible from the surface, the researchers recorded internal waves as tall as skyscrapers rocking the fjord. The slower, more sustained motion created by these waves prolonged water mixing, bringing a steady supply of warmer water to the surface while driving cold water down to the fjord bottom.

�Ұ�ä�ڴ� compared this process to ice cubes melting in a warm drink. If you don’t stir the drink, a cool layer of water forms around the ice cube, insulating it from the warmer liquid. But if you stir, that layer is disrupted, and the ice melts much faster. In the fjord, researchers hypothesized that waves, from calving, were disrupting the boundary layer and speeding up underwater melt.

Researchers also observed disruptive internal gravity waves emanating from the icebergs as they moved across the fjord. This type of wave is not new, but documenting them at this scale is. Previous work relied on site specific measurements from ocean bottom sensors, which capture just a snapshot of the fjord, and temperature readings from vertical thermometers. The data could help improve forecasting models and support early warning systems for calving-induced tsunamis.

“There is a fiber-sensing revolution going on right now,” said Lipovsky. “It’s become much more accessible in the past decade, and we can use this technology in these amazing settings.”

Other authors include, a UW graduate student in Earth and space science; a UW postdoctoral researcher in Earth and space science,,, , , of University of Zurich; , ,, of ETH Zurich;,, and of the Université Côte d’Azur; and of GEOMAR | Helmholtz Centre for Ocean Research Kiel; of Tufts University;, of École Polytechnique Fédérale de Lausanne; of Stanford University; and of the Université Paris Cité.

This research was funded by the U.S. National Science Foundation, the ����Ӱ�Ӵ�ý’s FiberLab, the Murdock Charitable Trust, the Swiss Polar Institute, the University of Zurich, ETH Zurich, and the German Research Center for Geosciences GFZ.

For more information, contact Dominik �Ұ�ä�ڴ� at graeffd@uw.edu.

There’s enough water frozen in Greenland and Antarctic glaciers that if they melted, global seas would rise by many feet. What will happen to these glaciers over the coming decades is the biggest unknown in the future of rising seas, partly because glacier fracture physics is not yet fully understood.

A critical question is how warmer oceans might cause glaciers to break apart more quickly. ����Ӱ�Ӵ�ý researchers have demonstrated the fastest-known large-scale breakage along an Antarctic ice shelf. The , recently published in AGU Advances, shows that a 6.5-mile (10.5 kilometer) crack formed in 2012 on Pine Island Glacier — a retreating ice shelf that holds back the larger West Antarctic ice sheet — in about 5 and a half minutes. That means the rift opened at about 115 feet (35 meters) per second, or about 80 miles per hour.

“This is to our knowledge the fastest rift-opening event that’s ever been observed,” said lead author , who did the work as part of her doctoral research at the UW and Harvard University and is now a postdoctoral researcher at Stanford University. “This shows that under certain circumstances, an ice shelf can shatter. It tells us we need to look out for this type of behavior in the future, and it informs how we might go about describing these fractures in large-scale ice sheet models.”

A rift is a crack that passes all the way through the roughly 1,000 feet (300 meters) of floating ice for a typical Antarctic ice shelf. These cracks are the precursor to ice shelf calving, in which large chunks of ice break off a glacier and fall into the sea. Such events happen often at Pine Island Glacier — the iceberg observed in the study has long since separated from the continent.

“Ice shelves exert a really important stabilizing influence on the rest of the Antarctic ice sheet. If an ice shelf breaks up, the glacier ice behind really speeds up,” Olinger said. “This rifting process is essentially how Antarctic ice shelves calve large icebergs.”

In other parts of Antarctica, rifts often develop over months or years. But it can happen more quickly in a fast-evolving landscape like Pine Island Glacier, where researchers believe the West Antarctic Ice Sheet has already passed a tipping point on its collapse into the ocean.

Satellite images provide ongoing observations. But orbiting satellites pass by each point on Earth only every three days. What happens during those three days is harder to pin down, especially in the dangerous landscape of a fragile Antarctic ice shelf.

For the new study, the researchers combined tools to understand the rift’s formation. They used seismic data recorded by instruments placed on the ice shelf by other researchers in 2012 with radar observations from satellites.

Glacier ice acts like a solid on short timescales, but it’s more like a viscous liquid on long timescales.

“Is rift formation more like glass breaking or like Silly Putty being pulled apart? That was the question,” Olinger said. “Our calculations for this event show that it’s a lot more like glass breaking.”

If the ice were a simple brittle material, it should have shattered even faster, Olinger said. Further investigation pointed to the role of seawater. Seawater in the rifts holds the space open against the inward forces of the glacier. And since seawater has viscosity, surface tension and mass, it can’t just instantly fill the void. Instead, the pace at which seawater fills the opening crack helps slow the rift’s spread.

“Before we can improve the performance of large-scale ice sheet models and projections of future sea-level rise, we have to have a good, physics-based understanding of the many different processes that influence ice shelf stability,” Olinger said.

The research was funded by the National Science Foundation. Co-authors are and , both UW faculty members in Earth and space sciences who began advising the work while at Harvard University.

For more information, contact Olinger at solinger@stanford.edu.

Watch: Researchers Brad Lipovsky and Marine Denolle explain how fiber-optic cables can be used to sense ground motion. Credit: Kiyomi Taguchi/UW

The fiber-optic cables that travel underground, along the seafloor and into our homes have potential besides transmitting videos, emails and tweets. These signals can also record ground vibrations as small as a nanometer anywhere the cable touches the ground. This unintended use for fiber-optic cables was discovered decades ago and has had limited use in military and commercial applications.

A ����Ӱ�Ӵ�ý pilot project is exploring the use of fiber-optic sensing for seismology, glaciology, and even urban monitoring. Funded in part with a $473,000 grant from the M.J. Murdock Charitable Trust, a nonprofit based in Vancouver, Washington, the new UW Photonic Sensing Facility has three decoder machines, or “interrogators,” that use photons traveling through a fiber-optic cable to detect ground motions as small as 1 nanometer.

“Fiber-optic sensing is the biggest advance in ground-based geophysics since the field went digital in the 1970s,” said principal investigator , a UW assistant professor of Earth and space sciences. “The UW Photonic Sensing Facility and its partners will explore this technology’s potential across scientific fields — including seismology, glaciology, oceanography and monitoring hydrology and infrastructure.”

The new center — the largest in the United States and the first of its kind in the Pacific Northwest — is among a handful of research hubs around the world that are beginning to explore fiber optics for sensing ground motion. This approach to monitoring could expand the amount of seismic data by thousands of times.

For now, one of the three UW interrogator machines is hooked up to a “dark fiber,” or unused cable, that runs between the UW campuses in Seattle and Bothell. The researchers will soon also connect to a similar underwater cable across Alaska’s Cook Inlet to sense volcanic, oceanic, glacial and tectonic systems there. The other equipment will be used for temporary deployments.

When the ground vibrates — due to a heavy truck, construction work, or an earthquake — the seismic waves travel out from the source like ripples on a pond. When a seismic wave reaches the fiber-optic cable, the cable stretches very slightly, and that disrupts photons that are naturally reflected back to the source. The researchers can detect this disruption in the returning light waves and determine where the cable was disturbed.

The technique is known as “distributed acoustic sensing,” or DAS, because the system is spread out and can be used to monitor both sound waves and ground motion.

The same technology can also record more gradual motions. Lipovsky, who studies glaciers, and UW graduate student carried equipment up to Easton Glacier on Mount Baker to monitor the rate of surface melt. The team installed a cable and used an interrogator to see how much snow was melting on the glacier.

In other pilot projects, UW researchers with the Pacific Northwest Seismic Network are exploring uses for seismology, including earthquakes, volcanoes and landslides. UW oceanographers will use fiber-optic cables connecting to a seafloor observatory to monitor ocean faults and even eavesdrop on whales. UW civil engineers will study whether this technology could monitor traffic collisions or building and bridge infrastructure.

The facility will include semi-permanent observatories in Seattle and other unused “dark” fibers, including a cable that runs to Whidbey Island. The team also plans to lay cables for temporary field deployments at Mt. Rainier and is exploring projects farther afield at a fjord in Greenland and at McMurdo Station in Antarctica.

“We’re getting to the ‘smart Earth’ concept, where we can listen to the Earth,” said , a UW assistant professor of Earth and space sciences. “This technology allows seismic sensing to go to places you could not go before — where it was too hard, or too expensive, to deploy sensors. The other aspect that’s new is a density of sensors beyond what we had before.”

Today’s seismometers record ground motion at a single point, whereas fiber-optic cables take measurements at many points along the cable — the test cable has 15,000 data channels. Denolle will use computing and machine learning to make sense of this new mountain of seismic data.

“In seismology, our data used to be just wiggles,” Denolle said. “This is the first time we can get 2D images, and even videos, of data streaming in.”

The grant was awarded in late 2021. Researchers have used the funds to hook up and test the equipment last spring, and a data-visualization room on campus is coming soon.

“Thanks to the M.J. Murdock Charitable Trust’s support, the UW is the first university to acquire so much equipment for this technique,” Lipovsky said. “This is in the pilot experiment stage, and we are excited to see where it goes.”

Other funders are the UW and the UW-based Pacific Northwest Seismic Network.

For more information contact Lipovsky at bpl7@uw.edu and Denolle at mdenolle@uw.edu.

Glacier ice is usually thought of as brittle. You can drill a hole in an ice sheet, like into a rock, and glaciers crack and calve, leaving behind vertical ice cliffs.

But new ����Ӱ�Ӵ�ý research shows that glaciers are also slightly compressible, or squishy. This compression over the huge expanse of an ice sheet — like Antarctica or Greenland — makes the overall ice sheet more dense and lowers the surface by tens of feet compared to what would otherwise be expected, according to published Jan. 19 in the Journal of Glaciology.

“It’s like finding hidden ice,” said author , a UW assistant professor of Earth and space sciences. “In a sense, we discovered a big piece of missing ice that wasn’t accounted for correctly.”

Compression of the ice lowers the surface by up to 37 feet (11.3 meters) on the Antarctic ice sheet and by up to 19 feet (5.8 meters) on the Greenland ice sheet. Averaged across the entire Antarctic ice sheet, the surface is lower by 2.3 feet (0.7 meters), which represents 30,200 gigatons of additional ice. For Greenland, the surface of the ice sheet is lowered by an average of 2.6 feet (0.8 meters), which represents 3,000 gigatons of ice.

The mass of the ice sheet is only partly to blame: Since a glacier’s temperature increases with depth, thermal compression makes the colder ice, near the surface of the ice sheet, denser, squishing the ice almost as much as its weight.

Together, the combined effects of gravitational and thermal compression add about 0.2% to the total mass of the ice sheet. Though that sounds small, including this effect will help improve calculations of glacier changes over time — especially as the newest satellites can make precise measurements of glaciers’ elevation to monitor their responses to climate change.

“The long-term behavior of the ice is that it flows, and it also slides a bit. But at the same time, if you hit the ice with a hammer, it goes bing, bing, bing,” Lipovsky said. “On short timescales the glacier is a solid, and on long timescales it’s a fluid.”

Currently even the long-term climate models don’t account for the compression, which becomes a bigger effect for large ice sheets like in Antarctica and Greenland.

“In the long-term flow models, ice is always treated as incompressible. I think if you had really pressed people, and said, ‘There’s seismic pressure waves in glaciers, they must be compressible,’ they would have agreed. But it’s not something people have been thinking about,” Lipovsky said.

The additional water content probably doesn’t matter to future sea-level rise — the new results might add 8 inches (20 centimeters) to the projected 260 feet (80 meters) of sea level rise in the very unlikely event of all the planet’s glaciers melting, Lipovsky said.

But compressibility affects measurements of the difference in glacier elevation between winter, when they are weighted with fresh snow, and summer, when much of that snow has drained off. These seasonal measurements are used to monitor how the glacier is changing over time. The new study estimates that adding ice compressibility could eliminate about one-tenth of the error around these estimates, improving the monitoring of large ice sheets as they respond to climate change.

“Going forward, I hope this will become a correction that’s more commonly accounted for,” Lipovsky said.

For more information, contact Lipovsky at bpl7@uw.edu.