When a Landslide Unleashes a Flood of Geospatial Data

The slope gave way in the night, like half a million dump trucks unloading all at once. A man camping beside the Chilcotin River heard the rumbling upheaval of boulders, silt, sand, and clay above him and fled his tent just in time.

The material slammed into the river—a tributary of the Fraser River, about 285 kilometers north of Vancouver, British Columbia—and dammed it, causing water to pool upstream in a lake that soon began to fill the valley.

On July 31, the day after the landslide, Brian Menounos, an earth scientist at the University of Northern British Columbia, got a call from the Ministry of Emergency Management and Climate Readiness (EMCR). The agency, which coordinates and supports emergency response across the province, was reaching out to subject matter experts to conduct an initial hazard assessment of the slide.

Menounos, a Hakai Institute affiliate who runs the Airborne Coastal Observatory (ACO) program, is no stranger to hazardous slides. He studied the massive 2020 landslide and tsunami in Bute Inlet, on the BC coast, and helped conduct an inventory of past, present, and future landslides in the Fraser River Canyon, after the Big Bar slide partially blocked the river in 2019.

Hopping into a helicopter, Menounos and Marten Geertsema, who studies landslides and other natural hazards with the province’s Ministry of Forests, flew south from Prince George to check out the scene.

They were impressed by what they saw. Trees blackened in a 2017 wildfire still stood upright on top of the enormous mound of sediment that filled the river. Downstream of the blockage, where Big Creek met the Chilcotin River, water was still flowing, but immediately below the dam, the riverbed was nearly dry.

Menounos and Geertsema recognized the urgency of the situation: they knew the earthen dam would fail. The lake behind the slide was growing steadily. Eventually, the water level would rise enough to spill over the top, eroding the dam and releasing water to surge downstream through the Chilcotin and Fraser Rivers. If the dam disintegrated rapidly—a worst-case scenario—the resulting flood could be catastrophic for communities and infrastructure in its path.

“It was really important to acquire data as quickly as possible,” Menounos says.

On top of the danger, the timing of the blockage was particularly bad for salmon. Sockeye and chinook salmon populations were actively migrating upstream in the Chilcotin River when the slide blocked their progress. Some fish, such as the chinook that spawn in the Upper Chilcotin River, had fortunately reached their spawning grounds before the slope collapsed. But others—including Chilko Lake sockeye, one of the largest salmon populations in the watershed—were still downstream.

The Tŝilhqot'in Nation leadership declared a local state of emergency. The EMCR issued flood warnings along the Fraser River and issued an evacuation order upstream and downstream of the slide.

Menounos and Geertsema, meanwhile, flew home to Prince George to pack field gear and camping equipment; they then drove back to the site. Working alongside government scientists, consultants, and First Nations representatives, their task was to estimate the size of the dam, the potential volume of the lake behind it, and how fast the lake was filling—all to help forecast when the water would begin spilling over the dam and trigger its failure.

The Hakai Institute’s ACO plane flew over the site multiple times with LiDAR, using lasers to create a 3D model of the area surrounding the slide. Drone operators from Spexi Geospatial, a Vancouver-based aerial imagery company, photographed the slide from all angles to map its contours and determine how it might be changing over time.

Back at the Hakai Institute office in Victoria, members of the geospatial team dropped everything to help, processing LiDAR data of the Chilcotin River. This was data that Kîsik Aerial Survey had serendipitously collected just months before, as part of a GeoBC effort to map the entire province; it allowed Hakai Geospatial to create a “before” model of the area and compare it to post-slide LiDAR data to see how the terrain had changed.

Drawing from all of these sources, the researchers determined that the dam was roughly 1,000 meters long, 600 meters wide, and 30 meters tall. Menounos estimated that the landslide had displaced enough soil, sand, and rock—six million cubic meters—to blanket the city of Victoria, British Columbia, up to an adult’s shins.

The lake backing up behind the massive pile of debris stretched 11 kilometers long. On the morning of August 5, six days after the slide, Menounos, Geertsema, and their on-site collaborators were watching the lake, now level with the height of the dam, waiting for the moment it would overflow.

“The water fills lower points in the dam first,” Menounos says. “We were guessing where we thought the dam was going to breach.”

As the water began trickling over the edge of the slide in miniature waterfalls, it cut into the sediment, carving channels in the top and front of the dam. These grew bigger and bigger, as more muddy water poured through, cascading into the river channel below. Chunks of the slide debris fell away, earth and rocks disappearing into the coursing water downstream.

But the dam didn’t collapse right away. The reason, Menounos says, is geology: some of the boulders, silt, sand, and clay that made up the slide had been heavily cemented together. This type of sediment, which geologists call a diamicton, almost behaves like rock; dense and compact, it is difficult for water to cut through.

The diamictons peppered throughout the dam slowed down its failure. As a result, the lake took roughly 20 hours to drain, and the downstream surge was smaller—and less destructive—than the worst-case scenario that everyone had feared.

“It would have been a much more rapid collapse if not for those deposits,” Menounos says. “If the dam had incised twice as fast, then the magnitude of the flood would have been much, much larger.”

Fortunately, no one was injured in the flood, though the high water damaged important Indigenous cultural sites along the Chilcotin and Fraser Rivers—including historical Secwépemc house pits and burial sites—and other properties downstream. The Tŝilhqot'in National Government closed a bridge over the Chilcotin River below the slide, worried it could be unstable.

Hakai Geospatial performed an additional ACO survey after the dam was breached, and again a few weeks later, looking for more landslides. Riverbanks that had been saturated with water along the edge of the lake will become less stable, Menounos explains, meaning those slopes can fail as the water recedes. The floodwaters churning downstream also eroded and undercut slopes, pulling more sediment into the river.

In mid-September, six weeks after Menounos and Geertsema returned home, two separate, smaller slides did occur downstream of the original site. Both also barred the upstream migration of adult salmon, though the river cut through each blockage within 24 hours.

By October, Fisheries and Oceans Canada had counted nearly 6,500 chinook and roughly 65,000 sockeye migrating past the landslide area—though the numbers were lower than expected for 2024, the slide’s impact had been less severe than the Big Bar emergency in 2019. Hakai Geospatial adapted a mapping tool originally developed for the Fraser River Canyon landslide inventory to depict which salmon populations were affected by the Chilcotin River slide, using data from Fisheries and Oceans Canada, the Pacific Salmon Foundation, and Simon Fraser University.

This region of the Fraser River watershed has seen events like these for millennia, Menounos explains. Several thousand years ago, a large landslide also ran right across the Fraser River downstream of Lillooet. And the landslide on the Chilcotin River occurred just upstream of Farwell Canyon, where a similar landslide also dammed the river in August of 1964. The Tŝilhqot'in name for the canyon is Nagwentled, which roughly translates to “landslides across the river.”

But never before have geologists had the luxury of having so much information from both before and after a landslide. “This was probably one of the best-documented damming and then subsequent breaching events along a river that we’ve ever had—not just in BC, but in probably many places in the world,” says Menounos.

“We’re kind of in an era of unprecedented availability of digital data and geospatial data.”