New measurements at Antarctica’s Thwaites Glacier show parts of the ice are melting at rates up to 43 m per year, far higher than the averages used in many models and concentrated along steep crevasses near the grounding line. These findings come from direct under-ice surveys and radar instruments that give scientists their sharpest view yet of how the so‑called “doomsday glacier” is being eaten away from below. The results matter for coastal communities worldwide because they point to faster and more uneven ice loss than simplified model physics have assumed so far.
Why Antarcticas doomsday glacier is melting faster than matters now
The sharpest warning signal comes from the base of Thwaites where the glacier first lifts off the seabed. There, a team using the Icefin robot and phase‑sensitive radar, known as ApRES, recorded melt rates along crevasse walls of about 43 m per year according to a Primary study in Nature. That figure does not represent the whole glacier, but it shows that narrow, steep features can lose ice many times faster than flat sections.
The same Icefin and ApRES observations show a strong contrast between relatively moderate average melt on smoother ice and extremely high melt in the crevasses and terraces near the grounding line according to the Primary research around Icefin. This pattern means a large share of the ice loss is concentrated in hidden undercut areas rather than spread evenly across the ice shelf base. As those undercut zones widen, they can weaken the glacier’s structural support and speed up the inland flow of ice.
For people living near sea level, the location of this aggressive melting matters as much as the raw number. Melt that hollows out the grounding zone can allow warm water to reach farther inland under the ice, which researchers describe as a pathway to rapid retreat according to the Primary work on Thwaites. That is why the discovery of such intense, localized melt has raised concern that earlier models, which often used smoother average melt rates, may have underestimated how quickly key parts of Thwaites can destabilize.
The evidence behind Antarcticas doomsday glacier is melting faster than
The Icefin campaign is only one piece of a broader effort to measure how Thwaites is changing. The same Nature study that deployed Icefin combined its robot footage with ApRES measurements to show that melt near the grounding line is highly heterogeneous, with terraces and crevasses taking a disproportionate share of ocean heat according to the Primary Icefin observations. That finding challenges the idea that a single average melt rate can capture what is happening beneath the ice.
Satellite evidence backs up the picture of rapid structural change. A Primary analysis using COSMO‑SkyMed repeat radar images documented grounding‑line retreat at Thwaites between 2016 and 2017 and found that the retreat rate in the so‑called butterfly extension of Thwaites and the Thwaites Eastern Ice Shelf doubled for 2011 to 2017 compared with 1992 to 2011 according to the COSMO‑based grounding‑line study. That doubling shows that the glacier’s response has already accelerated beyond its late twentieth‑century behavior.
NASA has linked this retreat to large cavities forming under the ice. A peer‑reviewed satellite study described a large sub‑ice cavity at Thwaites that was measured using radar instruments, and the agency framed it as a sign of rapid thinning and undercutting according to the Authoritative Thwaites cavity assessment. The cavity evidence fits with the Icefin view of concentrated melt in hidden pockets rather than smooth, uniform thinning.
At the same time, scientists working beneath the Thwaites Eastern Ice Shelf, or TEIS, have found places where melt is less aggressive than some standard formulas would suggest. Oceanographic measurements beneath TEIS show that strong stratification and weak currents can suppress basal melt rates by limiting the upward transport of heat according to a Primary study of suppressed melting under TEIS. The same research concluded that the widely used three‑equation melt formulation overpredicts basal melt rates under these stratified conditions.
That apparent contradiction between intense melting in crevasses and suppressed melt under stratified water helps explain why some earlier models have struggled. Many projections rely on simplified melt schemes that apply a version of the three‑equation formulation across large areas. The TEIS results show that, in some zones, those schemes run too hot, while the Icefin data indicate that the same broad averages can miss very high melt rates in narrow features according to the combined Primary work on Thwaites.
Longer‑term geological records add another layer. Seafloor evidence around Thwaites shows that the glacier has experienced episodes of rapid grounding‑line retreat in the pre‑satellite era when ocean and climate conditions allowed, according to a Primary reconstruction of past Thwaites behavior. That history supports the idea that the high melt and retreat rates now being measured are physically plausible responses rather than statistical outliers.
Looking ahead, a peer‑reviewed synthesis of recent observations and calibrated models found that the latest measurements of acceleration and ice loss at Thwaites are consistent with high mass‑loss trajectories over the next 50 years according to a Primary analysis of recent Thwaites trends. The study links observed changes to scenarios in which the glacier contributes more strongly to sea‑level rise than in lower‑loss cases, although it does not specify a single fixed outcome.
What remains unresolved for Antarcticas doomsday glacier is melting faster than
Even with direct robot footage, radar time series, and seafloor records, scientists still lack a clear answer to how fast Thwaites will retreat in the coming decades. One open question is how representative the extreme 43 m per year crevasse melt rate is across the grounding zone according to the Primary Icefin work. Icefin sampled specific channels and terraces, but there is insufficient data to determine how many similar hotspots exist under the wider glacier.
Another unresolved issue is how to reconcile the suppressed melt rates beneath TEIS with the rapid grounding‑line retreat seen in COSMO‑SkyMed data between 2016 and 2017 according to the Primary COSMO analysis and the TEIS stratification study. The TEIS observations show that ocean stratification can protect parts of the ice shelf, while the grounding‑line record shows that other sectors of Thwaites and TEIS experienced a doubling of retreat rate for 2011 to 2017 versus 1992 to 2011. The balance between these stabilizing and destabilizing zones will shape how the glacier evolves.
Model physics also remain a work in progress. The finding that the three‑equation melt formulation overpredicts basal melt under stratified conditions at TEIS according to the Primary TEIS study suggests that some projections may have exaggerated melt in protected areas. At the same time, the heterogeneous melting detected by Icefin indicates that models can underestimate melt in narrow crevasses if they smooth conditions over large grid cells according to the Primary Icefin observations. Improving both sides of that equation is essential for credible forecasts.
For coastal planners and insurers, the practical consequence is a wider range of plausible futures. Recent peer‑reviewed work that ties observations to high mass‑loss trajectories over the next 50 years according to the Primary Thwaites projection study signals that conservative sea‑level assumptions may no longer be safe. While there is insufficient data to determine an exact timeline for specific thresholds, the combination of rapid grounding‑line retreat, measured sub‑ice cavities, and extreme localized melt argues for planning that takes higher sea‑level scenarios seriously.
The latest publicly available updates on Thwaites come from these Primary field and satellite campaigns rather than from continuous monitoring networks. Until more real‑time systems are in place, the next key signals to watch will be new Icefin‑style under‑ice missions, fresh COSMO or similar radar analyses of the grounding line, and revised models that incorporate stratification effects and crevasse‑scale melt. For readers far from Antarctica, the technical terms translate into a simple stake: how quickly seas rise around cities from Miami to Mumbai will depend in part on how fast the “doomsday glacier” responds to the ocean heat now reaching its base.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
