New York --:--
For Companies
Climate

About 84 million cubic yards of rock came off an Alaska fjord wall and drove a landslide tsunami almost a third of a mile up the opposite slope, but the only things lost were a kayak and a camp

By SEP 12, 2026 7:50 PM 5 MIN READ
Tracy Arm tsunami landslide scar and stripped fjord wall in Alaska, after 64 millionTracy Arm tsunami landslide scar
Stay in the loopFollow us on Google News

A narrow fjord in southeast Alaska, gray water, cliffs on both sides, low cloud sitting on them.

High on one wall there is a line where the forest simply stops.

Above it, trees. Below it, bare scoured rock and splintered stumps, for a thousand feet down to the water.

Nothing burned and nothing was cut. Something came through and took the mountainside off.

It happened before six in the morning.

There was nobody there.

Why a retreating glacier drops a mountain into the sea

A glacier filling a valley does more than sit in it. It presses outward against the rock walls on either side.

That pressure holds fractured slopes in place, and geologists call the support buttressing because the ice works the way a buttress on a cathedral does.

When the ice thins and pulls back, the support goes with it, and a slope that has been held for centuries is suddenly free to move.

The rock does not fall the moment the ice leaves. It creeps, opens its fractures, and finds the surface it will eventually slide on.

Here the glacier below the slope retreated sharply across a single week at the end of July, and the wall came down five days later.

The trigger was the ice leaving, not an earthquake arriving.

What the fjord looks like now

The mass that came off is put at roughly 84 million cubic yards, which is enough to bury a square mile of ground more than 50 feet deep.

Its crown sat over 3,280 feet above the water, so the whole block had that far to accelerate before it hit anything.

The landslide tsunami it pushed out broke at around 328 feet where it left the impact zone.

Two miles away, directly across the fjord, water climbed the opposite wall to about 1,578 feet above sea level and stripped it bare.

An island further down the arm lost nearly all of its vegetation, and the trimline left behind is visible from a boat today.

The scar is permanent and the fjord now has a different shape.

Where the numbers come from and what they rank against

The event carried no earthquake, but it shook the ground hard enough to register across the regional seismic network.

Seismologists put the energy at the equivalent of a magnitude 5.4 earthquake, which is how the timing was fixed to the minute at 5:26 in the morning.

A federal tide gauge in Juneau, tens of miles away, logged a maximum of about 14 inches above the normal tide, with wave activity continuing for hours.

The runup figure of 1,578 feet is the second highest ever measured anywhere, behind a 1958 event in another Alaskan bay that reached about 1,730 feet.

The work was led from a Canadian university and published with eighteen authors, which is what a measurement of this kind takes.

One instrument network gave the clock, and field survey gave the height.

The days of warning nobody was listening for

Here is the part that matters beyond this one fjord, and it was found in the seismic record afterward rather than at the time.

For days before the collapse the slope was producing small signals, each roughly the size of a magnitude 1 or 2 tremor, arriving about once an hour.

Six hours before failure the interval collapsed to every 30 to 60 seconds.

In the final two hours the signals merged into continuous vibration, which is a slope sliding on its failure surface rather than cracking.

That escalation is a recognizable pattern, and it means a slope of this kind can announce itself if somebody is watching the right instrument.

Careful measurement keeps replacing folklore, the way a catalog of sightings found no earthquake link for giant oarfish.

The volume, the runup and the agencies involved are set out by the federal survey.

A readable warning is worth nothing without a monitored slope.

How close it actually came

This fjord is a cruise destination, and boats carrying hundreds of passengers work the same water most summer days.

Multiple ships had been in the arm the day before. On the morning itself the weather was wet and the hour was early, so nobody was ashore.

A seismologist on the study described it plainly as dodging a bullet, which is unusually direct language for a scientific coauthor.

What was lost was a kayak and a camp belonging to paddlers at the mouth of the arm, whose gear was swept off the shore.

Tour operators have since stopped running the fjord altogether and moved their trips to the arm next door.

Water can turn dangerous in places people treat as scenery, the way warm water spins up funnels under ordinary fall clouds.

The precursor signals and the warning about retreating ice are described by the Alaskan institute.

The scar is still shedding rock, and the next one will land in the same water with nobody watching.

Read the whole thing?

Get the week's signal, not the noise

Our sharpest reporting on energy, climate and nature — free, once a week.

Share this article
Hugo_writer
Hugo is an engineer with strong technical expertise and deep knowledge of the space industry. Multilingual from an early age, his writing combines technical clarity with a strong interest in science and energy.