It's not thinning and going away. It shatters.
Beneath one of Earth's most remote and consequential glaciers, a small robot has returned with news that reframes how humanity understands its own coastal future. The Thwaites Glacier — vast enough to raise global seas by more than two feet — is not quietly dissolving from below so much as it is shattering from within, its fractures outpacing the warm water long blamed for its decline. Scientists gathered this knowledge at great cost and with incomplete access, yet even a partial truth, honestly held, may be more valuable than the comfortable assumptions it displaces.
- A robot the size of a pencil dove beneath Antarctic ice and upended the prevailing model of how the world's most feared glacier is dying — not melting quietly, but fracturing violently.
- The discovery of crevasses as the dominant destructive force is alarming because ice that shatters and floats away raises sea levels faster and less predictably than ice that simply thins.
- Researchers could only reach the glacier's more stable eastern flank — the main trunk, cratered like a field of sugar cubes, remains beyond safe landing range for any aircraft heavy enough to carry equipment.
- The $50 million international mission found melting rates lower than feared in accessible zones, offering a brief exhale before the caveat: the worst areas remain unstudied and are likely worse.
- Scientists now warn that Thwaites' collapse may unfold through cascading fracture rather than gradual melt, a mode of failure that compresses timelines and complicates coastal planning worldwide.
Antarctica's Thwaites Glacier — nicknamed the Doomsday Glacier for the two-plus feet of sea level rise locked within it — has long been studied from a distance. In February 2023, that changed. A 13-foot robot called Icefin was lowered through a nearly 2,000-foot borehole and swam beneath the glacier's grounding line, sending back the first direct images of what is happening where ice meets ocean.
What the robot found was both reassuring and deeply unsettling. Basal melting — the warm ocean water gnawing at the glacier's underside — was slower than models had predicted. But the cameras also revealed something more alarming: a landscape riddled with crevasses, deep fractures that are splitting the glacier apart. Cornell scientist Britney Schmidt, who designed Icefin and co-led the research published in Nature, put it plainly: Thwaites isn't thinning away — it shatters. When ice fractures and calves into the sea, it displaces water and raises sea levels in ways that gradual melting does not.
The $50 million research effort was constrained by geography. Scientists could only access the glacier's more stable eastern section; the main trunk, where deterioration is most severe, is so fractured at the surface that no heavy plane can safely land. That inaccessibility leaves the most critical questions unanswered — though researchers believe the fracturing observed in the stable zone is likely worse where conditions are more extreme.
The findings carry a cautious thread of hope. Understanding the mechanism of failure — fracture over melt — gives scientists and policymakers a clearer target. Ted Scambos of the National Snow and Ice Data Center noted that while Thwaites will threaten coastlines for at least a century, this knowledge creates time to develop strategies that might slow the flooding. Whether that window is used wisely remains the open question.
Icefin's final images added an unexpected grace note: sea anemones drifting near the ice-ocean boundary, sediment glittering like stars in the robot's lights. Life persisting at the edge of collapse. The larger story, though, belongs to a glacier coming apart faster than predicted, and to rising seas that will redraw the world's coastlines for generations.
Antarctica's Thwaites Glacier has earned a grim nickname—the Doomsday Glacier—because of what it holds: enough ice to raise global sea levels by more than two feet if it all melts into the ocean. For years, scientists could only guess at what was happening to this Florida-sized expanse of ice. Now, for the first time, they have seen it up close, and the picture is more complicated than they feared.
A pencil-shaped robot named Icefin, just 13 feet long, was lowered through a 1,925-foot hole drilled into the ice near McMurdo Station in Antarctica. The robot swam beneath the grounding line—the critical point where the glacier first extends over open water—and sent back images that revealed something unexpected. The melting happening underwater, where warm ocean water gnaws at the glacier's underside, is actually slower than scientists had predicted. That's the good news. The bad news is that it barely matters.
What the robot's cameras showed instead was a landscape fractured by crevasses—deep cracks in the ice that splinter the glacier into pieces. These fractures, not the melting, are doing the real damage. "That's how the glacier is falling apart," said Britney Schmidt, the Cornell University polar scientist who created Icefin and led one of two studies published in Nature on February 15, 2023. "It's not thinning and going away. It shatters." The distinction is crucial. When ice breaks off from land and floats into the sea, it displaces water and raises sea levels. When it simply melts in place, the effect is different. The fracturing accelerates the glacier's collapse in ways that melting alone cannot.
The research represents a $50 million international effort to understand the world's widest glacier, which has been changing faster in recent years than it was five years ago. Scientists were able to explore only the eastern, more stable portion of Thwaites—the part where conditions are relatively less catastrophic. The main trunk, where the glacier is breaking apart most rapidly, remains largely inaccessible. The surface there is so cratered and crevassed that it resembles a pile of sugar cubes, with no safe place for a heavy airplane to land. A helicopter might eventually reach it, but the logistics are formidable.
This limitation means the full picture of Thwaites' deterioration remains incomplete. The crevasses that scientists observed in the more stable eastern section are themselves breaking apart faster than expected. If similar fracturing is occurring in the main trunk—and researchers have reason to believe it is—the glacier's collapse could accelerate beyond current projections. Paul Cutler, the National Science Foundation's Thwaites program director, warned that the glacier's "eventual mode of failure may be through falling apart," with fracturing potentially hastening the overall demise of the ice shelf.
Yet there is a sliver of hope embedded in the findings. The research provides a clearer understanding of the processes destroying Thwaites, and that knowledge, however sobering, creates a window for action. Ted Scambos of the National Snow and Ice Data Center noted that while Thwaites will remain a major threat to sea levels for at least a century, better understanding of how the glacier is failing gives humanity time to pursue strategies that might slow the pace of coastal flooding. The challenge now is whether that time will be used wisely.
When Icefin completed its journey through the underwater cavity, its cameras captured one more detail that seemed almost incongruous with the larger story of collapse: sea anemones, small creatures that should not exist in such a harsh environment, were drifting near the ice-ocean interface. In the background, sediment and rocks picked up by the glacier sparkled like stars. It was a reminder that even in a landscape of fracture and dissolution, life persists in the margins. But the larger narrative remains one of accelerating change, of a glacier that is coming apart faster than anyone predicted, and of rising seas that will reshape coastlines for generations to come.
Notable Quotes
That's how the glacier is falling apart. It's not thinning and going away. It shatters.— Britney Schmidt, Cornell University polar scientist and Icefin creator
Unfortunately, this is still going to be a major issue a century from now. But our better understanding gives us some time to take action to slow the pace of sea level rise.— Ted Scambos, National Snow and Ice Data Center
The Hearth Conversation Another angle on the story
So the robot found that melting is slower than expected. Shouldn't that be reassuring?
It would be, except that melting isn't actually the main problem. The glacier is fracturing—breaking into pieces—and that's what's driving the ice into the ocean and raising sea levels. The slower melting doesn't change that.
Why does fracturing matter more than melting?
When ice is still attached to land, it doesn't contribute to sea level rise. But when it breaks off and floats, it displaces water. The fractures are essentially breaking the glacier apart into pieces that can float away. It's a different mechanism entirely.
They only studied part of the glacier, though?
Right. They could only safely access the eastern section, which is actually more stable. The main trunk, where things are falling apart fastest, is too crevassed and dangerous to reach with a plane. So the worst part remains largely unseen.
That sounds like a significant gap in the research.
It is. But it also means the situation could be worse than what they've documented. If the fracturing they observed in the stable section is happening faster in the main trunk, the glacier could be collapsing even more rapidly than current models suggest.
Is there any reason for optimism here?
The understanding itself is valuable. Scientists now know more precisely how Thwaites is failing, which gives policymakers better information for planning adaptation and mitigation. It's not a solution, but it's time—if we use it.