Antarctica Is Melting from Below and Scientists Say It Is Worse Than Anyone Expected

There is something deeply unsettling about the way Antarctica is melting. You might picture it as a slow, visible process — icebergs calving into the ocean, glaciers retreating at their edges, dramatic cliffs of ice crumbling under a warming sun. And some of that is certainly happening. But the most alarming thing occurring in Antarctica right now is largely invisible. It is happening in the dark, in the frigid waters beneath the continent's ice shelves, hundreds of metres below the surface. Warm ocean water is creeping silently under the ice and eating it from the bottom up. And a series of major research findings published in 2026 has revealed something that should concern everyone living near a coastline: this process is moving faster than almost anyone predicted, and the climate models that world governments rely on to plan for sea level rise are not even accounting for the full picture.

What Is Basal Melting — and Why Does It Matter?

To understand why scientists are alarmed, it helps to understand how Antarctic ice actually holds back the sea. The Antarctic continent is covered by an immense ice sheet — a layer of ice so massive that if it melted entirely, global sea levels would rise by tens of metres. But the ice sheet does not sit neatly on land everywhere. At its edges, glaciers flow outward and extend over the ocean as floating platforms called ice shelves. These shelves act as natural buttresses or brakes — their weight and friction slow the flow of land ice behind them into the sea.

When the underside of an ice shelf is attacked by warm ocean water, the shelf thins and weakens. As it does, it gradually loses its ability to hold back the glaciers behind it. Those glaciers then accelerate toward the sea, calving off ice and raising global sea levels. This is called basal melting — melting from the base — and it is the primary mechanism by which Antarctica is currently contributing to sea level rise. The process is hidden from view, which is part of why it has been so difficult to measure accurately. But in 2026, researchers have made significant breakthroughs in understanding just how far advanced, and how self-reinforcing, this threat really is.

What the May 2026 Research Found

The most consequential new findings were published in the journal Nature Geoscience on May 15, 2026, by a team led by University of Maryland scientist Madeleine Youngs. The study, titled "Antarctic ice-shelf basal melt shaped by competing feedbacks," revealed something that most global climate models — including those used by the Intergovernmental Panel on Climate Change (IPCC) — have been leaving out entirely: a powerful self-reinforcing feedback loop that causes melting to accelerate its own pace.

The mechanism works like this. Under normal conditions, cold, dense, highly saline water sits on the continental shelf around Antarctica and acts as a barrier, preventing warmer deep ocean water from reaching the underside of the ice shelves. When ice melts, however, it releases freshwater. That freshwater mixes with the ocean, reducing its salinity and making it lighter. As the dense cold water layer weakens, the barrier breaks down, and warmer water from the deep ocean can flow up and underneath the ice shelves far more easily. More warmth means more melting, which releases more freshwater, which further weakens the barrier, which allows even more warm water in. It is a positive feedback loop — a self-amplifying cycle where each step makes the next one worse.

The scale of this effect, once properly modelled, proved startling. Youngs and her team found that this melt-driven feedback accounts for approximately two-thirds of the overall increase in melt rates projected across Antarctic ice shelves under future climate scenarios. To put that plainly: the warming of the atmosphere and ocean accounts for only around one-third of the accelerated melting scientists expect to see. The rest comes from the process of melting itself. "Most current climate models that inform international policy don't consider this feedback loop at all," Youngs said. "The IPCC treats melting as a fixed, rather than interactive, input." The implication is sobering: sea level rise projections based on those models may be significantly understating what is coming.

A companion study published in Nature Communications in May 2026 by Norwegian researchers added another layer of concern. That team used high-resolution modelling to show that channels carved into the undersides of ice shelves — long grooves in the ice that were previously thought to be relatively benign features — are in fact traps for warm ocean water. The warm water flows into these channels and becomes concentrated there, intensifying melting in specific areas by an order of magnitude compared to surrounding ice. This channelised melting does not just thin the ice from below — it drives the channels to grow wider and deeper over time, progressively destabilising the structural integrity of the shelf from the inside.

Which Ice Shelves Are Most at Risk?

Antarctica is not one uniform block of ice. It contains dozens of distinct glaciers and ice shelves, each interacting with the ocean in different ways and at different rates of vulnerability. Understanding which are most endangered is crucial for predicting where the greatest near-term risks lie.

The Thwaites Glacier in West Antarctica — known informally as the "Doomsday Glacier" — has long been the focus of greatest scientific concern, and recent findings have done nothing to ease those worries. Roughly the size of Florida and more than 2,000 metres thick in places, Thwaites is already responsible for about 4% of annual global sea level rise. Its full collapse would add approximately 65 centimetres to global sea levels on its own, enough to inundate vast areas of coastal land around the world. Crucially, the eastern ice shelf attached to Thwaites — the floating structure that acts as its primary brake — could begin breaking apart within the next few years, according to recent research. Once that natural brake is gone, the glacier's flow toward the sea is expected to accelerate significantly.

Thwaites does not stand alone. It acts as a dam for the broader West Antarctic Ice Sheet, which holds enough ice to raise sea levels by up to 3.3 metres if it were to melt entirely. Scientists have found that the continent's ice flows through interconnected basins, meaning that destabilisation in one region can cascade into others. The Amundsen Sea region, which includes both Thwaites and Pine Island Glacier, is experiencing rapid ocean warming — at approximately triple the historical rate, according to modelling studies — and widespread increases in ice-shelf melting. A separate April 2026 study of ocean sediments found that human-caused warming is triggering a long-term meltdown that appears, at this point, to be irreversible regardless of how aggressively emissions are reduced.

The Ross Ice Shelf, the largest ice shelf in Antarctica, is also attracting new attention. Research published in April 2026 in Science Advances found that the ocean deep beneath the Ross Ice Shelf is far more variable and dynamic than scientists had assumed, responding rapidly to tidal flows and the complex topography of the seafloor. That variability means heat can find its way under the ice far more readily than previously believed.

What This Means for Sea Level Rise Projections

The combined effect of these findings is that sea level projections — already sobering — need to be revised upward. The Youngs study's most alarming conclusion is that because the feedback loop between meltwater and ocean circulation has not been incorporated into IPCC models, current sea level rise estimates may be too conservative by a significant margin.

A convergence of evidence published in 2026 points toward an acceleration of sea level rise that could reach approximately one metre by the end of this century and potentially up to 1.5 metres in some equatorial regions. Other modelling scenarios put the plausible upper range higher still. Scientists studying sediment cores and paleoclimate records — the evidence of what happened to sea levels during past warm periods — have found that the average global sea level rise in the coming decades is likely to accelerate, potentially reaching three feet by 2100 under current emissions trajectories, with higher projections for the most vulnerable coastal regions.

These numbers are not abstractions. Every centimetre of sea level rise increases the risk, frequency, and severity of coastal flooding events. A one-metre rise transforms what was previously a once-in-a-century flood event into something that happens several times a decade. It erodes the buffer that protects coastal infrastructure from storm surges. And it permanently submerges low-lying land that billions of people currently call home.

What This Means for Coastal Cities

The cities that stand most in the crosshairs of Antarctic ice loss are, broadly, the world's great coastal metropolises and low-lying island nations. Half a metre of sea level rise — roughly the lower end of plausible projections by 2100 — would be enough to submerge large portions of cities like Bangkok and Manila, threaten significant stretches of the Netherlands, and expose the eastern coastline of England to far more severe and frequent flooding. At a metre of rise, Miami's flood problems — already severe today — would become existential. Jakarta, which is also sinking due to groundwater extraction, would face a compounded crisis. Mumbai, Shanghai, New York, New Orleans, and dozens of other major cities would need to invest hundreds of billions in coastal defences or accept the permanent loss of significant portions of their urban areas.

What makes the Antarctic findings particularly pressing is not just the scale of the risk but the timeline. The loss of Thwaites' eastern ice shelf, the activation of the meltwater feedback loop described in the Youngs study, and the channelised melting now identified in cold water ice shelves all point toward an acceleration that is already underway — not a threat that exists only in distant future scenarios. The direction of travel, as one climate scientist put it, has been clear for years. What 2026 has made clearer is how far along that road we already are.

What the Science Is Saying We Should Do

The honest scientific answer is that no amount of action can now prevent some significant sea level rise. The question is whether humanity acts decisively enough to keep that rise at the lower and more manageable end of projections, or whether continued emissions push it toward the higher end where adaptation becomes increasingly impossible for hundreds of millions of people.

Madeleine Youngs' team at the University of Maryland is already developing higher-resolution simulations that will trace projected trajectories through to 2100, with a particular focus on identifying which ice shelves are closest to what scientists call the "point of no return" — the threshold beyond which collapse becomes self-sustaining and irreversible. "The next step is understanding exactly when and where things tip — and what that means for all of us," Youngs said. That phrase, "for all of us," is worth sitting with. Antarctica may be the most remote place on Earth, but its fate is bound to every coastline on every continent.

Conclusion

The 2026 Antarctic research findings represent a maturation of scientific understanding — moving from alarming projections to alarming measurements, from models of what might happen to observations of what is already happening. The discovery that meltwater creates its own feedback loop, driving further melting at a rate comparable to direct atmospheric warming, is a landmark finding that demands a revision of how the world plans for the rising ocean. The identification of channelised traps that intensify melting by orders of magnitude adds another layer of urgency to an already critical picture.

None of this is yet a reason for paralysis. Cities can be defended. Infrastructure can be redesigned. Coastal populations can be prepared and, where necessary, relocated with enough time and political will. But all of those responses require first accepting the scale of what the science is now describing. Antarctica is melting from below, faster than the models predicted, driven by a self-amplifying process that most climate projections do not even include. The ice does not care about policy timelines. It responds to physics. And right now, the physics is not on our side.

*This article is for informational purposes only. The research cited includes studies published in Nature Geoscience (May 15, 2026), Nature Communications (May 2026), and Science Advances (April 2026), alongside related findings from The Invading Sea and ScienceDaily.*