A Landslide in Alaska Set Off a Tsunami. There May Be More to Come.

Nearly 500 feet up a near-vertical rock face, scraped clean of soil and alder trees, Bretwood Higman, a geologist, looked down across the Tracy Arm fjord in southeast Alaska at a scene of devastation.

At 5:26 a.m. on Aug. 10 last year, a mass of rock with a volume 24 times larger than that of the great pyramid of Giza crashed down the mountainside, sending a wave of water 1,578 feet up the opposite wall and setting off a tsunami that roared down the fjord. It swept over the ridge that Dr. Higman was now standing on. The whole thing took about a minute.

Dr. Higman was part of an international team investigating the aftermath of the geologic event, the second largest landslide-generated tsunami on record. Using computer models, the researchers were able to recreate the landslide and tsunami, as well as a standing wave called a seiche that sloshed back and forth for 36 hours after the landslide.

Among other things, the new study, published on Wednesday in the journal Science, revealed how tricky it is to predict such catastrophic landslides before they take place.

The Tracy Arm landslide was preceded by an unusually rapid retreat of the South Sawyer Glacier, leaving the rock slope that ultimately collapsed bare and unsupported. That same rearrangement of land elements is increasingly occurring throughout Alaskan fjords and around the world. As glaciers retreat and thawing permafrost lubricates slopes, these giant landslides may become more frequent.

Scientists have been sounding the alarm about the emerging hazards of climate-linked giant landslides in Alaska for years. In 2020, Dr. Higman discovered a slow-moving landslide in the Barry Arm fjord that he worried could collapse catastrophically and inundate the nearby town of Whittier with a tsunami.

“We’re rapidly approaching a totally new landscape that has way fewer glaciers in the Alps and really everywhere, and a lot of new lakes,” said Mylène Jacquemart, a glaciologist at ETH Zurich who was not involved in the study. She studied the Blatten landslide that buried a Swiss village in rock, ice and water last year.

Jackie Caplan-Auerbach, a seismologist at Western Washington University and an author on the new study, was among the first to hear about the tsunami: Her neighbors, whose boat was anchored at sea some 50 miles from the landslide, texted her about a strange surge of water that had hit their vessel. Other firsthand accounts trickled in from Harbor Island, where camping kayakers said their gear had been carried away by the wave, and from a 150-passenger cruise ship, the National Geographic Venture, that was sitting just outside the fjord.

The stakes are high for detecting these events ahead of time. Although no vessels were in Tracy Arm fjord proper when the landslide hit, that mostly came down to luck: It was early morning and not many boats were about.

But three large cruise ships carrying thousands of passengers and numerous small tour boats visit the fjord daily, ferrying tourists right up to the glacier’s calving face. Had the Venture been up the fjord, instead of at its mouth, the wave would have been “unsurvivable,” Dr. Higman said.

Increased cruise-ship tourism to glacial fjords, and more oil and gas exploration in the Arctic, mean “we, as a global society, are putting more infrastructure and people in harm’s way,” said Dan Shugar, a University of Calgary geomorphologist and the study’s lead author.

Scientists at the U.S. Geological Survey and its state counterpart in Alaska look for slopes along the vast coastline that are moving toward collapse, using satellite radar and optical imagery. Because resources are limited, only Barry Arm is monitored in real time, with on-the-ground scientific instruments. Detailed assessments of a handful more moving slopes are underway in Glacier Bay National Park, which is also frequented by cruise ships.

Dr. Shugar said Tracy Arm “throws a wrench in” the strategy of looking for slope deformation “because it happened, as far as we can tell, without much warning.” Scientists were unable to detect any deformation in the slope before the collapse.

But when Dr. Caplan-Auerbach dug deeper into the seismic data from the landslide, she noticed patterns of land movement similar to those that she had studied for decades, which sometimes preceded landslides on volcanic slopes.

These tremors were “probably tiny bits of slip on the base of the landslide, and it can do that only so much before it’s got to break apart and fall,” she said. Tiny coalescing fractures within the mountain eventually reach a crescendo, a threshold at which point the rock can no longer hold itself together, and the slope gives way.

It is not yet clear how many landslides display these seismic signals as precursors, but since there were no other cautionary signs, they provide a hope of early warning. If the signals are subtle — the slopes “whispering to us, not yelling,” as Dr. Caplan-Auerbach put it — it is possible the seismometer network in Alaska is spread too thin to typically pick them up.

“The bar is, can we do better than missing most of these,” said Noah Finnegan, a geomorphologist at the University of California, Santa Cruz, who was not involved in the study. “So getting a handle on why these precursors happen and what their relationship is to catastrophic collapse is an area many people are interested in.”

Last month, three cruise lines notified customers that their ships would not visit Tracy Arm this year, opting instead for nearby Endicott Arm and Dawes Glacier. “That’s probably a wise move, but there’s no reason why Endicott is any safer than Tracy,” Dr. Shugar said.

Stephen Hicks, a seismologist at the University College London and an author on the study, said the planet is entering a new era when warming has penetrated geology.

“When we think about climate change, we think about impacts in the atmosphere and rising sea levels,” he said. “We sort of look up and across, but we don’t often look down,” he added, but now “the ground has moved beneath all our feet.”

Many open questions remain. But among the biggest, Dr. Higman said, is whether we can expect a significant increase in such events as a result of climate change, as some studies suggest.

“If it’s a dice you roll every 50 years, well, maybe that’s all right,” he said. “But if it’s one you’re rolling once or twice a year, then this is really, really urgent.”