From Alaska to Antarctica, thousands of glaciers flow over the land and out to the ocean. These tidewater glaciers are rapidly retreating and melting, like much of Earth’s ice, continually adding to rising sea levels. But to date, scientists have struggled to pinpoint where on the face of a glacier’s terminus the most intense melting occurs—and exactly how fast it is happening—because of the difficulty and danger involved in getting close enough to these frozen behemoths.

Now, however, a team of researchers has figured out how to directly probe these melt processes and has tested the method out on one glacier in Alaska. What it found, published this week in Science, is worrying: the glacier is melting far faster than current theories had suggested “The melt rates that we measured were about 10 to 100 times larger than what theory predicted,” says lead study author David A. Sutherland, an oceanographer at the University of Oregon.

Unstable tidewater glaciers such as the one the team measured flow into the sea relatively quickly and calve, or spit off icebergs, often. Although these glaciers tower above the ocean, most of the melting action happens beneath the waterline. Meltwater on the glacier’s surface trickles down through cracks in the ice, creating a stream beneath the glacier that lubricates its flow over land and spills into the ocean. It stirs up the salty, comparatively warm seawater and pushes it up the glacier’s face, causing the ice to melt faster and ultimately calve. Rising temperatures increase surface melting; scientists expect more meltwater will trickle down as the climate warms, exacerbating this process and ultimately threatening coastal cities around the world.

Time-lapse video taken over eight hours from a ridge above LeConte Glacier shows the strong meltwater plume flowing away from the glacier, as well as icebergs calving and swirling in the near-glacier fjord. Credit: Jason Amundson University of Alaska Southeast

Getting accurate readings of glacier melt is critical to predicting future sea-level rise with any meaningful precision. “Computer simulations need to have accurate information about how ice sheets and oceans interact,” says glaciologist Twila Moon of the National Snow and Ice Data Center in Boulder, Colo. “If we're basing our projections of the future on theory that is not aligning with our observation, we're not going to do as good a job at projecting the future.”

But acquiring those data has required getting close to the glacier’s terminus, which could become deadly if an iceberg suddenly breaks off and sends large waves surging. “These are really difficult measurements to make because the end of a glacier, where it is contacting the ocean, is a really active and pretty dangerous environment,” says Moon, who helped place some of the team’s instruments but was not involved in any analysis of the data.

Sutherland and his team found a way around this problem during two field campaigns in August 2016 and May 2017, which took place near LeConte Glacier in the Alaska Panhandle’s LeConte Bay. They spent multiple days in a boat, bobbing in the iceberg-strewn waters, while instruments anchored to their craft collected data about water temperature, salinity and flow near the glacier’s terminus. Their innovation was to deploy a multibeam sonar instrument—similar to technology commercial fishers use to spot a prizewinning catch—to repeatedly map the topography of the glacier’s face from a safe distance. By tracking the glacier’s shape over time, they could see which parts of it were melting. When they combined that information with GPS data that tracked the glacier’s movement, they could chart the rate of meltwater flow.

Time-lapse video taken between March 31, 2016, and August 8, 2016, shows the flow of LeConte Glacier (from right to left in the picture). The glacier’s terminus retreats backward as summer progresses, even as ice flows rapidly toward the sea. Credit: Jason Amundson University of Alaska Southeast

The sonar data revealed seasonal differences, with higher melt rates in August than in May, as well as changes in where melt occurred. In May rates were higher on the top of the glacier; in August they were higher along the bottom, deep below the waterline. As the summer sun beat down, more meltwater flowed beneath the glacier and into the ocean, pushing the relatively warm seawater against the submerged face of the glacier and melting it. This process causes the glacier to become top-heavy and thus more likely to calve. It is a vicious cycle: as more ice cleaves into the sea, there is less friction to keep the glacier from sliding farther into the water—and it therefore melts faster. “The glacier starts speeding up because it’s like removing a plug in front of the glacier,” says glaciologist Eric Rignot of the University of California, Irvine, and NASA’s Jet Propulsion Laboratory.

Sutherland and his colleagues also noticed variations across the face of the glacier, which suggested melt rates were higher than they had expected in places besides where the meltwater flows into the ocean. This finding means factors other than meltwater discharge also influence how and where the glacier melts. The next step, Sutherland says, is to identify those mechanisms.

He suspects melt rates could actually be even higher than his team reports, because frequent calving changed the face of the glacier too quickly for the sonar to get accurate readings during some parts of the study.

Rignot, who was not involved in the new research, is not entirely convinced about the high rates of melting it found. Previous work he did to model ocean conditions in front of a glacier face at a high resolution returned much lower rates. He says he would like to see more detail in comparing these most recent data to those of previous models to sort out the differences.

Rignot does think the process used in the study could be very useful for getting a better handle on how fast glaciers are melting. “The methodology is fantastic. The results are fantastic,” he says. “This is the kind of data set that we need to make progress” in understanding interactions between ocean and ice along the face of a calving glacier.