Fractures spreading across the floating edge of Antarctica’s Thwaites Glacier have reached a point that alarms the scientists who study them most closely. A cluster of peer-reviewed studies published over the past two years shows that the Thwaites Eastern Ice Shelf, a critical buttress holding back one of the continent’s largest and fastest-moving glaciers, is breaking apart from within. The driver is not primarily warm ocean water eating the ice from below. Instead, rifts and cracks that began propagating around 2016 have accelerated in sharp bursts, weakening the shelf’s structure faster than many models anticipated.
The stakes are enormous. Thwaites contains enough ice to raise global sea levels by roughly 65 centimeters, about two feet, on its own. If its collapse destabilizes neighboring glaciers across the wider West Antarctic Ice Sheet, the total contribution could eventually exceed three meters. Much of this research has emerged from the International Thwaites Glacier Collaboration (ITGC), a joint U.S.-U.K. effort involving more than 100 scientists and one of the largest Antarctic field campaigns ever mounted.
What is verified so far
The clearest picture of the shelf’s decline comes from a study published in the Journal of Geophysical Research: Earth Surface that documents stage-by-stage weakening through the evolution of shear-zone fractures. The researchers tracked how crack orientations shifted over time and how damage zones interacted, concluding that the ice shelf is moving through identifiable phases of structural failure rather than heading toward a single, sudden break. Each phase brings more extensive cracking and a measurable loss of mechanical integrity, evidence that the shelf is already well along a path toward disintegration.
A separate analysis in the Journal of Glaciology attributes the shelf’s destabilization primarily to rift propagation rather than basal melting. By reconstructing a timeline of rift growth beginning around 2016 and highlighting rapid extensions that the study’s authors identified in 2021 and 2022, the paper shows that surface and full-thickness fractures have become the dominant agents of change. The distinction matters: if fracturing, not ocean heat alone, is now the main driver, then projections built largely on water temperature and melt rates may underestimate how quickly the shelf can lose its ability to hold back inland ice.
Field observations published in Nature reinforce that paradox. Instruments lowered through boreholes into the eastern Thwaites grounding zone found that warm deep water was not freely flushing the area where the glacier lifts off the seabed. Stratified, relatively calm water beneath the ice was limiting direct erosion of the shelf’s underside. Lead author Peter Davis of the British Antarctic Survey and colleagues reported melt rates well below what leading models had predicted, a finding that initially sounds like good news but carries a troubling implication: the ice is structurally deteriorating even where the ocean is not aggressively undercutting it.
Satellite and geophysical work adds further detail. A study led by Pietro Milillo and colleagues used radar interferometry with daily-repeat ICEYE synthetic aperture radar data from March to June 2023 to reveal widespread seawater intrusions beneath grounded ice across a grounding zone spanning several kilometers. Those intrusions suggest the boundary between grounded and floating ice is more diffuse and unstable than earlier maps indicated, allowing ocean water to penetrate inland along weaknesses in the glacier’s bed. Such incursions can lubricate the ice-bed interface and concentrate stress into pre-existing cracks.
Research published in Nature Geoscience showed that fracturing drives short-lived speedups and slowdowns in the flow of Thwaites’ floating extensions. When rifts extend or link up, flow can briefly accelerate; when stresses redistribute, the ice can temporarily slow. Over time, these pulses contribute to a net trend of thinning and retreat. A complementary study in The Cryosphere found that the glacier thins and retreats most rapidly where basal channels intersect its grounding zone, providing a geographic explanation for why instability clusters in particular corridors. These channels focus both ocean heat and mechanical stress, making them natural pathways for accelerated damage.
Oceanographic observations in Pine Island Bay, the waters adjacent to the Thwaites Eastern Ice Shelf cavity, have documented conditions consistent with deepwater warming and episodic pulses of heat reaching the continental shelf, according to measurements reported by researchers involved in the ITGC. Although these heat pulses do not yet appear to be the primary cause of the shelf’s current failure, they represent a looming hazard that could amplify fracture-driven weakening if circulation patterns shift.
What remains uncertain
Despite the growing body of evidence, several gaps limit how precisely scientists can forecast what comes next. No publicly available datasets track fracture propagation rates beyond mid-2023, leaving the most recent pace of crack growth undocumented in the peer-reviewed record. The ICEYE-based analysis, while unprecedented in its temporal resolution, covered only a four-month window. Seasonal effects, evolving stress patterns, and potential new rift initiations since then remain open questions as of early 2026.
The interplay between suppressed basal melting and fracture-driven instability has not yet been captured in a single, unified framework. Glaciological studies tend to focus on the mechanics of crack initiation, propagation, and coalescence, treating the ice shelf as a deforming solid. Oceanographic work emphasizes heat transport, stratification, and circulation beneath the ice. How these two systems feed back on each other, whether fractures can reorganize water flow under the shelf, or whether subtle changes in melt can tip a damaged shelf into rapid collapse, remains unresolved.
Coupled models that explicitly integrate fracture physics with ocean forcing are still in early stages of development. Existing ice-flow simulations often parameterize damage in simplified ways or assume smooth grounding lines, while ocean models may not resolve narrow basal channels and evolving rift geometries. Bridging that gap will be essential for determining whether Thwaites’ retreat is likely to proceed gradually or transition into a more abrupt, nonlinear phase as structural thresholds are crossed.
Grounding-line mapping from NASA’s National Snow and Ice Data Center, drawing on differential satellite radar interferometry data spanning more than three decades, provides the longest continuous baseline for tracking where the glacier detaches from the seabed and begins to float. These maps clearly show substantial inland migration of the grounding line, particularly along zones of concentrated thinning and basal channel activity. Yet translating that retreat into precise sea-level timelines requires assumptions about how quickly upstream ice will respond and whether neighboring ice shelves can continue to provide any buttressing. Different research groups make different choices about those assumptions, producing a range of plausible futures rather than a single forecast. Current estimates for Thwaites’ contribution to sea-level rise by 2100 vary from several centimeters to tens of centimeters, depending on how fracture dynamics and ocean warming interact.
How to read the evidence
The most robust conclusions center on what has been directly measured rather than modeled. High-resolution satellite imagery, radar interferometry, and borehole instruments all point to a structurally compromised Thwaites Eastern Ice Shelf whose fractures and rifts have expanded markedly since the mid-2010s. The evidence that basal melting in key parts of the grounding zone is currently lower than expected strengthens the case that mechanical failure, not just thermal erosion, is driving the present phase of change.
At the same time, the record does not yet support confident statements about the exact timing or manner of a major shelf collapse. The lack of continuous, multi-year fracture monitoring and fully coupled ice-ocean models means that projections still rely on extrapolating from a relatively short, intense period of observation. It is clear that the shelf is weakening and that its loss would accelerate ice discharge from the Thwaites catchment, with significant implications for coastal communities worldwide. It is less clear whether that transition will unfold over decades or could be triggered more abruptly by a combination of rift linkage, grounding-line retreat, and shifting ocean conditions.
For readers following new developments on Thwaites, the emerging consensus as of May 2026 is that the glacier’s most immediate vulnerability lies in its fractured, thinning ice shelves rather than in runaway melting alone. Structural damage has already advanced to the point where large sections of the Eastern Ice Shelf are primed for further disintegration, even under only moderate ocean forcing. Future work combining detailed fracture mapping, long-term satellite records, and refined ocean observations will be critical for narrowing the range of outcomes. Until then, the evidence supports viewing Thwaites as a system already in transition, with cracks opening the door for ocean heat to play a larger and potentially more destabilizing role in the years ahead.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
