A thawed slope turns a tremor into a landslide
A PNAS study from early September reconstructs how a 6.5-magnitude quake off Arctic Norway set off a giant rock avalanche that buried a glacier. The new ingredient is not the earthquake but the decades of warming that had already weakened the frozen slopes waiting to fail.
On 10 March 2025 a magnitude 6.5 earthquake shook Jan Mayen, a small volcanic Norwegian island halfway between Iceland and Greenland. Seismically the island is no stranger to shaking, and earthquakes there are common. What was unusual was what happened next: within minutes a large rock avalanche tore down a volcanic slope and spread across the surface of the Kjerulf Glacier, burying the ice almost to the shore, and satellite data recorded a separate outbreak of ice breaking away from the neighbouring Weyprecht Glacier. A study published in PNAS in early September argues that the earthquake alone does not explain the scale of the destruction. The slope had been quietly primed for decades.
The argument is about what holds a steep Arctic mountainside together. In permafrost terrain, the cracks and fissures in the rock are bound by frozen water, ice that acts like cement in concrete. That ice has stabilized these volcanic slopes for thousands of years. As the climate has warmed, the permafrost has degraded, and the binding ice in the fractures has thinned. The rock has not fallen yet, but it is already weaker than it was. When a strong earthquake arrives, the shaking that an intact, ice-cemented slope would have absorbed instead sets the weakened rock in motion. The earthquake is the trigger. The warming is the reason the trigger had anything to pull.
The reconstruction is an unusually careful piece of detective work. The international team, led by Guilherme de Melo of the GEOMAR ocean research centre in Kiel, combined local and regional seismic records, models of how the shaking travelled, high-resolution satellite imagery and infrasound measurements to pin down the exact moment the slope failed, alongside air temperature and climate records. The result is a minute-by-minute account of a cascade: ground shaking, rock failure, debris runout across the glacier, and then the calving on the next glacier over. No single data source could have shown all of that.
The most telling comparison is what did not happen earlier. Forty years of satellite observations over Jan Mayen show no earthquake-triggered avalanche of comparable extent, even though the region has been seismically active throughout that period. The authors read that absence as evidence that something has changed in the slopes themselves, and the most plausible candidate is long-term permafrost warming.
The wider point is about cascading hazards in a warming world. A thawed slope does not have to wait for a landslide to come to it. It can fail under a perfectly ordinary earthquake, a routine storm, or a sudden melt pulse, and the debris can then bury a glacier, choke a river or bury a road that nobody thought was at risk. The work extends the familiar worry about Arctic thaw, which is usually told in terms of carbon released or coastlines eroding, to the more concrete problem of where the next mountain will move when the ground beneath it is no longer frozen solid.
Jan Mayen is uninhabited apart from a weather station crew, so this particular cascade cost no lives. The lesson is portable. Any steep, frozen, earthquake-prone slope on Earth is, by this logic, more fragile than it used to be, and the shaking that used to be harmless now comes with a longer list of things it might set off.