Beneath Antarctica’s floating ice shelves, warm ocean water is carving hidden channels into the ice from below, and a study published in April 2026 reveals that scientists have been significantly underestimating how much ice those channels destroy. The research, published in Nature Climate Change, found that this focused, channelized melting has been undercounted by 42 to 50 percent, a correction large enough to force a rethink of how quickly some of the continent’s most vulnerable ice platforms could weaken and break apart.
The findings land at a moment when separate oceanographic evidence confirms that the warm deep water responsible for this erosion has been creeping closer to the continent over the past two decades. For the roughly 900 million people living in low-elevation coastal zones worldwide, the stability of Antarctic ice shelves is not an abstraction. These floating platforms act as buttresses, bracing the vast land-based ice sheets behind them. When a shelf thins or collapses, the glaciers it once restrained can accelerate toward the sea, raising ocean levels.
Hidden channels, sharper measurements
The breakthrough behind the new numbers is a leap in resolution. The study was led by researchers at the University of California, San Diego’s Scripps Institution of Oceanography, with lead author Susheel Adusumilli and colleagues combining stereo satellite imagery with satellite altimetry to produce basal melt maps at roughly 50-meter resolution, fine enough to pick out narrow melt channels carved into the underside of ice shelves. Previous satellite products smoothed over these features, averaging them into broader melt estimates that missed the concentrated erosion happening along structural weak points.
“The channels we resolved are not minor surface features,” Adusumilli noted in a summary accompanying the paper. “They represent focused pathways where ocean heat is doing the most damage to the structural integrity of the shelf.”
Think of it this way: older maps were like looking at a riverbed through frosted glass. You could see that water was flowing, but you could not make out the individual channels cutting into the rock. The new method clears the glass. And what it reveals matters, because wider, deeper channels act like fault lines. They precondition the ice above them for fracture, making shelves more prone to calving and eventual collapse.
This high-resolution work builds on a continent-wide melt record assembled from satellite radar altimetry, ice-velocity measurements, and snow-compaction modeling. That foundational dataset, published in Nature Geoscience in 2020, produced time-averaged basal melt rates for 2010 to 2018 and a meltwater flux time series stretching back to 1994. Together, the two products paint a picture of basal melting that is not only larger than previously measured but also highly variable from year to year, with sharp spikes in meltwater discharge that feed directly into the Southern Ocean.
Warm water on the move
The physical driver behind the accelerating melt is well documented at Pine Island Glacier Ice Shelf, one of the fastest-changing outlets in West Antarctica. There, wind-driven upwelling lifts a layer of relatively warm deep water over the continental shelf edge and pushes it beneath the floating ice. Research published in Science showed that changes in ocean circulation beneath the shelf drove rapid thinning, and NASA’s summary of that work describes how shifts in regional wind patterns can rapidly increase the volume of warm water reaching the ice base, turning a steady background melt into a surge.
Now, a broader oceanographic trend is amplifying the concern. Repeat ship-based observations and water-mass classification published in Communications Earth & Environment in 2026 confirm that Circumpolar Deep Water, the warm, salty layer that sits below the cold surface waters of the Southern Ocean, has shifted and expanded poleward over roughly the past 20 years. In plain terms, the heat source responsible for eating away at ice shelves has moved closer to the ice.
That proximity matters. When warm deep water sits near the continental shelf break, relatively modest changes in wind strength or direction can push it underneath floating ice. The Pine Island scenario, once treated as a regional case study, starts to look like a preview of what could happen at other vulnerable shelves if the warm water continues its poleward migration.
What scientists still cannot pin down
For all the clarity these new measurements provide, significant gaps remain. The 50-meter-resolution melt maps so far cover selected vulnerable shelves, not the entire Antarctic coastline. The 42 to 50 percent underestimate may not apply uniformly. Some shelves sit in colder ocean settings where channelized erosion plays a smaller role, and extending the new method everywhere will require additional satellite passes and processing time.
The continent-wide meltwater flux record ends in 2018, leaving several years unaccounted for during which ocean temperatures and wind patterns have continued to shift. Without updated observational records, researchers cannot yet confirm whether the interannual melt spikes seen in the 2010-to-2018 window have intensified or leveled off.
Fieldwork at the Ross Ice Shelf grounding zone has offered a rare ground-truth glimpse. Researchers drilled through the ice and found a thin water column where tidal cycles drive mixing and internal waves lift heat toward the ice base, according to reporting published by Phys.org from The Conversation. Those in-situ measurements, however, come from a single borehole and cannot yet be generalized across the vast Ross shelf, the largest in Antarctica.
Perhaps the most consequential unknown is the timeline. The newly measured channelized melt has not yet been fed into ice-sheet models capable of translating it into centimeters of global sea-level rise over defined periods. That next step requires a new generation of simulations that can ingest 50-meter-scale melt patterns and run them forward under different climate scenarios. Until those models are completed and peer-reviewed, converting the revised melt estimates into precise sea-level projections remains premature.
Antarctica currently contributes roughly 0.3 to 0.5 millimeters per year to global sea-level rise, according to recent assessments, but that rate is expected to grow as ice loss accelerates. The question these findings sharpen is not whether the contribution will increase, but how fast.
Why the direction of change is clear, even if the speed is not
The strongest evidence in this story comes from peer-reviewed remote-sensing analyses and ship-based oceanographic surveys published in Nature-family journals or supported by NASA research programs. These are primary observational datasets, not model projections or opinion pieces. The 50-meter melt maps represent a direct measurement advance: they show what is happening beneath the ice now, not what a simulation predicts might happen. The 42 to 50 percent correction is a comparison between this new product and older, coarser satellite-derived estimates, so it reflects an update to the observational record rather than a speculative forecast.
The meltwater flux time series from 1994 to 2018 similarly rests on observational inputs. Its value lies in showing how basal melt varies over time, which matters because climate models that assume steady melt rates will miss the surges that stress ice shelves most. The Pine Island case study, anchored in a Science paper and amplified by NASA, provides a concrete physical narrative: wind shifts drive warm water upwelling, and that upwelling feeds rapid basal erosion. It is a well-documented example, though conditions at Pine Island do not automatically apply to every shelf around the continent.
For readers weighing these findings, two points stand out. First, the direction of change is not in dispute. Observations show that many Antarctic ice shelves are thinning from below faster than previously measured, and warm ocean water is the primary driver. Second, the remaining uncertainty centers on the pace and geographic spread of that change, not on whether the underlying physics is real.
What emerges is a picture of an ice-ocean system that is more dynamic and more sensitive to small shifts in wind and water temperature than earlier coarse-scale maps suggested. Focused channels of melt beneath Antarctic shelves act like structural seams, preconditioning the ice for fracture while funneling freshwater into the Southern Ocean in intermittent pulses. Recognizing and measuring those hidden pathways is a necessary step toward more reliable projections of coastal risk around the world, even as the exact timeline of future sea-level rise remains an open and urgent research question.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
