OPINION:
Antarctica’s Thwaites Glacier is changing. It is thinning, retreating in places and losing ice to the ocean. Those observations are real. Still, instability is not the same thing as collapse, and the difference matters.
Recent media coverage has leaned heavily on projections suggesting that the Thwaites Glacier could lose on the order of 200 gigatons of ice annually by 2067. Presented without context, that number sounds catastrophic. In context, it is not.
The Thwaites Glacier is estimated to contain roughly 600,000 gigatons of ice. A 200-gigaton annual loss would amount to about 0.033% of its total mass. Even if such a rate were sustained, which itself depends on model assumptions, then complete disintegration would unfold over centuries, not decades.
Scale matters. So do percentages.
As summarized in the report “Climate at a Glance — Antarctic Ice Melt,” Antarctica holds enough ice to raise global sea levels by nearly 200 feet. Yet the continent is vast and dynamic. While parts of West Antarctica are losing mass, other regions have experienced net ice gains. The overall Antarctic system does not behave as a single uniform block sliding toward the sea.
Understanding the Thwaites Glacier also requires understanding what is actually melting it.
The International Thwaites Glacier Collaboration deployed instruments beneath the ice shelf and directly measured intrusions of Circumpolar Deep Water — ocean water roughly 34 degrees — flowing under the glacier and melting it from below.
The mechanism is primarily oceanographic. It involves shifting ocean currents transporting subsurface heat beneath the ice shelf. That distinction is important. The dominant process here is not surface air temperature melting the glacier from above. It is warm seawater interacting with the glacier’s underside.
The delivery of that water is influenced by complex wind patterns, changes in ocean circulation and seafloor topography. Reducing this to a simple atmospheric warming narrative obscures the actual physics.
None of this minimizes the significance of what scientists are observing. The Thwaites Glacier is retreating in some sectors. Ice shelves are fracturing. Basal melting is measurable. These are serious changes that warrant continued monitoring, but the data does not prove an inevitable collapse on a specific timetable.
To move from “instability is present” to “collapse by midcentury” requires speculative computer modeling. Ice sheet systems are nonlinear and sensitive to thresholds that are not yet precisely defined. Whether retreating becomes self-sustaining and irreversible on human timescales depends on assumptions about ocean forcing, grounding-line dynamics, and feedback mechanisms, which remain active areas of research.
That is why projections vary. Different parameter choices produce different outcomes. The observational record establishes cause for scientific concern; it does not, by itself, establish a countdown clock.
It’s also worth remembering that ice sheets have advanced and retreated repeatedly over geological time. The West Antarctic Ice Sheet likely collapsed and reformed during past warm periods. These systems operate over centuries and millennia. They are capable of substantial change, but rarely on abrupt, cinematic timelines.
Sea level rise itself reflects this reality. Global mean sea level has been increasing by roughly 0.12 inches per year over the past few decades. Even high-end assessments in the Intergovernmental Panel on Climate Change Sixth Assessment Report describe the recent rise in terms of feet, not dozens of feet, and emphasize the uncertainty surrounding dynamic ice-sheet contributions.
The Thwaites Glacier deserves attention and continued scientific study. What it does not deserve is the conflation of measured instability with predetermined collapse.
Instability is a condition. Collapse is a conclusion. Science shows the former. It has not yet established the latter.
• Anthony Watts (awatts@heartland.org) is a senior fellow for environment and climate at the Heartland Institute in Schaumburg, Illinois.
Please read our comment policy before commenting.