Models of Glacial Collapse in Antarctica Improve Predictions of Sea

With approximately one third of the world's population living within a day's walk of a coastline, sea-level rise poses a serious global threat. Every centimeter of sea-level rise could displace 1.5 million people, and current estimates of rise range from half a meter to 2 meters by the end of the 21st century—which means that potentially hundreds of millions of people could be affected. A more reliable prediction of sea-level rise is crucial to prepare coastal infrastructure and communities around the world for these changes.
The Pine Island Glacier, located in West Antarctica, is the fastest flowing glacier on the continent and is the region's largest contributor to sea-level rise. Its steady slide into the ocean is slowed by the Pine Island ice shelf, the glacier's floating edge on the sea. Ice shelves act like a cork in a bottle, providing what is known as a buttressing stress on the heavy body of the glacier, to hold back its flow.
In 2017, a large chunk of the Pine Island Ice Shelf broke off into the ocean to form an iceberg in what is called a calving event, and the glacier has since increased its speed of flow by 20 percent. Understanding when and where other ice shelves may calve into icebergs is critical for predictions of sea-level rise.
In new work, Caltech researchers used nine years of satellite data on the flow of Antarctic glaciers to measure and model how and why ice shelves like Pine Island's collapse.
The study is described in a paper appearing in the journal Proceedings of the National Academy of Sciences on August 11. The work was conducted in the laboratory of Brent Minchew (PhD '16), professor of geophysics.
"The potential population displacement caused by sea-level rise is on the order of the displacement seen during World War II, except sustained over entire lifetimes," Minchew says. "That scale of disruption historically comes with conflict. Coastal adaptation is costly, and roughly 90 percent of people at risk live in low- to mid-income countries. It is critical to understand the future scenario we're facing so that we can most efficiently use resources to prepare."
The behavior of glaciers, which cover thousands of square kilometers and persist for thousands of years, is much more complex than that of the ice cubes in a cold drink or ice in the lab, making it challenging to model. Over time, glacial ice accumulates tiny defects in its crystalline structure that add up to create larger damage, like jagged crevasses that can extend hundreds of meters into the ice. Glaciologists aim to understand the exact processes that cause the collapse of such massive long-lived structures.
"In 2002, the Larsen B ice shelf—which had an area comparable to Rhode Island—in Antarctica collapsed over the course of just six weeks, leading to its glaciers speeding up four- to sixfold," says Sarah Wells-Moran, lead author on the new paper and a former undergraduate and master's student in Minchew's previous lab at MIT. "A lot of the uncertainty in sea-level rise projections comes from our lack of ability to accurately model when these ice shelves are going to collapse because we have gaps in our knowledge of the material properties of ice. How and when does it break? How does it sustain damage?"
The Pine Island Glacier is currently flowing into the sea at a rate of about 4.8 kilometers per year, about one-hundredth the speed of the average garden snail. Though this seems like a literal "glacial pace," the glacier is currently the largest Antarctic contributor to sea-level rise.
In the new study, the team examined data taken by the European Space Agency's Sentinel-1 radar satellites over Antarctica from 2015 to 2024. The satellites measured the velocity of Pine Island's movement and showed that calving events at the ice shelf, particularly the one in 2017, increased the glacier's velocity by 20 percent. Overall, the velocity of flow has increased by more than 100 percent since 1973.
The study also examined the glacier's shear margins, the sides of the glacier where it rubs up against rocky Antarctic outcroppings. As with a thick fluid squeezing through a tube, the pressure on the sides of the glacier provides an important frictional force to slow its progress. The new study found that calving of the ice shelves at the front edge of the glacier weakened the shear margins at the side, reducing their ability to hold the glacier back. The data appear to show that the glacier had been completely decoupled from its damaged shear margins by 2020.
"The shear margins have visible damage in them prior to the 2017 calving event," Wells-Moran says. "After this calving event, the damage just intensifies. As the ice speeds up, that puts more force onto the shear margins because the ice is now going faster than the ice next to it, which generates more damage, which further weakens the margin and causes more speedup."
The findings provide a unique opportunity to test models for the retreat of Pine Island Glacier into the ocean. Conventional understanding suggests that Pine Island will undergo a multiyear period of retreat and increased mass loss because of the reduction of buttressing. But the specifics of how and how fast this retreat will occur are unclear. Models used to project sea-level rise need to show accuracy in predicting the responses to these changes before they can be trusted to inform coastal planning and adaptation.
"The difference between a future with manageable levels of sea-level rise and the extreme scenarios lies in the behavior of the West Antarctic ice sheet, specifically the evolution of Pine Island Glacier and its neighbor, the Thwaites Glacier," Minchew says. "How these glaciers evolve depends on how their floating ice shelves continue to buttress and resist the seaward flow of ice."
Minchew's group is currently studying possible methods to slow the rate of mass loss and actively stabilize the ice sheets by leveraging the glacier's own natural processes of healing structural defects.
The paper is titled "Near-total loss of buttressing stresses observed on Pine Island Ice Shelf, West Antarctica." In addition to Wells-Moran and Minchew, Bryan Riel (PhD '17) of Zhejiang University in China is a co-author. Funding was provided by the National Science Foundation.

Facts Only

* Approximately one third of the world's population lives within a day's walk of a coastline.
* Sea-level rise could displace 1.5 million people due to displacement caused by sea-level rise.
* Current estimates of sea-level rise range from half a meter to 2 meters by the end of the 21st century.
* The Pine Island Glacier, located in West Antarctica, is the fastest flowing glacier on the continent and a major contributor to sea-level rise.
* The Pine Island ice shelf slows the glacier's flow by providing buttressing stress.
* A large chunk of the Pine Island Ice Shelf broke off into the ocean in 2017 in a calving event, increasing the glacier's speed by 20 percent.
* Caltech researchers used nine years of satellite data on Antarctic glacier flow to model ice shelf collapse.
* Calving events increased the glacier's velocity by 20 percent.
* The study examined shear margins where the glacier rubs against rocky outcroppings, finding calving weakened these margins.
* The glacier had been completely decoupled from its damaged shear margins by 2020.

Executive Summary

Sea-level rise poses a significant global threat, with every centimeter of rise potentially displacing 1.5 million people, and current estimates project rises between half a meter and two meters by the end of the 21st century, affecting hundreds of millions. Understanding the future scenario requires reliable predictions to prepare coastal infrastructure. The Pine Island Glacier in West Antarctica is a major contributor to sea-level rise; its flow is influenced by the Pine Island ice shelf, which acts as a buttress slowing the glacier's movement. A calving event in 2017 caused the iceberg formation and a 20 percent increase in the glacier's flow speed. Researchers used nine years of satellite data to model how ice shelves like Pine Island’s collapse, and the subsequent changes, occur. The study found that calving weakened the shear margins, reducing their ability to hold back the glacier, a process accelerated when the glacier sped up. This research highlights that uncertainty in sea-level rise projections stems from the difficulty in accurately modeling the behavior of ice shelves and their response to structural damage.

Full Take

The analysis of glacial dynamics underscores a critical gap between observed physical processes and predictive modeling for large-scale environmental change. The central tension lies in the complexity of ice sheet mechanics—how small, accumulated defects in crystalline structure lead to catastrophic failure—which creates massive uncertainty in projecting future sea-level rise. The mechanism described, where calving reduces buttressing, accelerates flow through weakening shear margins, introduces a non-linear feedback loop that is difficult to quantify for large-scale projections. This highlights how uncertainties in local ice behavior translate into massive uncertainties regarding global coastal adaptation scenarios, especially given the historical context linking such disruptions to conflict and varying levels of socioeconomic vulnerability among at-risk populations. The focus on the Pine Island Glacier suggests that reliance on conventional models must incorporate dynamic feedback loops resulting from catastrophic calving events rather than static predictions of mass loss alone. What are the specific mathematical constraints or material property assumptions within current sea-level rise models that fail to account for this type of rapid, stress-induced structural failure observed in ice shelves? Does framing coastal adaptation solely around measured rates ignore the non-linear risk introduced by these geological uncertainties?

Sentinel — Human

Confidence

The text appears to be a high-quality summary of scientific research, effectively weaving complex geophysical concepts with specific observational data and expert commentary.

Signals Detected

Moderate sentence length variance and the inclusion of direct, cited quotes suggest human editorial pacing.

The text flows logically from a broad threat (sea-level rise) to a specific mechanism (glacier dynamics) and then to research findings, demonstrating integrated narrative structure.

Citations of specific researchers, journals (PNAS), satellite data sources (ESA Sentinel-1), and explicit methodological links provide high specificity typical of human reporting or academic summary.

The internal logic linking calving events, shear margins, and velocity changes is complex and specific, making simple LLM fabrication less likely, though the structure of the findings could be polished.

Human Indicators

Direct engagement with specific academic work (Citations to PNAS, names, lab affiliations) suggests grounding in specialized knowledge beyond generic synthesis.

The nuanced description of physical mechanisms (buttressing stress, shear margins, kinematic feedback loops) indicates domain expertise woven into the narrative.