WWU professor Jackie Caplan-Auerbach co-publishes paper on second-largest tsunami ever recorded

As the sun came up on August 10, 2025, Jeffery and Christine Smith, the owners of a small cruise ship anchored in a cove within the Endicott Arm of Alaska’s inside passage, noticed an irregularity. They should have seen calm waters and a receding tide; instead, their vessel was rising and falling as much as 10 to 15 feet over water surging in over the sandbar. Neither is a scientist by trade, but fortunately, their neighbor and friend of over 20 years would help solve their mystery.

Western professor and seismologist Jackie Caplan-Auerbach had just gotten up for her morning coffee when she was contacted by the Smiths via Starlink and learned of the unusual waves. The Smiths had already confirmed that no earthquake had occurred to cause a tsunami, so the next likely story was a landslide in the area. Caplan-Auerbach found the nearest seismometer was far away from the Smiths, so identifying one lone wiggle that could confirm a landslide was proving difficult.

“I study weird seismic signals that are accidentally recorded. And one of the signals that I've studied for the past 20 years or so occurs in association with landslides. Some slides, when they're about to fail, exhibit a slight stuttering behavior prior to failure. We call that a precursory earthquake sequence," said Caplan-Auerbach. “They're not big earthquakes. They’re tiny, tiny, tiny, little stuttering events, and it's not a hugely common signal. We've probably seen it about 15 to 20 times total. After doing some processing of the Tracy Arm data, I saw that there was a booming precursory signal. And I knew at that point that it was a landslide.” 

Once Caplan-Auerbach confirmed the landslide, the next step was contacting the United States Geological Survey (USGS) and the Coast Guard to get the word out. The Smiths told Caplan-Auerbach about a report of kayak campers who had their gear swept away by a large wave, which indicated that this event was potentially big enough to endanger people’s lives.

“The landslide analysts in Alaska did some analysis to figure out where this signal had come from, and they were able to say that it had taken place not in the fjord where Jeffrey and Christine were anchored, but in the next one up,” said Caplan-Auerbach. “And that told us two things: that there was a big event in that area, and that it had to be of a pretty decent size to make a wave that we estimated went somewhere around 47 or 50 miles between that slide and Jeff and Christine.”

On Wednesday, the 13th, scientists with the USGS took a helicopter flight to survey the area and confirmed that the landslide had occurred off the South Sawyer Glacier in the Tracy Arm fjord. Further study showed the tsunami wave was so large that it reached >1,570 feet up the side of the mountain, nearly rivaling the 1,720-foot wave triggered in Lituya Bay, Alaska, in 1958. When the wave crashed back into the fjord, it stripped vegetation and trees up to 300 feet on either side. Miraculously, due to the timing of the event, no one was reported injured.

“Had anybody been there at that time, I think we would be telling a very different story today. It happened at 5:30 in the morning, Alaska time, and nobody had yet entered this fjord to go view that glacier, which they do many, many, many times a day. On any given day, there could be between five and 25 ships in that fjord,” said Caplan-Auerbach. “We know that there were people who had planned to be in the Tracy Arm on that day, and for a variety of reasons, had said, ‘we're gonna stay somewhere else tonight.’ So we just got ludicrously lucky. Since this event, two major cruise lines have stopped their cruises to the Tracy Arm next year.“

Liam Toney, a research geologist in USGS’s Landslide Hazards Program, noted the difficulties in detecting and reacting quickly to landslides like the Tracy Arm event. Some landslides, like the Barry Arm landslide in south-central Alaska, are possible to track from space with satellites due to observable slow movements that might signal eventual failure. However, not all landslides are as easy to monitor.

“Landslide monitoring looks a little bit different from monitoring other types of hazards. You could draw an analogy to volcano hazard monitoring. The main difference is that with volcanoes, we already know where they are. Because we know where they are, we can deploy tools, methods, and instruments that take advantage of that. And with landslides, we don't necessarily know where they are going to fail before they do, and that means that we have to employ a different approach, and that can make things a bit more challenging,” said Toney. “There's no one recipe that you can employ that's going to work for all types of landslides. But by studying this landslide, especially the seismic signals that it produced prior to failure, we might be able to understand how we could monitor some subset of these landslides that have a similar failure behavior.”

The South Sawyer Glacier has been both thinning and retreating over the years, and there’s a possibility that the glacier’s retreat had de-buttressed that side of the fjord. Caplan-Auerbach said it’s very likely the landslide was caused by instability resulting from the glacier’s thinning.

“This is happening all over Alaska. It's happening all over the world. We know glaciers are retreating. We know they're thinning. We know that's caused by human-induced climate change. All that's not in question. The question is, to what extent might we expect to see an increase in these [landslides]? And we don't totally know,” Caplan-Auerbach said.

This was the driving force behind the collaborative paper co-authored by Caplan-Auerbach alongside 18 other scientists and researchers and recently published in Science, one of the most prestigious scientific journals in the world. The research presented in the paper provides a comprehensive analysis of the landslide and the resulting tsunami, and emphasizes the potential danger of similar future events at less opportune times. Caplan-Auerbach hopes that this paper will inspire future research into these events.

“It has motivated a multi-disciplinary attack on this problem. What was really exciting for me about this paper was that I study earthquakes, and I do a little bit of landslides. The lead author of this paper studies geomorphology, landscape studies, and glaciers. We have people who do tsunami modeling. We have people who are fundamental landslide modelers. We have people who study climate science. There are so many different scientists who worked on this paper,” said Caplan-Auerbach. “And that brought a lot of us together to say, what can we do to study these things from all these different directions? Science is always better with more voices, and always better when each of those voices brings a perspective that is either scientific, cultural, or regional, that is different.”

For the foreseeable future, fjords like Tracy Arm will likely see fewer visitors like Jeffery and Christine Smith entering their waters. But with scientists and researchers like Caplan-Auerbach drawing more attention to Alaska’s landslide risks, there may yet be future improvements to predicting and mitigating these events. 

Ava Nicholas is the 2025-26 Tahlequah Intern, a paid internship made possible by a generous gift through The Foundation for WWU & Alumni and overseen by the Salish Sea Institute and the Office of University Communications. She is a Journalism-Public Relations and Theatre major, with an Honors minor.