Antarctic Peninsula Models Show 20% Sea Ice Loss at Highest Emissions, Muted Change at Low
Climate projections for Antarctica have long carried an unsettling quality: the uncertainties are large, the timescales are long, and the consequences - sea level rise, disrupted ocean circulation, ecosystem collapse - are planetary in scope. A study published in Frontiers in Environmental Science attempts to sharpen that picture by quantifying the difference between trajectories rather than simply describing worst cases. The result is a comparative portrait of two futures separated by emissions choices made in the next few decades.
The research team, led by Professor Bethan Davies of Newcastle University in collaboration with colleagues including Professor Peter Convey of the British Antarctic Survey and Professor Martin Siegert of the University of Exeter, focused on the Antarctic Peninsula - the finger of land extending northward from the continent that is among the fastest-warming regions on Earth and also the most extensively studied. They modeled outcomes under three emissions trajectories: a low scenario producing approximately 1.8 degrees Celsius of warming compared to preindustrial levels by 2100; a medium-high scenario reaching 3.6 degrees Celsius; and a very high scenario reaching 4.4 degrees Celsius. Eight environmental systems were assessed across those scenarios: marine and terrestrial ecosystems, land and sea ice, ice shelves, the Southern Ocean, atmospheric conditions, and extreme events such as heatwaves.
What Changes at Each Level
Under the low-emissions scenario, Davies described the future as difficult but manageable. The trends currently observable - glacial retreat, more frequent extreme events, some sea ice variability - would continue but at reduced rates. Winter sea ice coverage would remain close to current levels. Sea level contributions from the Peninsula would be limited to a few millimeters. Most glaciers would remain recognizable, with their supporting ice shelves still intact. The changes would be real, but their scale would allow ecosystems and infrastructure to adapt at a pace that avoids the most abrupt disruptions.
The very high emissions scenario produces outcomes of a fundamentally different character. Sea ice coverage could fall by 20% - a loss that cascades through the food web, because Antarctic krill, the foundation of the food chain for penguins, seals, and whales, depends on sea ice for larval development and winter refugia. If krill populations collapse, warm-blooded predators with some temperature tolerance may survive but face starvation if their prey cannot. The warmer Southern Ocean would accelerate melting of ice shelves - floating extensions of land-based glaciers that currently buttress those glaciers against faster seaward flow. When ice shelves collapse, the glaciers they supported can accelerate toward the ocean, contributing to sea level rise at rates that existing models struggle to constrain precisely.
The Current Trajectory
Davies described the current emissions trajectory as heading toward a medium to medium-high outcome - somewhere between the optimistic and pessimistic scenarios modeled. That assessment is consistent with projections from the Intergovernmental Panel on Climate Change, which have consistently placed current policy commitments on a path to roughly 2.5 to 3 degrees Celsius of warming by 2100. The Antarctic Peninsula responds to global temperature increases with amplified warming - the region has warmed roughly three times faster than the global average over the past several decades - meaning a global trajectory landing at 2.5 to 3 degrees translates to substantially more warming at the Peninsula.
Professor Convey, who first visited Antarctica in 1989 as a researcher stationed at Signy Station in the South Orkney Islands and has returned many times since, provided personal context for what change looks like from the perspective of field observation: what registers as unremarkable ice and snow to a first-time visitor is visibly and significantly reduced to anyone who has compared it against the same landscape decades ago. Long-term ecological observation captures dynamics that modeling alone cannot, and the study integrates both.
Why Peninsula Changes Do Not Stay in Antarctica
The study's framing explicitly positions Antarctica not as a remote concern but as a system whose changes propagate globally. Ice shelf collapse drives glacier acceleration and sea level rise that affects coastal communities worldwide. Freshwater influx from melting ice disrupts the Atlantic Meridional Overturning Circulation and other ocean current systems that redistribute heat across the planet. Changes in Southern Ocean temperature and chemistry affect the carbon cycle, because the cold Southern Ocean is a significant absorber of atmospheric CO2 - a role that diminishes as it warms. Extreme weather patterns in the mid-latitudes are partly driven by the temperature gradient between the poles and the tropics, a gradient that polar warming progressively reduces.
Professor Siegert noted that the 2019 iteration of this research characterized the Antarctic Peninsula's response to the 1.5 degree Celsius scenario; the 2026 update extends that analysis to scenarios now understood to be more likely, given where global emissions have gone in the intervening years. He described the prospect of exceeding 1.5 degrees for the Peninsula as frightening.
Research Infrastructure Under Threat
One element of the study's scope that rarely appears in Antarctic climate coverage is the effect of changing conditions on scientific research itself. The infrastructure that supports Antarctic science - research stations, transport networks, equipment calibrated for specific temperature ranges - is designed for conditions that are shifting. More frequent extreme weather events damage and disrupt that infrastructure. Scientists face greater difficulty accessing field sites as ice conditions change. The data collection that makes improved projections possible is compromised precisely when better data is most needed. The study identifies this feedback loop as a meaningful constraint on the scientific community's ability to reduce the uncertainties in future projections.
Davies' conclusion reflects the fundamental structure of the climate choice the study quantifies: changes under the high scenario would be irreversible on any human timescale. Glaciers cannot be rebuilt on decades-long schedules; wildlife that has moved south or gone locally extinct will not return. The gap between the low and high scenarios is not a difference in degree but in kind - and the window in which the choice can still be made toward the lower trajectory is finite.
Species at the Intersection of Multiple Stressors
The ecological consequences of the scenarios modeled in the study do not act through single mechanisms. Higher ocean temperatures affect Antarctic krill both directly - by changing the temperature range of their habitat - and indirectly, by reducing the sea ice that supports their larval development and provides refuge during winter. Krill are the pivot point of the Antarctic food web: they feed on ice algae and are themselves the primary prey of penguins, seals, whales, and many fish species. A 20% reduction in sea ice coverage under the highest emissions scenario does not simply reduce krill habitat; it compresses and fragments it in ways that interact with krill reproductive cycles, predator foraging ranges, and the spatial distribution of productivity across the Southern Ocean.
Penguin species face these pressures differently depending on their ecology. Adelie penguins, which are heavily dependent on sea ice habitat, fare worst under high warming scenarios. Chinstrap and gentoo penguins, more adaptable in their habitat use, may shift their range southward as warming opens new territories - but whether those new territories provide the prey resources needed to sustain populations is uncertain. The study flags these interactions as genuinely difficult to model at the detail needed to make species-specific predictions, which is why the emissions-scenario comparison is expressed in terms of ecosystem-level changes rather than species population forecasts.