Lessons of a landslide detective

From atop a narrow ledge some 2,000 feet high, the tiny white speck of a 144-passenger tour boat stands out against the jade surface of Alaska’s Portage Lake as it putters up to the face of a steep glacier that towers over the shoreline. On the far end of the lake, just out of view, is the visitor center, a 1980s brutalist mass of glass and concrete hanging out over the water. Below this ledge is a slow-moving landslide. It’s not a landslide as you might picture one—a rapid flow of dirt and debris rushing downhill after heavy rain. Instead, it’s a mass of bedrock moving around six feet per year, which could accelerate to a sudden collapse. If that happens, a resulting tsunami in the water below could capsize the tour boat, wipe out the visitor center with a wave several hundred feet high, flood the valley, and maybe spill over Portage Pass, a narrow gap between the Chugach and Kenai Mountains, inundating an airstrip and a cruise ship terminal four miles beyond us.

Geologist Bretwood Higman—Hig to everyone else, including his wife—puts the chances of collapse in any given year at one in 30. He’s taken Robes Parrish, the Chugach National Forest’s new geologist, and me along for fieldwork on the landslide, having brought us to this ledge only after we scrambled over the hundreds of vertical feet of loose, slowly slipping bedrock. It’s research Hig is eager to get done as quickly as possible, for obvious reasons.

“Here’s the part that’s tearing itself apart,” Hig says. The tundra ahead splits wide open. A patch of green mountain heather is interrupted by a gaping chasm. We watch as fresh snowmelt streams through large shards of slate, disappearing into a series of cracks. Hig believes that far below our feet some 40 million cubic yards of bedrock are creeping toward the glacier below. That means the base of the mountain is shifting faster than where we are at the top—leaving the part we’re standing on precariously supported in midair. “And that scares the crap out of me,” he says.

The mountain could continue to slide slowly for decades or, with little notice, suddenly accelerate and then collapse catastrophically. So Hig unzips his hiking backpack and pulls out a custom-made sensor, contained in a simple screw-cap mason jar. Government geologists like Parrish often have to pick their battles, working with both state and federal officials to first identify potentially dangerous weak spots along mountains like these and then commission formal studies to determine if the potential risk warrants further tracking. As an independent geologist who specializes in diagnosing deep-seated landslides and the tsunamis they can generate, Hig has developed his own tools and techniques to challenge the bureaucracy that can slow down government scientists, like those at the U.S. Geological Survey. “[He] provides a service that, frankly, the USGS can’t,” Parrish says.

Over the past decade, Hig has become known as something of a scientific vigilante, calling public attention to landslides whenever governments seem slow to respond while simultaneously working to determine how scientists can get better at knowing when and where the next big one will hit. The race to prevent such catastrophes has never been more urgent: As Alaska warms several times faster than much of the world, the Portage Glacier and many others are retreating at an unprecedented rate. They leave valley walls unsupported, a phenomenon called de-buttressing. At the same time, stabilizing permafrost is thawing and rain events are trending heavier and more frequent. The rock in many locations is already weak—Hig calls it “Chugach Crud.” Alaska’s mountains are literally falling apart.

Around the world, more places with glaciers, including Norway, New Zealand, Switzerland, Iceland, Greenland, the Himalaya, and Chile, are seeing deep-seated landslides as a growing threat. In Alaska, however, these processes are happening at a scale that makes the landscape an invaluable laboratory. Tsunamigenic landslides in the state have increased roughly tenfold in just the past 10 years, according to Hig’s research, which includes landslides that have been dated back to the 1850s.

(This scientist is racing to predict deadly volcanic mudslides.)

Today the planet’s top landslide scientists flock to the region and often embed with Hig, whose no-frills approach to fieldwork extends to his steady diet of trail mix and his trail-running shoes that are held together with fishing line. Alaska has so many landslides, international scientists can pick and choose the best one to answer their specific research question. On the coast, they can study how glacial retreat leaves fjords unstable or, further inland, track how permafrost thaw “unglues” landmasses held together by ice. Together with Hig, they’re hoping to learn more about why slopes fail and how to stay ahead of the next collapse.

“The most likely scenario is something really, really bad is going to happen before we are on top of any of it,” Hig says. “And I want to be able to say I did something that was really helpful. And the odds are not really with me on that.”

In late 2015, roughly 180 million tons of rock crashed into Taan Fiord, a tributary of Icy Bay along the southern coast of Alaska. The impact generated a wave that ascended more than 600 feet up a slope and stripped forest from eight square miles of Wrangell-St. Elias National Park. The fjord, Hig says, was unoccupied at the time, but it empties into Icy Bay, which is well trafficked by cruise ships. As an international group of scientists scrambled to investigate, Hig volunteered to plan logistics. The researcher already knew Alaska as few other geologists do.

Hig had grown up in Seldovia, an isolated village of stilt homes along Alaska’s Kenai Peninsula that is accessible only by boat or bush plane. As a graduate student at the University of Washington, he had been one of the first tsunami scientists on the scene in Sri Lanka after the 2004 Indian Ocean tsunami killed more than 200,000 people. He remembers entering village after village that had 99 percent fatality rates. Decades of tsunami research hadn’t saved any lives. Hig thought of his own hometown, itself in a tsunami inundation zone. “It was profoundly influential to be in this place that just had been so devastated,” he says. “You’re in this world that’s just been torn apart.”

After receiving his doctorate in 2007, Hig set out with his wife, Erin McKittrick, on a year-long, 4,000-mile trek by foot, ski, and pack raft from Seattle to the Aleutian Islands on the far southwestern tip of the state. Now, both realms of expertise were colliding. “He does these very dangerous things that have a very thin margin of error, but he’s very scientific about all of them,” says Noah Finnegan, a geomorphologist who studies landslides in California, and a college friend of Hig’s.

The Taan Fiord investigation concluded that the landslide was due in part to glacial retreat, a factor researchers were just beginning to consider at the time. But after academic papers were published and the scientists went home, Hig was left with a nagging feeling. Alaska has a lot of retreating glaciers, and the landslide-generated tsunami at Taan Fiord wasn’t even the most dramatic on record. In 1958, an estimated 40 million cubic yards of rock slid into Lituya Bay, an inlet in what’s now Glacier Bay National Park. The impact generated the highest wave crash ever recorded on Earth, reaching more than 1,720 feet above the shoreline and killing five people.

Hig wondered: Could other mountains spontaneously collapse? He began obsessively studying publicly available satellite imagery, laser-imaging datasets, and historical aerial photos for signs of movement. He chartered helicopter and fixed-wing plane flights to conduct his own aerial surveys. The research didn’t come easily; Hig is colorblind and dyslexic. But it didn’t take him long to find something. In June 2019, his sister Valisa, then an artist in residence with the Chugach National Forest, was boating when she noticed a mile-long section of slope near the retreating Barry Glacier hanging precariously above the water.

Hig pulled up satellite images and zoomed in on the Barry Arm fjord, comparing the current view with the historical record. Then he panicked. The entire slope appeared to be sliding toward the water at about 85 feet per year. By his estimate, the landslide—a 650 million-cubic-yard chunk of sliding bedrock—could completely dislodge within five years, generating an initial 500-foot wave. The surge could quickly cover the 40-mile distance southwest to Whittier, sending a 30-foot swell to swamp the town. All told, Whittier’s 275 residents would have no more than 20 minutes to evacuate uphill around a freight rail yard that separates the town from the shoreline. If cruise ships were docked at the town’s terminals, though, several thousand people would be at risk.

Hig wrote to a local Forest Service manager and suggested closing the area to recreational traffic. “Wow, the images are striking,” the manager wrote back. But, he informed Hig, the Forest Service would neither close the area nor issue a warning.

“I’m kind of naturally inclined to be contrary to authority,” Hig says. So he enlisted the help of Anna Liljedahl, a Swedish permafrost hydrologist for the Woodwell Climate Research Center based in Alaska, and drafted an open letter to state officials signed by a dozen other researchers. Weeks went by. “They were not releasing the information,” Liljedahl tells me. Worrying a collapse could be imminent, the researchers decided to share their letter with the press. And on May 15, 2020, the entire state of Alaska, including those in the inundation zone, learned about the Barry Arm crisis from a New York Times article with the headline “It Could Happen Anytime.”

The impact on the local tourism industry was immediate: There was a sharp decline in boat traffic near Barry Arm. Kelly Bender, former co-owner of Lazy Otter Charters, a water taxi and sightseeing company, who is now the city’s chamber of commerce president, remembers feeling blindsided. “I tried to be gentle,” she says about her later discussions with Hig. “But this just put us into a tizzy.”

The alarm bell, by some accounts, galvanized Congress to pass the Landslide Preparedness Act, which authorized the U.S. Geological Survey to study and implement new tracking measures at Barry Arm. Soon federal and state officials were working together to triage the situation. Initially, Hig and Liljedahl were included in the working group’s meetings. But Hig says that disagreements over bureaucracy and how to communicate with the public led to a falling-out.

Still, he got results. The working group eventually installed a suite of high-tech instruments that now act as an early-warning system. It includes a synthetic aperture radar array to map landscape changes, infrasound sensors to listen for rockfall, seismometers to monitor ground movement, high-resolution cameras, and tidal gauges to register water surges—all powered by solar panels and large batteries that can keep everything running through the snowy, gray winter. In all, the federal government spends more than three million dollars annually monitoring the landslide.

For any independent scientist, that federal action would be a huge victory. But Hig doesn’t think it’s enough. In 2024, a series of five landslides collapsed into Surprise Inlet in Prince William Sound, a short distance from Barry Arm; the site had not been flagged as an imminent hazard. The same year, in Kenai Fjords National Park, another undiscovered slide triggered a tsunami that swept under the occupied guest cabins of a nearby lodge. A year later, in 2025, a slide in Tracy Arm fjord generated a wave that ran up nearly 1,600 vertical feet of terrain, carried away the gear of kayakers camping on a nearby island, and threatened a cruise ship. It also was not on any state or federal watch list.

Barry Arm is far from the only threat, but it’s the only major landslide risk area in Alaska that’s getting such focused scientific attention. Hig thinks the next horrible disaster will “most likely” be somewhere else.

Perched on the narrow ledge above Portage Glacier, buffeted by wind, Hig pulls a rock drill from his battered backpack, drills into the boulder, slides a steel anchor into the hole, and bolts the mason jar full of radar electronics to it. Close by he has anchored a metal reflector about the size of his palm.

The contraption essentially bounces radar off the reflective surface and transmits the distance reading over a local radio system to a computer. For Hig, the technological solution arrived after he watched a childhood friend in Seldovia tinker with a cheap radar sensor to monitor the oil level in his household fuel tank. The same sensor, he realized, could be mounted next to a landslide and aimed at the moving portion to track its movement in real time.

Having forgotten a sledgehammer, Hig whacks the whole assembly with a rock, driving the anchor into the boulder. The piece of slate breaks after three hits, an irony that isn’t lost on the geologists. Parrish wields a caulk gun, attempting to waterproof the sensor against the deep snowfall that will drift over this spot in the winter.

In all, each mason jar instrument costs about $300. Hig hopes the basic construction will fare well against goats and the jaws of curious bears, which have demolished enough scientific instruments in the area that there are now academic papers tracking the phenomenon. He also hopes that the price might encourage more at-risk communities to invest in the devices. It’s an idea that’s already being implemented in the Philippines, where a researcher named Roy Kaimo has worked with some locals to install relatively cheap tilt sensors in the ground to measure changes in incline on more than 50 landslides. Several villages now monitor and service their own early-alert systems, deciding if and when they would evacuate their homes should heavy rain or an earthquake accelerate unstable rocky slopes.

Parrish certainly sees the appeal. “We don’t need to spend however many millions they’re spending at Barry Arm on each and every possible landslide,” he says. “That’s why I like Hig’s approach, where he’s really trying to scale this in a very modest way.”

At the Portage Glacier site, the Forest Service plans to use his data to augment an existing system staff use to monitor risks that takes into account weather and rockfall sightings. But Hig has also spent several years working in Glacier View, a community of 375 residents about two hours outside Anchorage. There, he’s learned that just identifying slow-moving landslides isn’t enough to save communities. You also have to convince the people who live there that they’re at risk. In some ways, that’s proved even trickier than the cutting-edge science.

In 2023, Hig emailed Bill Billmeier and Betsy Young, a pair of former backcountry guides who run an aerial imaging and geospatial data company out of Glacier View. He’d noticed signs of deformation in the valley’s mountains and wanted to compare notes on an area eventually known as the Matanuska Narrows Instability that was situated on a slumping mountain above the town. As with Barry Arm, Hig grew apprehensive. “This one has just bumped way up my ‘terrifying’ list,” he wrote to them in an email. Hig calculated roughly 130 million cubic yards of weak rock were held together largely by the strength of permafrost. In the worst-case scenario, the landslide would take out the only highway and Glacier View’s internet and power lines—and dam the Matanuska River far below, causing flooding in the valley and, once the dam broke, a surge of water toward communities downstream.

At first, Hig thought this might be the perfect test case for his new sensors and a chance to right his mistakes at Whittier. There would be no press releases or letters to officials. No blindsiding residents. This time, he’d go to the community first.

In mid-2023, Hig attended a community council meeting with a poster and a slide presentation, hoping to charm everyone with his transparency. He likes to say that he’s an open book, and so are his instruments, in their clear glass jars, which appeal to an Alaskan culture that values canning and a good-quality weld.

It went well enough at first. The council voted to send a letter to borough and state officials asking for assistance to fund studies. They stockpiled 30 days of food supplies for tourists who might end up trapped in Glacier View. They formed a geology committee and explored the ways that a rural community might secure more federal funding for disaster preparedness.

Then Hig recruited Susan Conway and Costanza Morino, two European geomorphologists interested in permafrost landslides, to study Glacier View more deeply. They covered slopes with GPS sensors and flew drone surveys. Their research helped Hig identify areas that could benefit from his real-time monitoring system and drove home the urgency of this kind of work. “The quantity of landslides in Alaska is quite impressive. The mountains seem to be falling apart,” Morino says. It’s puzzling, she notes, given how many permafrost degradation landslides there are in Alaska, that they haven’t received more scientific attention. “There should be more people working on this.”

On the coast, the landslides are more clearly linked to de-buttressing glaciers. Inland, in places like Glacier View, there are other mysteries. Although these mountains appear to be held in place by permafrost, clearly some fail before the permafrost has fully thawed. It’s possible, Hig says, that the more a slope moves, the more thaw accelerates due to friction, triggering a feedback loop. In Denali National Park, a slide of this type destroyed the only road through the park at its halfway point. By studying Alaska, Morino and Conway hope to inform proactive management in places like the northern Arctic and the Himalaya where permafrost hasn’t degraded to the same degree yet.

After their fieldwork, Morino and Conway organized a workshop that brought together state scientists, borough employees, and locals. “I had a bit of a goose bump,” Morino says. “When you talk with people face-to-face about the fact that something might happen, you feel a little bit of responsibility.”

That’s when everything fell apart. The scientists presented a map of Alaska with a cluster of red dots over Glacier View. Each highlighted a permafrost landslide with some degree of movement. The borough’s emergency manager warned residents they could be stranded for roughly 90 days without help if a landslide destroyed the highway. It all was suddenly too real.

“The tone of it changed,” says Sarah Barton, a Glacier View resident. Instead of welcoming the information, some locals began voicing concerns about the research’s impact on property values and insurance coverage. “It was an awkward time,” Barton says. “And I think it still is in some ways.”

In late 2024, the council considered an anonymously submitted resolution that some interpreted as a ban on public discussion of landslide hazards in the community—with an apparent threat to sue if any ongoing research negatively affected the community. The council members rejected the resolution but voted to disband the geology committee and hired a lawyer to protect themselves from liability. “It was really upsetting,” says Young, who was on the geology committee. “It’s hard not to take that as a threat: ‘If you talk about this, we might sue you.’ ”

For his part, Hig defends emphasizing worst-case scenarios. While uncertain, they highlight the stakes for gathering more data to rule them out. And some in Glacier View still share that opinion, including Mike Dreiske, the director of nearby Victory Bible Camp, who has always been vocally in favor of Hig’s research. Dreiske encouraged Hig to install five mason jar sensors on a landslide and allowed him to place antennas atop the Miracle Lodge, one of the camp’s main buildings, to assist in data collection.

On a warm morning in July 2025, Hig meets Dreiske inside Miracle Lodge’s bustling dining hall. They’re surrounded by dozens of kids of all ages who have shown up for breakfast. Some of the older campers are wearing ranch-themed flannels and boots, while some younger ones are still dressed in animal-print onesies. The camp, which occupies an isthmus set between two lakes and has 80 buildings, draws 160 campers a week when at capacity. Meanwhile, the steep slope of the mountain far above us has become a moonscape littered with the disintegrated remnants of once frozen boulders. The peak is slowly splitting as it thaws, sloughing down toward the valley.

Over coffee, Hig mentions how quickly the landslide above us is moving, and says one instrument shows it “ripping along” at half an inch per day.

Dreiske looks startled. “A day?” he says loudly. Despite his word choice, Hig assures him that this qualifies as “slow-moving”—it’s the sort of shift anyone would easily outrun. Still, some of the rock is moving at a rate of 30 feet per year, leaving the mountain peak increasingly unsupported. If it maintains that speed, it would take the landslide about a century to reach the camp. Unless it collapses catastrophically; then it could arrive in minutes.

Hig admits he doesn’t know which will happen. All he and Dreiske can do is watch and monitor for anything alarming. “If it means that it increases insurance or we can’t have insurance, I can’t control that,” Dreiske says at one point. The recent flash flood that occurred at a Christian camp in Texas, killing 27, was still fresh in his mind. “If it means something catastrophic happens and it wipes this place off, I can’t control that. But if I can make sure there’s nobody here, that I can control.”

With enough sensors collecting real-time data, Hig hopes that scientists will start to make progress on the biggest question in landslide science: What causes a slope to suddenly collapse?

For now, though, triggering is “a total black box,” says Finnegan, Hig’s longtime friend who studies slow-moving landslides, including one in California’s Diablo Range, outside San Jose. Engineers typically interpret landslide creep as a “transient evolution towards catastrophic failure,” Finnegan says. Or, in simpler terms: More movement is scarier. This idea underlies some scientists’ approach to landslides, but researchers are learning that it might not be quite right. Only about a quarter of the slides Hig has studied exhibited precursory signs before collapse. Sometimes a slope collapses right after it stops moving. Or a stationary section collapses after the movement nearby slows down.

Based in part on Hig’s research, Finnegan is developing a different model for predicting collapse. It draws on seismology principles often noted by researchers studying earthquakes. As faults shift, some rock types appear to be weakened by sliding movement, but others are strengthened. So a constant slow creep at fault lines sometimes releases stress that would otherwise cause an earthquake, stabilizing the system.

By the same token, it’s possible some giant slides might become more stable as they move. To learn more, scientists are conducting laboratory tests on the rocks that make up different slides and trying to observe more slopes in real time. This is where Hig’s real-time monitoring comes in. “As we document more and more of these cases, I think signals will start to emerge from that chaos,” says Finnegan, who has deployed similar instruments at his own test sites in California.

In Alaska, officials are aware of the increasing hazards and are working closely with federal authorities to prioritize the most concerning landslides. Dennis Staley, a USGS geomorphologist, says the agency hopes to build on its monitoring efforts at Barry Arm and deploy a more economical suite of sensors on slides across the state. But some experts admit that they are often trailing behind Hig, envious of his speed and lack of hierarchy, and frustrated when he unexpectedly draws attention to a new hazard.

“I don’t think I could work with a lot of Higs,” says Michael West, who runs Alaska Earthquake Center and the statewide seismic monitoring system for landslides. “But the world needs people like that. The world needs instigators.”

In that way, getting through to just one person like Dreiske at Victory Bible Camp holds a larger accomplishment for Hig. It starts a chain of potentially lifesaving action that grows in momentum. The borough’s emergency services department is collaborating with Hig and the camp on monitoring and emergency planning, and the camp now covers some of the costs associated with Hig’s work. Another recent victory: Hig has secured $25,000 in funding and a five-year commitment from Parrish and the Forest Service at Portage Lake to continue work on a risk-assessment dashboard officials can use in deciding whether to evacuate the visitor center. Real-time data from the sensors that he recently installed will soon appear in the system, and officials there have begun developing evacuation protocols and plan to soon hold preparedness drills for staff.

In September, the Forest Service put those plans to the test when, after a stretch of unseasonably warm days, over a foot of rain fell on the landslide. A tour boat captain observed increased rockfalls in the area, so the Forest Service decided to evacuate and closed the area early for the season. Meanwhile, Hig’s newly installed instruments showed that the terrain temporarily sped up tenfold before slowing back down. This is the kind of real-time data that will arrive automatically to guide decisions. After a year of federal government cuts and turnover, it’s another essential win for the intrepid geologist. One place in Alaska—a popular tourist destination—is a little bit safer.

If Hig celebrates these wins, he doesn’t show it. He knows that, like a landslide, it can all collapse in an instant. “I live in unending fear that I’m going to just drive home some message really successfully and it’s going to turn out to be counterproductive,” Hig says. “Nonetheless, I still think I have a better chance of having a positive impact if I do try, rather than abandoning the effort.”