× Dismiss

Never Miss an Update.

Oceans Deeply is designed to help you understand the complex web of environmental, social and economic issues facing the world’s oceans. Our editors and expert contributors work to bring you greater clarity and comprehensive coverage of ocean health.

Sign up for our newsletter to receive weekly updates, special reports and featured insights as we cover some of the most critical issues of our time.

Robots Dive Below Antarctica Ice Sheet to Collect Climate Change Data

Scientists know little about how the West Antarctica Ice Sheet is melting from below, but a mission to deploy autonomous vehicles in the ocean underneath the glacier could provide crucial answers.

Written by Matthew O. Berger Published on Read time Approx. 5 minutes
Antarctica float1
Scientists will deploy four autonomous floats under the West Antarctica Ice Sheet to collect data on melting sea ice.Paul G. Allen Philanthropies

A few days after Santa returns to the North Pole, robots will head to Antarctica.

A network of three autonomous gliders and four autonomous sensor-equipped floats is set to be deployed beneath a portion of the West Antarctic Ice Sheet, a massive frozen shelf extending into the ocean that scientists suspect could be melting due to climate change. If it thaws, the inland ice it holds back could eventually flow into the sea, triggering a catastrophic global sea-level rise of several feet.

The bottom of the glacier is likely melting the fastest, where the warmer saltwater meets the ice in pitch-black caverns in the underbelly of the ice shelf. Deploying sensors there to understand the icy landscape, how it’s changing and how quickly those changes are happening below an ice sheet that is 1,000–1,600ft (300– 500m) thick has been mostly impossible due to the inhospitable conditions.

But a new generation of robotic research instruments – and new funders willing to risk losing that pricey hardware – may open a new age of underwater exploration that could advance critical research into the melting of Antarctica ice sheets.

A few days after Christmas – midsummer in the Southern Hemisphere – the ship carrying the robots developed by University of Washington researchers in partnership with Microsoft cofounder Paul G. Allen’s philanthropic foundation, will set sail from Littleton, New Zealand. About two weeks later, the vessel will reach the Amundsen Sea, where several glaciers extend into the ocean off Antarctica’s west coast, and researchers will deploy the gliders and floats. When the ship leaves Antarctica, it will leave the robots behind to continue to collect data on water temperature, salinity and current direction and speed.

This graphic shows how self-driving Seagliders and floats will track conditions below an Antarctic ice shelf. Inside these caves, warmer saltwater flows in on the bottom, carrying heat that may eat away at the ice, and fresher glacial meltwater flows out above. (University of Washington)

The instruments consist of floats that can adjust their buoyancy to move deeper or higher in the water column and Seagliders that can propel themselves like small torpedoes. The floats have a five-year lifespan, while the gliders can last a year in the ocean. The researchers built both, but if bought commercially the floats would run about $25,000 and the gliders about $200,000 each, said Craig Lee, a University of Washington oceanographer working on the mission. He noted that the far bigger cost is the daily operation of the ship during the mission.

The floats will be sent under the ice shelf and will be carried with the current until they reemerge. At first, the gliders will go a little ways under the ice and come back to open water to surface and transmit data via satellite. Then they’ll go a bit further before surfacing again. Then a bit further. Summer will turn to winter, and robots will be left to fend for themselves as the ice cover expands.

“If we’re lucky and still know where they are and can have a ship that’s capable of retrieving them, we’ll hopefully retrieve them between January and March [of 2019],” said Knut Christianson, a professor of Earth and space science at the University of Washington.

But retrieving the hardware isn’t the important part. What matters is showing that gliders and floats can gather useful data and survive long enough to deliver it to researchers, filling in the blanks in their models.

Christianson and Lee called the effort high risk, high reward. If successful, it could fill in the blanks about what’s happening in potentially one of the most significant places where climate change is unfolding. “For these glaciers, we know they’ve been thinning and retreating for at least the last 20 years and likely a lot longer than that, and the reason for that change has been warm water,” Christianson said.

Researchers know the water around Antarctica has warmed, but measurements of the water and ice melt under the floating part of the glacier are lacking. Even knowledge of the topography of the bottom of the ice shelf and the seafloor below it, which can influence water temperature, flow and ice melt, is limited.

That underwater thawing would accelerate melting on the broader ice sheet. “We have large floating tongues of ice that essentially buttress the ice behind the tongue,” said Lee. “If you take away that big floating tongue, you take away the buttress.” The latter is what keeps inland ice from flowing out to sea, where it will continue to melt once in contact with the relatively warmer saltwater.

Unfortunately, he said, “our ability to predict how that works is limited because of lack of observation. We don’t know as much as we need to know about the dynamics of this ice sheet.”

Understanding the rate at which the Antarctic ice sheet is disappearing is one of the best ways to firm up predictions about sea-level rise, Lee said.

University of Washington researchers lower a Seaglider into Puget Sound for an early November test. Three custom Seagliders will travel in December to explore the water below an Antarctic ice shelf. (Paul G. Allen Philanthropies)

“It could very well be that we don’t get these instruments back,” Christianson said, adding that they don’t even know whether they’ll get results back from the robots. “If we do get results, they could be very important, but we don’t know that we will.”

It’s more of an exploratory mission, said Lee. He was cautious about even calling attention to the project before it’s accomplished anything. “We’re trying not to say we’ve done anything astounding because we haven’t actually done anything yet.”

He says the scientists should know whether they’ve succeeded within a month, once data about what’s happening beneath the ice slab starts being transmitted back – or not.

But simply having the opportunity to send expensive equipment somewhere from which it may not return may be an accomplishment in itself. Most of what scientists know about the ice shelf comes from dropping instruments down holes drilled 0.6 mile (1km) or more through the ice. Those methods are constrained both by space (the instruments are tethered and can only travel a little way from the bottom of the borehole) and time (before the borehole ices over).

There have also been a couple of earlier expeditions that sent autonomous vehicles below the ice, most notably a British initiative that deployed the 23ft (7m)-long, nearly 4-ton Autosub 2, which was lost under the ice in 2005, as well as its replacement, Autosub 3.

But the ability to dispatch small and maneuverable robots more frequently, at lower costs and for longer durations than the Autosub would be a huge boon to climate and marine science

“The holy grail of sub-ice shelf science is putting gliders and other [robots] down there,” said Matthew Siegfried, a Stanford University glaciologist who studies the Antarctic ice sheet, the polar ice cap that covers the continent.

But it’s clear that this sort of research is too risky for the usual funders, namely the government. “Science has limited resources through the normal means of government funding,” said Siegfried, who isn’t involved in the University of Washington project, “so it’s just incredible that the philanthropic community is coming through and taking on some of that risk.”

A previous version of this article incorrectly stated the year the robots will be retrieved. It is 2019, not 2018. 

Become a Contributor.

Have a story idea? Interested in adding your voice to our growing community?

Learn more