⚡ Fresh stories daily

JITTERSNAP

News from the edge of reality

← Back to feed
Edge of Extinction

Antarctica's Ice Cliffs Fall Fast, Then Stop

A survey of 15,330 ice cliffs found they collapse faster than models assume — but the runaway retreat those models feared did not follow.

By JitterSnap•2026-11-20•3 min read

The Cliff That Should Have Run

Antarctica's ice shelves are thinning and breaking up. As they retreat, they expose walls of ice hundreds of metres tall at the edge of the grounded sheet.

For years one scenario dominated the worried literature: marine ice cliff instability. If a cliff grows past a critical height — often quoted around 100 metres — its own weight should make it fail. The collapse would expose a taller cliff behind it, which would fail in turn, and retreat would feed itself inland.

A study in Science has now tested that idea against a far larger set of real cliffs than the models were built on.

15,330 Walls of Ice

Caroline Needell, Catherine Walker and Jeremy Bassis measured 15,330 ice cliffs around Earth's ice-sheet margins using laser altimetry from NASA's ICESat-2 satellite.

Their result is neither a reprieve nor a confirmation. Both halves matter.

Tall cliffs, they report, calve faster than existing parameterisations assume, which implies models may underestimate ice loss on sub-annual timescales. But the retreat did not run away. Dynamic thinning, and retreat into regions where the ice-thickness gradient is low, kept the process from feeding on itself.

Cliff height also turned out to be far from the only control. Height correlated strongly with melange thickness — the pack of broken ice floating at the cliff foot — with water depth, and with crevassing.

What That Means for Sea Level

The practical read is narrower than headlines usually allow.

  • Faster calving: short-term losses at tall cliffs may exceed what current models project.
  • No demonstrated runaway: in this dataset, collapse did not cascade indefinitely.
  • Not a clean bill: the study constrains one mechanism. It does not clear the ice sheet, and it produces no sea-level number.

Instability now depends on more than a height threshold. Melange, bathymetry and the shape of the ice inland all enter the calculation, which makes projections harder to compress into a single figure.

The cliff fell. The glacier stopped. That is the part nobody modelled.

The Timeline

Antarctica's contribution to sea level is a slow variable punctuated by fast local events. This study suggests the fast events are faster than assumed, and the slow catastrophe is less automatic than feared. Both corrections push research toward better local observation: what is actually holding each cliff's foot, and for how long.

Speculation note: turning this into a revised sea-level range for 2100 would require model runs the paper does not present.

#antarctica#sea-level#ice-sheets#climate#calving

MORE FROM EDGE OF EXTINCTION

⏎ Back to all stories