A Mountain Fell Into an Alaskan Fjord and Water Climbed 1,580 Feet

A landslide that sent water 481 metres up an Alaskan fjord wall has prompted researchers to propose alerts for unstable slopes above waters used by cruise ships and kayakers. ^[1][3]

The tsunami struck Tracy Arm in southeastern Alaska early on August 10, 2025, after a mountainside collapsed near South Sawyer Glacier. Its run-up reached about 1,580 feet on the opposite wall, making it the second-highest tsunami recorded, according to the researchers. No one was caught in it. ^[4][5]

Michael West and Ezgi Karasözen, researchers at the Alaska Earthquake Center, say the retreating glacier, heavy rain and thousands of small tremors before the collapse could have supported escalating alerts if a monitoring program had existed. Their proposal turns the reconstruction of a near-disaster into a practical question: how much warning can these slopes give? ^[1][3]

Dan Shugar, who led the study published in Science on May 6, 2026, offered a more cautious assessment. He said the slide gave little warning, its seismic noise was too slight to attract attention, and the event came as a surprise, according to the University of Calgary. ^[4]

Whether the tremors could have been recognised in real time as a precursor to collapse, rather than understood afterward, is unresolved. The warning proposal is a case for building and testing monitoring, not evidence that this landslide could reliably have been forecast. ^[3][4]

A mountain falls into a narrow fjord

Tracy Arm lies about 80 kilometres south of Juneau. Its steep walls and glaciers draw cruise passengers, boaters and kayakers into a landscape where a collapsing slope can transfer enormous force directly into confined water. ^[3][4]

On the evening before the collapse, passengers aboard the Hanse Explorer visited South Sawyer Glacier and then left the fjord. The tsunami struck around 5:30 the following morning, when no ships were caught in the wave. Some kayakers woke to water passing their tent and carrying away a kayak and equipment, according to the University of Calgary. ^[4]

The Alaska Earthquake Center estimates that about 100 million cubic metres of rock failed. The collapse released seismic energy equivalent to a magnitude 5.4 earthquake. That comparison describes the energy recorded, not the cause of the tsunami: falling rock drove this event. ^[2]

The 481-metre measurement describes how high the water reached up the opposite slope. It is a run-up height, rather than the height of a wave travelling across an open ocean. The fjord's confined shape is central to understanding the extraordinary number. ^[3][5]

Shugar's team reconstructed the event using eyewitness accounts, satellite observations, seismic records and numerical modelling. Together, those methods allowed the researchers to examine the slope before it failed and trace what happened when the rock entered the water. The study was an observational reconstruction, rather than an experiment. ^[4]

The human outcome depended heavily on timing. The fjord had visitors the previous evening, but the collapse happened before ships arrived the next morning. The same physical event in an occupied stretch of water would have presented a very different emergency. ^[4]

The ice that held the slope

The study links the slope failure to glacier retreat. South Sawyer Glacier had occupied the foot of the mountainside, providing support that diminished as the ice pulled away. Researchers call that loss of support "debuttressing". ^[5]

Think of the ice as a brace against the lower slope. Removing it changes the forces acting on the rock above. Rain and water entering fractures may add stress to an already vulnerable mountainside. ^[3][5]

West and Karasözen say the glacier retreated more than a third of a mile in the two months before the collapse. Heavy rain also fell, and thousands of small tremors originated in the area during the preceding days. They see that combination as a potential basis for warnings. ^[1][3]

Shugar's assessment puts the difficulty squarely on detection. A signal visible during a detailed reconstruction may still be too slight to draw attention while a slope is moving. His account of limited seismic warning sits alongside the earthquake researchers' argument that a dedicated program could have made more of the signs. ^[3][4]

Glacier retreat makes the question relevant beyond one Alaskan fjord. Other parts of Alaska, Greenland and Norway combine steep coastal terrain with retreating ice. The Tracy Arm reconstruction gives researchers a specific failure to investigate as they assess similar slopes elsewhere. ^[3][5]

Three chances to raise an alarm

West and Karasözen propose a system with several levels of urgency. Glacier conditions and weather could support broad alerts, indicating that a slope deserves closer attention. Increasing seismic activity could then prompt a more urgent warning. ^[1][3]

A final layer would detect a landslide already under way. That would serve a different purpose from forecasting collapse: it would identify an event in progress and trigger a rapid alert. The proposal therefore spans both advance assessment and emergency detection. ^[3]

The researchers say the United States has no landslide monitoring system operating at the scale they propose. Establishing one would require testing which combinations of ice loss, rainfall and ground movement warrant action. A tremor count alone does not supply that decision. ^[1][3]

Travel operators have already acted on the hazard. In its April 2026 review, the Alaska Earthquake Center said two major cruise lines had removed Tracy Arm from their itineraries. ^[2]

The next task for researchers is to test the proposed warning levels against real slope behaviour. Agencies would then have to decide who monitors the signals, who issues alerts and what visitors should do when one arrives.

West and Karasözen have outlined the system they want built. It has yet to become an operating safeguard. ^[1][3]