Over the past three decades, Antarctica’s ice sheet has lost its grip on roughly 12,820 square kilometers of seafloor, an area spanning nearly 5,000 square miles, about 10 times the size of Greater Los Angeles. The retreat, documented in a study published in the Proceedings of the National Academy of Sciences in March 2026, represents the most comprehensive continent-wide accounting of grounding-line migration ever assembled. For the hundreds of millions of people living in low-lying coastal areas, these grounding lines function as physical brakes on the ice sheet: when they pull back, glaciers accelerate toward the ocean and sea levels climb.
33 years of radar, one continent-wide map
Researchers at the University of California, Irvine stitched together 33 years of satellite radar data, spanning multiple European and NASA missions from 1992 through 2025, to track the boundary where Antarctic ice lifts off the bedrock beneath it. The technique, called differential synthetic aperture radar interferometry (DInSAR), detects tiny surface deformations caused by ocean tides flexing the ice. By analyzing those flex patterns across thousands of satellite passes, the team mapped grounding-line positions for every major glacier system on the continent.
“This is the first time we’ve been able to see the entire Antarctic grounding line evolve over multiple decades,” said Eric Rignot, a glaciologist at UCI and lead author of the study, in a university news release. The team reported the total loss at 12,820 plus or minus 1,873 square kilometers, with the full dataset, including glacier-by-glacier retreat calculations and GIS-ready grounding-line vectors, publicly available through the Dryad data repository.
That openness matters. Any glaciologist in the world can now download the raw files, check the numbers, and test them against independent ice-flow models. It is a level of transparency that turns a single study into a shared baseline for the entire field.
West Antarctica bears the worst of it
The damage is not spread evenly. Roughly 77 percent of Antarctica’s coastline remained stable over the 33-year window, a figure reported by the same UCI study team in its institutional press release rather than by an independent assessment. But the losses that did occur are concentrated in West Antarctica, and they are severe.
The Amundsen Sea Embayment, home to the Thwaites and Pine Island glaciers, recorded a maximum grounding-line retreat of about 43 kilometers, the largest pullback documented anywhere on the continent. The European Space Agency published a visualization showing how that retreat unfolded year by year, with individual glacier fronts migrating steadily inland.
None of this is entirely new territory. A 2014 study published in Geophysical Research Letters (Rignot et al., 2014) used NASA radar observations covering 1992 through 2011 to flag rapid retreat at the Pope, Smith, and Kohler glaciers in the same embayment, with the authors describing the losses as potentially irreversible. What the new PNAS study adds is more than a decade of additional data and, critically, a continent-wide framework that places those individual glacier retreats in context. The Amundsen Sea sector is not just losing ice quickly; it is losing ice faster than anywhere else on the continent, and it has been doing so for at least 33 years.
Elsewhere, the picture is more mixed. Parts of the Antarctic Peninsula and some East Antarctic outlet glaciers show modest retreat. Others appear nearly stationary. A few have even advanced slightly where local ocean currents or heavier snowfall favor ice growth. Those localized gains do not offset the large-scale losses in the west, but they explain why the majority of the coastline is classified as stable. The ice sheet is not collapsing uniformly. Specific sectors are undergoing dramatic, sustained change while others have so far held firm.
What the numbers cannot yet tell us
Mapping where grounding lines have retreated is not the same as predicting where they will go next. The PNAS study provides a historical record, not a forecast. Translating that record into a sea-level projection requires assumptions about future ocean warming, ice dynamics, and the shape of the bedrock beneath the ice, all of which remain actively debated.
Tidal cycles add a layer of complexity. Ocean tides can push a grounding line inland or seaward by several kilometers within a single day, meaning any individual satellite pass captures a snapshot that may not represent the long-term average. The DInSAR method accounts for this by combining repeat-pass observations and analyzing tidal flex patterns, but the magnitude of tidal noise varies by glacier and by mission. The study’s reported uncertainty margin of plus or minus 1,873 square kilometers reflects, in part, that variability.
There is also the question of feedbacks. As ice pulls back into deeper basins, warm ocean water can intrude beneath it more easily, potentially accelerating further retreat. Conversely, if the bedrock rises or narrows inland, it can act as a stabilizing ridge, slowing or pausing the pullback for decades. The new dataset gives modelers a baseline against which to test those competing scenarios, but it does not resolve which feedbacks will dominate later this century.
No economic impact assessment has been released alongside the study. The connection between grounding-line retreat and sea-level rise is well established in glaciology, but converting 12,820 square kilometers of lost grounded ice into centimeters of global sea-level change requires coupling these observations with ice-flow models that are still being calibrated against the new data. The Intergovernmental Panel on Climate Change’s most recent assessment report estimated that Antarctica could contribute anywhere from a few centimeters to more than half a meter of sea-level rise by 2100, depending on emissions. Datasets like this one are designed to narrow that range.
What coastal planners should watch next
For anyone tracking flood risk, infrastructure investment, or insurance exposure in coastal cities from Miami to Mumbai to the low-lying Pacific islands, the practical signal is clear: the physical structures that restrain Antarctic ice flow have weakened across a measurable, significant area, and the fastest changes are happening in the sector glaciologists have long considered most vulnerable.
The new map does not assign a deadline or a dollar figure. What it does is identify, with unprecedented precision, where the ice sheet is already losing contact with the seafloor and where further retreat is most plausible. As modeling groups fold these observations into their projections over the coming months, the combination of high-resolution satellite measurements and improved ice-flow physics should begin to tighten the range of plausible futures. For communities making decisions about seawalls, drainage systems, and buildings meant to last 50 or 100 years, that narrowing cannot come soon enough.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
