It was 10:30am on a Wednesday, and Tery Kozma was at the Bremner Trio Hydro camp, halfway up the west side of Harrison Lake. She and a group of others were enjoying the January day on a deck overlooking the water when something caught her eye. Across the water, Mount Breakenridge was falling.

Or rather, pieces of Mount Breakenridge were falling. Dust billowed around the base of the mountain as rocks and dirt tumbled along an ancient fault line in the mountain. Some of that debris landed in the water, joining a cone of debris that had piled up around the base of the mountain for decades. Other rocks came to a stop among the trees on the slope.

The slide earlier this year injured no one, destroyed no property, and left the forest service road on the east side undamaged. But Marten Geertsema, a provincial researcher studying Mount Breakenridge, kept his eye on the mountain.

“It worried me initially, because sometimes that’s a precursor to a big event happening,” Geertsema said. “Often large landslides are preceded by lots of small ones.”

Geertsema would know. He is a research geomorphologist for the province and a professor at UNBC and UVIC studying natural hazards and terrain analysis. Specifically, he has been looking at landslides on the mountains around Harrison Lake, and the possibility that one of those could trigger a tsunami that could destroy communities along the lake.

Landslide-caused tsunamis have killed before

The concept isn’t as fictional as it seems. Large volumes of rock falling into lakes or fjords have caused devastating tsunamis across the west coast before. Oral historians recall a landslide-caused tsunami in Knight Inlet that completely destroyed an Indigenous community about 300 years ago. In 1958, a landslide in Alaska caused a tsunami that killed five people in the bay and tossed a fishing boat out into the open ocean. (Everyone on board that boat survived.) And just 14 years ago, in 2007, a rockslide at Chehalis Lake only 10km west of Harrison Lake caused a several-metre-high wave that damaged three recreation sites and left a section of shoreline as bare rock. (There were no injuries, as no one was at the campsites.) Climate change has made these kinds of events more likely, as thawing permafrost and receding glacier ice destabilizes mountain slopes.

In the 1990s, researcher Steve Evans found there was a potential for a large landslide coming off the mountains around Harrison Lake, including Mount Douglas and Mount Breakenridge. Subsequent studies found it was possible that 200 million cubic metres of rock (about 80 Egyptian pyramids) could collapse into the lake, sending a wall of water rushing south towards Harrison Hot Springs. He estimated the waves could be seven storeys tall at Echo Island, and as high as five metres at Harrison Hot Springs. A wave like that would overtop the lagoon, pushing water through the village at five metres a second. The wave could last a minute and a half, which would flood the area at least as far as the municipal hall, if not further.

The hazard was real, but no one knew how likely it was, or if it would be as bad as feared. Real-time monitoring of the Mount Breakenridge slope was recommended, but never implemented. Then, more than two decades later, Geertsema was hired by the province to study the slopes and find out what, exactly, could happen on the lake.

Beneath Harrison Lake’s waves

The first place he looked was underwater.

Along with student Katie Hughes, who published a paper on their discoveries last August, Geertsema found three large deposits of rock in the lake beneath Mount Breakenridge, Mount Douglas, and the more southern Silver Mountain. Those deposits, based on their size and position, likely caused previous tsunamis on the lake—and were the first confirmed evidence of past tsunamis on Harrison Lake.

Next the team looked on land, in nearby wetlands. They were hoping to find a layer of sediment that may have come from the lake during a tsunami event. If they found that, they could determine how large the wave was, and how long ago it happened. They didn’t find anything. No evidence of a tsunami that would have submerged Harrison Hot Springs. No evidence of anything at all.

“We were thinking that we might find something that’s a few thousand years old,” Geertsema said. “We didn’t find that—even though we took core samples and excavated looking for [evidence]. So that was a blind alley.”

But that is also a part of science. Sometimes your hypothesis doesn’t give you the evidence you expected—and when it comes to landslide-caused tsunamis, no evidence can be a good thing. If there had been a destructive tsunami that washed over Harrison Hot Springs thousands of years ago, it would have left organic matter in the wetland. Geertsema would have found it. But he didn’t. And that means it’s possible a tsunami that destructive didn’t happen—and it’s back to the drawing board for Geertsema’s research.

“It’s a type of hazard that we know exists, and we’re learning more about. But they’re tricky things,” Geertsema said about the landslide-tsunamis. How many rocks fall, how fast they fall, and how far up the mountain the slide starts all impact how destructive a tsunami might be. “Those are all pieces of the puzzle that we’re trying to bring together.”

No warning system in place

Two days after Kozma saw rocks tumbling down the side of Mount Breakenridge and shared her video on social media, staff and government officials from communities around Harrison Lake met with the province to discuss the impact of the slide.

Geoscientists from the province took a helicopter ride around the mountain to assess the damage and tsunami potential. There was good news for the short-term: slides were occurring in the top level of soil on the lower third of the slope, and there was no increased risk for a tsunami-causing slide.

What was less comforting was the continued lack of early warning systems in the area. Although it was recommended in the 1990s, the province does not consistently monitor slope stability on Mount Breakenridge, nor does it have an early warning system for large slides. (Emergency Management BC noted that it works with Public Safety Canada to provide some satellite monitoring for the area.) Municipal and regional governments are required to prepare emergency plans and maintain an emergency response team, but no one has been willing to pay for the equipment that could detect landslides in rural areas like Mount Breakenridge.

According to Harrison staff, the “conversation is ongoing regarding the potential for a Harrison Lake tsunami” between the village, the Fraser Valley Regional District, and the province. There is currently little in the area’s emergency response plan about landslide-tsunamis, although more could be added if the risk becomes greater. But right now, it doesn’t look like the risk is as high as people feared.

“I don’t want to say too much about hazards and risk, because we’re still trying to figure things out,” Geertsema, who is now trying to map the deposits in the lake back onto the mountainside using advanced imaging, said. “But I think we’re moving away from a worst-case scenario.”

Want to read every story as soon as it’s published? Sign up for our newsletter.

Correction: An earlier version of this story incorrectly stated the 1990s studies anticipated the volume of a landslide would be 200 cubic metres (26 dump trucks). It is actually 200 million cubic metres (80 pyramids). The article has been fixed and we regret the error.