Sasha Calvey heard the tsunami coming. It was after 5:30 a.m. on Aug. 10, 2025, and she was camped with two friends on a remote island in Southeast Alaska. They were 74 days into a sea kayaking trip from Washington.
Heavy rain had fallen overnight, with gusts that swayed the spruce and hemlocks overhead. From her sleeping bag inside the tent, Calvey heard waves tumbling against the shore. But it was an alarming new sound that roused her: Like a waterfall but deeper, and getting louder as it got closer until it almost shook the ground.
“Everything in my gut told me to check outside,” Calvey said.
The three friends scrambled out to find that 15-foot waves had just slammed into the nearby beach, shoving water into the forest. Seaweed dotted the moss beside their tent. A few steps away, water drained from the woods down to the gravel beach where they’d stowed their food and gear, half of which was now gone. One kayak was wedged against a tree. Another hung from a bluff, and the third floated in the distance.
They were lucky. Had the tsunami arrived three hours earlier, when the tide was 10 feet higher, water could have crashed through the area, toppling trees and sweeping the trio, trapped inside their tent, into the ocean.
“The water was chaotic for hours,” Calvey recalled, describing powerful swells and giant whirlpools. With so much gear missing, they radioed for rescue, ending their trip.
Later, they learned that the tsunami had been caused by a massive landslide roughly 30 miles away. It happened at the end of Tracy Arm fjord, a narrow channel that snakes between mountains south of Juneau. The slide came down where the South Sawyer Glacier is retreating; scientists believe that glacial retreat can contribute to landslides.
The tsunami wave climbed an incredible 1,580 feet high on nearby mountainsides — a measure called its run-up — stripping away trees, brush and soil. It was the second-largest run-up ever recorded. As the tsunami raced away from the slide, it scoured steep shorelines along the approximately 30-mile fjord. Its remnants hit Calvey’s beach, outside the fjord’s mouth.
Evidence suggests a sharp uptick in the number of landslide-generated tsunamis in Alaska over the last decade or more. In 2015, a slide at Taan Fjord, 400 miles northwest of Tracy Arm, launched a 630-foot run-up. And in Prince William Sound, a creeping landslide near the Barry Glacier threatens to launch a tsunami potentially even larger than Tracy Arm’s.
Scientists, land managers and tour operators are scrambling for answers. And they expect more slides as glaciers continue to shrink. This emerging climate-related hazard threatens communities, infrastructure and boats big and small, including the cruise ships that bring nearly 2 million visitors to Alaska yearly.
While researching her trip, Calvey found warnings about tsunamis in Alaska’s fjords. But they were generally tied to earthquakes, including the 1958 shaker that caused a landslide and 1,720-foot run-up in Glacier Bay National Park’s Lituya Bay. Nothing mentioned giant, tsunami-causing landslides triggered by retreating glaciers. But she hopes they get attention now.
IT’S HARD TO OVERSTATE the good fortune that day, considering Tracy Arm’s popularity: Each summer, it receives over 200 visits from lumbering cruise ships full of people hoping to see where the massive Sawyer and South Sawyer glaciers flow down to the ocean. The fjord, inside the Tongass National Forest, attracts an estimated 500,000 boat-based visitors annually. Dozens of vessels visited the area in the days before the slide, even as gravity quietly tugged at the soon-to-collapse mountainside.
But traffic is light during early mornings. The closest-known vessel was the over-100-passenger National Geographic Venture, which was in Tracy Arm and motoring toward the glaciers. In deep water and still about 20 miles from South Sawyer when the mountain slid, it was unharmed by the tsunami.
If the landslide had hit a few hours later, the Venture would have been at the glacier, where passengers climb onto skiffs to see it up close. In that zone, whitewater waves over 100 feet high full of icebergs and debris could have destroyed the tour boat and any already-launched skiffs.
“Everything in my gut told me to check outside.”
Christine and Jeffrey Smith, owners of a 65-foot tour boat called the David B, were also lucky. The previous evening, with six guests aboard, they had planned to anchor overnight five miles from South Sawyer. The tsunami would have struck them minutes after the slide. But because of stormy weather, they anchored 45 miles away in neighboring Endicott Arm instead.
“We’re definitely feeling our mortality,” Christine Smith told me later. It was hard to imagine surviving the chaos.
But even in Endicott Arm, the tsunami reached them. Cooking breakfast in the galley, Smith watched 10-foot swells suddenly overtake a shoal and an 80-foot vessel fight an abnormally strong current.
The Smiths sent a satellite text to Jackie Caplan-Auerbach, a geologist friend in Washington, asking her to check nearby seismic stations for signs of landslides. It was the first report of the Tracy Arm tsunami to the outside world.
Caplan-Auerbach was sitting down to coffee and a Wordle when the Smiths’ text dropped. The seismologist and Western Washington University professor quickly pivoted to scrolling for seismic stations near the David B. The data from many stations is publicly accessible. But there are few seismometers in Southeast Alaska, a vast and remote area where monitoring is expensive.
At a sensor in the Tlingit village of Angoon, 60 miles west of the David B, Caplan-Auerbach saw a telltale signal of a landslide somewhere close by. She texted colleagues, and soon a widening circle of scientists pinpointed its location in Tracy Arm.
“This is a big deal,” she recalled thinking. The landslide must have displaced a huge volume of water for the tsunami to travel 45 miles to the David B.
Caplan-Auerbach also noticed that for hours before the slide, the Angoon seismometer had been picking up small signals from the slope. “They’re tiny,” she told me, adding that no one would have noticed them had the landslide not followed. Geologists later spotted them on other seismometers, too. They stopped abruptly once the mountainside gave way.
Caplan-Auerbach has studied these “precursory” signals for nearly 20 years. She and others wonder whether they could inform a future landslide warning system. But any such system is a long way off. “Landslides are notoriously difficult to anticipate,” she said, explaining that rock type, precipitation and other wide-ranging factors complicate predictions. Plus, she added, precursory signals don’t precede every slide.
IN OCTOBER, Bretwood “Hig” Higman and I paddled tiny packrafts along the cliffy shores of Portage Lake, an hour southeast of Anchorage and 500 miles from Tracy Arm. Higman, an independent geologist working through his nonprofit Ground Truth Alaska, needed to check sensors on a slow-moving landslide above the Portage Glacier at the far end of the lake.
Dense fog blanketed the water. As we paddled toward the glacier, a rockslide rumbled somewhere high in the mist, a reminder that steep mountains loomed overhead. Higman, wearing wire-rim glasses and a fraying ballcap, cocked his head toward the sound with boyish interest.
In May 2020, Higman and over a dozen other scientists made their own big splash. They sent an open letter warning of a potentially catastrophic creeping landslide near the retreating Barry Glacier in Prince William Sound to The New York Times, which published a story about the dangers. It was an unorthodox move, and some locals wished they’d learned about the threat from an official announcement rather than from national media coverage that triggered a scare before tourism season.
The Barry Arm landslide, 40 miles north of Portage Lake, is intimidating. In summer, rocks tumble down its gravelly surface, kicking up dust or splashing into the ocean. Nearby, the Barry Glacier booms as it sheds ice. Modeling shows that if the whole slope went, it could unleash a huge tsunami threatening boats, cabins, campsites and all the people in them across a broad area. A diminished yet still powerful 6-foot wave could hit the small city of Whittier 30 miles away, endangering anyone along low-lying shores.
Soon after the researchers sent the open letter, state and federal agencies laced the Barry Arm fjord with helicopter-loads of high-tech instruments, including weather stations, seismometers, live-feed cameras, radio repeaters and ground-based radar. Together, they provide real-time observations of the unsteady slope. There’s also an experimental tsunami warning system that scientists can use to notify Whittier officials or alert boaters via VHF radio. In addition, research using the Barry Arm instruments has bolstered scientists’ ability to pinpoint more distant landslides, including the Tracy Arm event. The whole project represents millions of dollars in state and federal investment.
As we paddled Portage Lake, our reflections crisp on the water, Higman explained why certain of Alaska’s thousands of retreating glaciers stand out as hazards. The main ingredients are visibly unstable slopes perched above thinning ice; a body of water big enough to throw a tsunami; and proximity to people and infrastructure.
Like Barry Arm, Portage has it all. A Chugach National Forest visitor center sits lakeside and bustles with up to 1,000 visitors daily in summer. Many board the 140-
passenger Ptarmigan tour boat, which laps the lake throughout the day. There’s also a nearby road, railroad, campgrounds and trails, all in a tsunami’s potential path.
We landed at a wedge in the shore and hung our boats from a cliff so that curious bears wouldn’t mess with them. Shouldering packs in the cool air, we hiked up through thick alder brush, then steep green tundra. Far below, the fog dissipated, revealing the Portage Glacier, a giant river of ice between bulging mountains. Bergs floated where its face met the lake. But along its edges, rubble-darkened ice shrank away from mountain walls. The glacier looked sick.
It could be an example of “debuttressing,” Higman explained. Healthy glaciers erode the adjacent mountains, steepening or even undercutting their walls. But they also press against the slopes, supporting them.
When a glacier thins, as is happening globally due to climate change, the buttress disappears. That can cause steepened slopes to collapse. Climate change is also thawing alpine permafrost and worsening extreme rain events, which can trigger debuttressing-related landslides.
But the exact mechanisms that determine if or when a slope fails remain unclear. Landslides do not always immediately follow glacial retreat. And sometimes they don’t happen at all. Over the last decade or more, the Barry Glacier has pulled away from its unstable slope’s bottom, but the slope hasn’t slid. At least not yet.
“It’s frustrating that we don’t know more about how these systems work,” Higman said, looking down at where the wasting glacier met Portage Lake.
AS WE ANGLED HIGHER above the glacier, Higman pointed out thin cracks in the ground, evidence that the slope was crumbling. They’d opened so recently that living heather still straddled them, desiccated roots dangling in the air. More cracks appeared as we continued upward, until we were winding around seemingly bottomless chasms several feet wide. Far below, the Portage Glacier thundered as it dropped bergs into the lake.
The tundra abruptly ended at the landslide’s most active edge. Stretching above and below and wide as a football field, the slope was a mess of upturned rock and soil, with leaning spires, shattered boulders and deep, gravelly sinkholes. We heard the occasional hard knock of tumbling rocks. The whole thing was slumping down toward the glacier by several meters annually, Higman said. Other factors may contribute, but he believes the thinning glacier has debuttressed the entire slope.
He and others worry that it may collapse catastrophically into the lake in the next few years as the glacier recedes from underneath it. That’s what happened at Tracy Arm. Satellite imagery shows that the South Sawyer Glacier swiftly withdrew from beneath the doomed mountainside in the weeks and days before it collapsed.
As we paddled toward the glacier, a rockslide rumbled somewhere high in the mist.
Higman, Caplan-Auerbach and several other scientists, led by University of Calgary researcher Dan Shugar, recently published a paper in the journal Science that details how the glacier’s retreat primed the slope for the landslide and tsunami.
Still, many factors can contribute to landslides, including weather and slope weakness, said Dennis Staley, the U.S. Geological Survey scientist leading the multi-agency effort at Barry Arm. The recent research renews focus on the tie between retreating glaciers and landslides large enough to cause tsunamis, but, Staley told me, “It’s hard to know a specific trigger.”
Higman agreed, choosing his words carefully: “There could be many causes” for the Tracy Arm slide. But he called the timing of the glacier’s rapid retreat “an extraordinary coincidence” that raised imminent concerns about Portage.
We side-hilled across the steep, barren slide. Dirt and scree sloughed downward beneath our boots as we navigated around wobbly boulders many times my size. Eventually, Higman reached over the edge of a sheer column to retrieve an extensometer, a metal and glass canister smaller than a water bottle that he’d previously drilled into the rock.
“It’s like an electronic tape measure,” he said, using a drill to open its casing and change its battery. All day, it snaps measurements of the rock’s movement. It’s still experimental, but he hoped that someday he or perhaps visitor center staff could view its data online, including during periods of heavy rain or quickening glacial retreat.
Ideally, Higman said, more precise instruments would pepper the Portage slide, like at Barry Arm. They would link to satellites and a government office where someone could “push a red button” to alert the public if movement accelerates. The array could also detect any potentially critical
precursory signals.
Closing the extensometer, Higman described the benefits of his small homegrown effort. Low costs, for example, mean more instruments in more places. But the rudimentary sensors must be placed “at just the right crack” to show movement. For now, though, they’re the only sensors at Portage.
We scrambled upward to the crown of the slide and back onto firm tundra. For an hour, we traversed the slope to other sensors, each an eye on what Higman hoped were the right cracks.
ON A DRIZZLY EVENING a few weeks after the Tracy Arm tsunami, Staley, the scientist leading the Barry Arm monitoring, drove to Whittier for a public meeting. Dozens of companies launch boat and kayak tours into Prince William Sound from the town. Captains, guides, officials and a few locals had gathered aboard a spacious tour boat moored in the harbor, chatting and munching pizza, to hear Staley give an update on Barry Arm.
Staley spoke briefly, then invited conversation. As low clouds and darkness settled over the calm harbor, operators asked questions or shared their evolving strategies for safely bringing people to the glaciers that support their businesses. While many operators had quit visiting Barry Arm in 2020, after Higman’s initial warning, some had returned in the years since monitoring equipment was installed. Now, after Tracy Arm and a 2024 landslide and small tsunami at the Surprise Glacier near Barry Arm, some were avoiding the area again. It’s a back-and-forth that’s becoming familiar to people who work near Alaska’s melting glaciers.
Later, Jeff Pedersen, operations manager of Alaska Wildland Adventures in nearby Girdwood, called it an ongoing assessment. In 2024, his company experienced minor damage from a large landslide near the retreating Pedersen Glacier in Kenai Fjords National Park, 60 miles south of Portage.
The slide slammed into a lagoon during a night of heavy rain, causing a tsunami that reached 50 feet high or more on nearby slopes. Farther away, at the company’s backcountry lodge, a 3-foot wave overran boardwalks near a cabin where guests slept. Staff reviewed protocols and now avoid that glacier.
Christine and Jeffrey Smith of the David B responded similarly after the Tracy Arm slide, with sober conversations about safety and where to anchor at night. They’ll still visit glaciers, they said, but their practices will evolve around changing conditions and ideally be informed by more research.
Major cruise lines, including Carnival, Holland America and Royal Caribbean, have also taken notice. In early April, nearly all the large cruise ship companies operating in Alaska announced they would avoid Tracy Arm in 2026. They will reroute over 300 cruise ship visits into neighboring Endicott Arm, a similar fjord where Dawes Glacier is receding from beneath steep mountainsides. None of the large cruise companies contacted for this story responded to interview requests.
“We were taught to not be loud, we were taught to not damage the landscape, to be careful where you go.”
Melenda Lekanof, secretary and treasurer of the Yakutat Tlingit Tribe Council, is also concerned. Yakutat, a village of 600, sits between Southeast Alaska and Prince William Sound, not far from the giant Hubbard Glacier. Lekanof recently coordinated a community workshop attended by Higman and other scientists. It was part of a regional initiative called the Ḵutí Geohazards Project, which combines Indigenous and Western science to better understand landslide risk.
“We have a lot to learn about our changing landscape,” Lekanof told me, including increasing landslides and melting glaciers.
While cruise ships don’t call on Yakutat, they frequently pass the village on their way to the Hubbard Glacier, in Disenchantment Bay. Scientific models indicate a possible landslide and destructive wave along their route. Lekanof worried that a lack of coordinated emergency planning among the cruise ship companies, state and other entities would put the burden of disaster response on the small village. And that a cruise ship accident could cost lives or cause environmental harm where her people fish, hunt and visit sacred lands.
Lekanof would also like to see a monitoring array like Barry Arm’s in Disenchantment Bay. But that would require expanding the kind of funding that the Trump administration has tried to cut.
Traditional teachings about glaciers and tsunamis reflect lessons her ancestors learned over millennia, Lekanof said. “We were taught to not be loud, we were taught to not damage the landscape, to be careful where you go.” But she has heard cruise ships blasting their horns near the glacier. “Spiritually, for the birds and mammals there, I find that to be disruptive and concerning.”
A MONTH BEFORE I VISITED Portage Lake, the captain of the Ptarmigan tour boat snapped photos of fresh debris at the base of the slow-moving landslide, in a zone recently uncovered by the Portage Glacier. She shared them with visitor center staff, who consulted Higman.
Then a multiday storm dumped 8 inches of rain. Weighing the possibilities, the Forest Service evacuated a school group and staff from the visitor center, according to a statement from the agency. And the Ptarmigan canceled tours, citing landslide and tsunami risk. This kind of adaptability might reflect a new way of operating around Alaska’s retreating glaciers.
And while that may seem hopeful, Higman noted that Tracy Arm raises questions about how scientists decide which slopes to monitor. Visible slope movement, like that at Portage and Barry Arm, is important, but so is glacial retreat. Higman and other researchers have gathered preliminary data suggesting that the longer a newly exposed slope had been covered by a glacier, the more likely it may be to slide.
Endicott Arm and LeConte Bay, two busy fjords south of Tracy Arm, deserve investigation, Higman added. After all, the Tracy Arm mountain was a solid-looking mass with none of Portage’s warning signs. And no sensors.
“As scientists, we need to own up to having totally missed the ball on that one,” Higman said.
We welcome reader letters. Email High Country News at editor@hcn.org or submit a letter to the editor. See our letters to the editor policy.
This article appeared in the July 2026 print edition of the magazine with the headline “First the glacier retreated, then came the landslide and tsunami.”
