An iceberg looms off the deck of a cruise ship near the Antarctic Peninsula in September 2005.  (Ann Manner / Getty)

A Radical New Scheme to Prevent Catastrophic Sea-Level Rise

A Princeton glaciologist says a set of mega-engineering projects may be able to stabilize the world’s most dangerous glaciers.

PRINCETON, N.J.—Geo-engineering, its most enthusiastic advocates will tell you, isn’t only possible. It’s already happening.

To hear more feature stories, see our full list or get the Audm iPhone app.

We know, they say, because we’re doing it—we just call it global warming. As humanity dumps billions of tons of greenhouse gases into the atmosphere every year, we’ve engineered a different climate system: one that is hotter, wetter, and more unwieldy than what people have lived in since the dawn of agriculture.

So far, the most promising—and least expensive—forms of reverse-engineering this change have taken a similarly whole-world approach. Researchers speculate that planes could periodically spray a clear gas into the high atmosphere that would prevent some sunlight from reaching Earth’s surface, cooling the globe in turn. There’s a lot of buzz about this idea, which is dubbed solar geo-engineering: More than 100 scientists discussed it at an off-the-record gathering in August; Harvard University has opened a $7.5-million center to study it.

But any downsides of this technology would be unpredictable. It might create winners and losers, cooling some regions while kicking off droughts in others. What if there was a more focused approach? What if scientists could prevent one catastrophic symptom of climate change—a rapid rise in global sea level, for instance—without messing again with the weather?

Michael Wolovick, a glaciology postdoc at Princeton University, believes it may be possible.

For the past two years, Wolovick has studied whether a set of targeted geo-engineering projects could hold off the worst sea-level rise for centuries, giving people time to adapt to climate change and possibly reverse it. He is exploring whether building underwater walls at the mouth of the world’s most unstable glaciers—huge piles of sand and stone, stretching for miles across the seafloor—would change how those glaciers respond to the warming ocean and atmosphere, dramatically slowing or reversing their collapse.

If they work as planned, these large walls could make glaciers last as much as 10 times longer than they otherwise would. In rudimentary simulations, the walls make a glacier that would collapse in 100 years last for another millennium.

Wolovick presented his work in December at the annual meeting of the American Geophysical Union, where I saw his work. We talked in the weeks after the conference.

“Part of the reason I’m putting this forward is it might be better to have a more targeted intervention. While the broad-brush solar geo-engineering is more planetary scale, the problems may be more planetary scale too,” he told me.

His proposal, which has not been previously described at length in the press, tries to work at the source of the problem. The glaciers in Greenland and Antarctica which will unleash the fastest sea-level rise are relatively contained right now. Engineering them should be different than trying to wrench a change into a tumultuous global weather system.

“Their geographic scale is smaller,” he said. “You get much more bang for your buck, in terms of how many societal impacts come from these specific ice streams and outlet glaciers.”

“We have to start thinking about how we can address the problem,” said Robin Bell, a professor of glaciology at Columbia University and the president-elect of the American Geophysical Union, a professional organization of more than 60,000 Earth scientists.

“As scientists, we can do individual action, and we also spend a lot of time trying to understand how the earth works,” Bell told me. She was Wolovick’s adviser during his doctoral study at Columbia, when they used radar to study how ice sheets bent and contorted as they moved across bedrock.

“At the same time I think he’s one of the very few people who’s said, okay, is there something we can do to slow how [ice] is going to change and how it will flow in the future?” she said. “This is kind of high-risk for young scientists, because everybody wants you to do what everyone else is doing. But somebody has to take those first steps.”

Though Wolovick has spent two years studying his proposal at Princeton, his ideas remain hypothetical. They will need years of further study before they become feasible. And even if his proposal seems to work, it will not reduce humanity’s need to reduce greenhouse-gas emissions. Slowing the rise of global sea levels will not alter other consequences of climate change, like deadly heatwaves, decade-long mega-droughts, or the widespread destruction of coral reefs.

It would just buy us some time to slow the rising seas. But for the more than 150 million people who live on land less than five feet above sea level, that may be enough.

Here’s a guide to Wolovick’s idea: how it works, the science behind it, and what other experts think.

Visitors look through an arch in a massive tabular iceberg off the coast of Antarctica in December 2016. (David Merron / Getty)

Wolovick’s plan requires the construction of what he calls “sills”: large, flat piles of material that sit on the seafloor. “It’s nothing particularly technologically advanced,” he said. “I’m imagining something like a big pile of sand or other loose aggregate, and maybe an outer layer of boulders to protect against tides.”

Simply constructing these large walls in front of the world’s most unstable glaciers, says Wolovick, might stop them from collapsing.

How? It seems counterintuitive. Wolovick’s sills wouldn’t rise above the ocean surface. They wouldn’t be sea walls or levees, like those that surround New Orleans today, meant to keep water hemmed into one place. They would just be changes to the underwater topography of the ocean floor.

Nonetheless, our current understanding of why the world’s largest ice sheets melt suggests that they have a good chance of working.

“The big vulnerability we’ve seen in the past two decades or so of high-quality Antarctic data isn’t so much warm air, but warm water,” said Wolovick.

The surface of the ocean near most ice sheets today is quite cold. But that frigid water is just the top layer, and it sits on a second layer of dense, warm water. As the oceans move, that warm water is upwelled out of the middle depths, onto Antarctica’s continental shelf, and toward the continent’s enormous glaciers that end in the sea.

When this warm water reaches the glacier, it sloshes around at the base of the glacier’s “grounding line,” which is what scientists call the place where the glacier’s icy front wall is exposed to the sea. Warm ocean water erodes and melts the exposed ice face at the grounding line. Glacier ice becomes ocean water, causing sea levels to rise and the glacier to recede.

This points to a key fact for would-be geo-engineers: As temperatures rise around the world, not all of the world’s giant ice masses will melt in the same way. Greenland, which holds the world’s second largest ice sheet, sits mostly above sea level, and it touches the ocean only in a select few spots. “The Greenland ice sheet sticks its nose down into the North Atlantic,” is how Bell puts it.

Right now, warm ocean water is eroding some of Greenland’s fastest-moving ice flows—including Jakobshavn, which produces more icebergs than any other glacier in the world. But Greenland is also nestled between Canada and Northern Europe, and it catches more pockets of warm air than its antipodal twin. About half of its annual mass loss is due to surface melting, which is what happens when the air above the ice sheet gets too hot for ice to remain.

The towering icebergs produced by Jakobshavn Glacier, one of the few large glaciers in Greenland to experience ocean-induced melting, meet the sea at the end of Ilulissat Icefjord, above. Wolovick imagines one of the first sills being built at this spot. (Robinson Meyer / The Atlantic)

By design, Wolovick’s geo-engineering proposal can only address ocean-driven melting. But that’s okay: Surface melting is steady but slow. Ocean-driven melting is fast and unpredictable, and it’s how the most cataclysmic sea-level rise would occur in the 21st century.

That’s because of Antarctica—and, specifically, the unique geography of the West Antarctica Ice Sheet, or WAIS.

In the late 1950s, as scientists first mapped the southernmost continent, they discovered that the ice sheet in West Antarctic differed from the one in Greenland. Unlike Greenland, which sat on bedrock above sea level, WAIS sat inside a kind of giant bowl in the Earth. Most of its bedrock was well below sea level. Odd physics seemed to keep the whole thing in place: “With a bed below sea level, the ice sheet is anchored to its bed only because it is too thick to float,” David Vaughan, the director of the British Antarctic Survey, explained in a recent paper.

Twenty years later, John Mercer, a glaciologist at Ohio State University, connected that unusual feature to the recent idea that humans were warming the globe with carbon-dioxide pollution. In 1978, he warned in Nature that warm ocean water and WAIS’s bowl of bedrock could interact catastrophically.

In any ocean-terminating glacier, warm seawater erodes and melts the grounding line, driving sea-level rise and retreat. But the bedrock of WAIS slopes toward the center of the continent—which means that the glacier stores most of its water close to its center, since it also is tallest at its center. These two facts combine into a hideous, runaway mechanism: For every foot the WAIS recedes, it introduces comparatively more water to the ocean than the previous foot did. At the same time, even as the glacier recedes, the tremendous weight of each glacial flow would still push it forward, toward the hungry ocean.

The glaciers that link WAIS to the sea won’t just go steadily over time. They’ll accelerate into the moment of their death, dumping more water into the ocean every decade until WAIS has completely disappeared. They’ll collapse, in other words, sending global sea levels 15 feet higher than they would be.

This is the mechanism that Wolovick’s walls try to stop. His models suggest that simply constructing a sill on the ocean floor will keep warm water at depth from reaching the glacier. With less warm water to paw at their their grounding line, glacial retreat stops, and often they actually gain mass.

Both Pine Island Glacier and Thwaites Glacier face the Amundsen Sea in West Antarctica. (NASA / GSFC / SVS)

Take Thwaites glacier, one of the largest outlet ice flows for the West Antarctic Ice Sheet and one of the glaciers that most worries scientists. Right now, Thwaites is retreating about 3,300 feet (1 kilometer) every year. When Wolovick turns on the model, he first lets it run for 100 years without building the sill, in order to simulate the passage of time and the onset of severe global warming. By the end of the run, the grounding line of Thwaites recedes 62 miles (100 kilometers) from its current position.

Then he builds the virtual sill. “And then it stabilized and was able to recover,” he told me. “In some cases, Thwaites grows beyond its present-day volume, and in those cases the grounding line advances onto the sill itself.”

In the most optimistic models, the ice shelves—the floating plain of ice that extends out from the grounding line—actually expands and attaches to the sill. This slows down the forward movement of the glacier, allowing the grounding line to advance forward.

Even in the most pessimistic scenarios—when Wolovick asks the simulated glacier, for instance, to quickly erode and destroy the sill over time—it still buys humanity some time, extending the life of the glacier by 400 or 500 years.

Wolovick cautions that the models he’s using are rudimentary enough that the time spans should be seen as promising avenues for further studies, not confident predictions of the future. “You shouldn’t read too much into the time scales of my model,” he says. “The model process removes a lot of small scale bumps [in the seabed], and those bumps can stabilize the grounding line temporarily.”

He recommends that coast-concerned humans build these sills in two places. First, they should construct them in the outlet fjords of Greenland’s largest glaciers, like Jakobshavn. These fjords are often only a mile or two wide, and an underwater dredging project there would resemble successful large-scale civil engineering projects, like the Palm Islands in Dubai. Greenland is also under the shared control of Denmark and the Greenlandic national government, two entities that might decide to undertake the construction project together.

If the sills seem to work in Greenland, then he recommends humanity moves to build them in Antarctica. This would be politically difficult—control of Antarctica is shared by 53 countries—and it would exceed the scale of any previous mega-engineering project. The ocean-facing front of Thwaites glacier is more than 60 miles across. Pine Island Bay glacier, another unstable ice flow linked to the WAIS, is about 25 miles across. And concerned countries might have to use submarines to build in either place, because some of the best construction sites available are beneath ice shelves that float on the surface of the sea.

Hikers march along an ice cliff in Antarctica in late January 2014. (David Merron / Getty)

They’ll have to work fast. In the past two decades, scientists have built a constellation of satellite observatories above the southernmost continent. Their measurements confirm: The retreat of West Antarctica has already begun. The glacier is shorter, faster-moving, and less heavy than it used to be. Whether it’s in full-scale collapse won’t be known until about 2050.

If it does go, the failure of the WAIS would be especially catastrophic for the United States. The world’s largest glaciers are so enormous that they have their own gravitational fields, which pull small amounts of ocean water toward them. The Atlantic and Pacific seaboards fall within the heart of WAIS’s gravitational halo, boosting any global jump in sea-level rise by as much as 25 percent.

Over the past several years, some scientists have identified a number of new mechanisms that may cause WAIS to fall apart even faster. One of them is called marine ice-cliff instability: As some glaciers in WAIS retreat further and further back, their icy fronts will tower 2,000 feet above above the sea floor. The ice simply won’t be strong enough to hold that much weight. Instead, the ice will crumble, and skyscraper-sized hunks of white will plunge into the water.

Another is hydrofracturing: As air temperatures get hotter in Antarctica, pools of water could form on the floating ice shelves. These pools of water could quickly disintegrate the ice beneath them, as happened in the Larsen Sea in 2002, when a Rhode Island-sized piece of ice fell apart in weeks. When ice shelves vanish, the landed glaciers behind them quicken their march to the sea.

Not every glaciologist agrees that the computer models get these mechanisms right yet. Last year, for instance, Robin Bell and her colleagues found an enormous waterfall on an ice shelf in Antarctica, as well as a number of other features that suggested pools of meltwater don’t always force ice shelves to disintegrate.

But when they are factored into computer models, the results are worrying. In 2013, the International Panel on Climate Change projected that sea levels would not top 3 feet, 2 inches (98 centimeters) by 2100. In a paper published last month, scientists accounted for those two new mechanisms and said that 2100 sea levels could actually hit 4 feet, 9 inches (146 centimeters). Some 153 million people, many of them Americans, would see their homes inundated.

Rob DeConto, a climate scientist at the University of Massachusetts, Amherst, said he was skeptical of Wolovick’s technique but understood why it was worth pursuing.

“I think my main reaction is, okay, so, maybe in the near term that would slow things down,” he said. “At what point do you decide that this is really happening? And is it worth the global-scale investment in engineering?”

He also worried that Wolovick’s proposal would only address warm water, when his research suggests that warm air can cause melt pools that are themselves catastrophic.

“In high-emission models, you get more prolonged periods of high air temperatures in the summertime, and then you get a lot of melt water—in certain circumstances, we know that’s really bad for ice shelf, regardless of what the ocean temperatures are,” he said. “We can save Thwaites in terms of the melt from the bottom up, but what happens when that whole margin is covered with massive amounts of meltwater every summer?”

Ken Caldeira, a climate scientist at the Carnegie Institution for Science, said that he would want to hear from engineers before investing further in a seafloor plan. “Without some numbers and some consultation with engineers, it is just a modeling thought experiment,” he said in an email. “I do not have the expertise to evaluate this proposal, but I am quite skeptical.”

Schemes to geo-engineer the WAIS have failed in the past. Glaciologists had once floated an idea to pump sea water into the middle of the Antarctica, so it could freeze solid and reduce the rise of global sea levels. In 2016, Katja Frieler at the Potsdam Institute for Climate Research and her colleagues studied the idea and found that doing so would actually accelerate the flow of Antarctic glaciers—while using up seven percent of the global energy supply.

“In terms of geo-engineering, I’m always in favor of simply leaving the fossil fuel in the ground- and relying on proven technology that already exists, like renewables,” said DeConto in an email.

So is Wolovick, for that matter. “It’s important to emphasize that any sort of geo-engineering is not a substitute for emissions reductions,” he told me. “Rising sea levels are not the only negative consequence of climate change, and glacial geo-engineering doesn’t do anything about thermal expansion, much less ocean acidification and heat waves.”

“And the other thing is, it couldn’t last forever,” he added. “The ultimate fate of the Antarctic ice sheet is closely tied to the total cumulative carbon emissions. If we burn all the carbon in the ground, then all of Antarctica will eventually go.”

It’s a warning that every glaciologist issued—and a fitting one for our ambiguous moment in history. Nearly every dire projection of sea-level rise assumes that people continue to burn fossil fuels at ferocious rates, especially in the most undeveloped parts of the world. Will this prediction come to pass? Events contradict each other too much to know now either way. Consider recent news: China may ban the use of gas-burning cars even as its companies keep building coal plants. American carbon-dioxide emissions continue to fall even as its federal government promises to open nearly the entire Atlantic and Pacific coast to oil drilling. Solar power is the fastest growing source of new energy, but India says it will run coal plants “for decades to come.”

In December, at the same conference where Wolovick unveiled his idea, DeConto presented early modeling evidence suggesting that if the world manages to prevent global temperatures from rising by more than two degrees, it may be able to stave off the collapse of the WAIS entirely. “It’s attainable,” he told me. “It just would take an internationally coordinated will to make it happen.”

There’s little sign of that international coordination right now. And that opens the door to another world, where global carbon emissions skyrocket, and future geo-engineers must add many more glaciers to their list.

“It’s not just about Thwaites, right?” said DeConto. “Thwaites is getting a lot of attention because that’s where the action is that we’re observing right now. But there are other outlet glaciers we’re observing across the continent. And there’s way more ice that can contribute to sea-level rise in the deep basins in East Antarctica—way more, vastly more than WAIS. There are outlet glaciers there that could become responsive if things warm up enough.”

Then it really won’t be about just Thwaites, or Pine Island Bay, or Jakobshavn. And no pile of sand and stone will hold back the tide.