Are more tornadoes coming to Ontario? Twister-trackers see signs

BARRIE—Adam Hunwicks had just walked downstairs from the bedroom of his Barrie home when his father called: get to the basement, immediately.

Hunwicks, then 15, fled downstairs with his mother and the dogs. Less than a minute later, what sounded like a train passed overhead.

When Hunwicks and his mom emerged onto the street, there was “stuff all over the place.”

“People’s washing machines were in front of our house,” Hunwicks says.

The roof of his home was missing, and so was most of his neighbour’s second floor: pink insulation was floating down from the sky like snow. A cement utility pole had been uprooted and tossed into the street like a pick-up stick.

As Hunwicks stood amid the wreckage, his phone pinged: tornado alert.

In Canada, the “bulls-eye” of tornado activity has historically been centred on the Prairies. But in recent years, a team of tornado-tracking scientists at Western University has observed a dip in activity within this bulls-eye.

Instead, Ontario is picking up hits, along with Quebec.

The team, the Northern Tornadoes Project, is still finalizing its analysis, part of a broader effort to better account for Canadian tornadoes and their destruction. But the preliminary findings mirror an effect that’s been documented by multiple research teams in the United States, which have discovered an eastward shift in America’s infamous Tornado Alley.

Scientists in both countries say they are still trying to understand why this shift may be occurring and what role climate change may play. Regardless of the cause, a shift would have enormous consequences. In Canada and in the U.S., these regions have more people, more built structures and lower awareness of how to protect lives, homes and critical infrastructure when a twister does hit.

Canada has another problem: a substandard alert system. When the scientists at Western assessed the performance of Environment Canada’s warning system, they found warnings were never issued for the majority of tornadoes, and only a small fraction came with at least 10 minutes of lead time — a short but critical window to safety.

In the days after the July 2021 twister in Barrie, 71 houses were ultimately deemed uninhabitable, including Hunwicks’s. The house across the street remains jaggedly topless nearly two years later, a totem of trauma in this otherwise tidy-as-a-pin community.

Just like when extreme heat waves grip cool climates, wildfires rage through regions that rarely see them, and during the unprecedented air pollution crisis that gripped the naive eastern seaboard in early June, the hazards of a natural disaster are multiplied when the public is unprepared. Had Hunwicks’s father, a paramedic, not seen a fierce wind whip a garbage can into a car windshield and sent warning, the teenager might still have been sitting in his top-floor bedroom as a twister ripped the roof off.

Hunwicks says he had heard of twisters in Barrie, but not since before he was born. Until that day, he never thought about them.

“It’s not like we live in Oklahoma.”

A web of scars and a big surprise

America’s Tornado Alley is so well-known it has provided the backdrop for several Hollywood blockbusters. Before the Northern Tornadoes Project was created in 2017, much less was known about Canadian twisters.

“We knew that Canada has way more tornadoes than are reported,” says Greg Kopp, lead researcher for the project and ImpactWX Chair in Severe Storms Engineering. “We’re a big country with low population density.

“So we thought, ‘Well, let’s see if we can find all those missing tornadoes.’ ”

Almost immediately, the team discovered some surprises.

That first winter, the researchers were scanning newly available high-resolution satellite imagery from the previous summer when they stumbled across a web of scarred tracks in the northern Quebec forest. They kept looking, and discovered evidence of a massive tornado outbreak — a flurry of twisters all spawned by the same group of storms. The team eventually identified 16 tornadoes that touched down in one day, a record for the province, and seven the day prior. The very next year, though, another tornado outbreak in Quebec broke the single-day record again.

“In a place where they had told us, ‘Oh, we don’t get many tornadoes here,’ we were very surprised they get all these big outbreaks,” says David Sills, executive director of the Northern Tornadoes Project.

Since then, the team has discovered tornadoes up north that are over two kilometres wide and tens of kilometres long.

“They’re like the big beasts you get in Oklahoma,” Kopp says. “They just happen to only hit trees.”

One in northwestern Ontario left behind a track so massive that Sills — who has spent his entire career studying tornadoes — didn’t initially believe it could from a tornado alone. He says there is no weather-related reason a tornado like this formed over a forest instead of a city, just good luck.

“That’s the kind of event that, if it happened over Winnipeg or Ottawa, it would be a major, major event.”

The U.S., with its warmer climate and longer summers, always outranks us for twisters. But the researchers found that Canada’s tornado activity had been seriously underestimated: before the Northern Tornadoes Project was created, Canada used to identify about 60 tornadoes a year. For the last two years, the team has almost doubled that.

When a tornado falls in the forest

Knowing the true number of tornadoes is important to understand the real risk. But calculating their intensity is equally important. Tornadoes are rated on the “Enhanced Fujita” or EF Scale by reconstructing wind speeds retroactively, based on how much damage occurred: an EF-0 might peel off a couple of shingles, whereas an EF-5 can wipe a family home off its foundation.

Most of the “damage indicators” the scale uses to estimate peak wind speeds are based on structures — houses, schools, utility poles.

“Houses have been very widely studied,” Kopp says. “We have a lot of confidence about the wind speeds for when houses are damaged.”

The Canadian researchers, however, deal with dozens of tornadoes every year that land deep in the forest. Some of the challenges this presents are mundane: until recently, to study tree-fall patterns, a human being used to have to hand-label the direction of thousands of snapped or uprooted trunks. Now they do it with AI in a fraction of the time, thanks to the input of a savvy intern.

Other issues are more consequential. The scale isn’t sophisticated enough to account for how trees in the Boreal forest above the Canadian Shield might be rooted in just a few centimetres of soil, toppling more easily.

Findings from the Northern Tornado Project are being used to help update the EF Scale again, offering a more sophisticated set of indicators for tree damage that will allow scientists everywhere to accurately assess damage caused by storms in the forest.

The project received initial and then expanded funding through private donations from a fund called ImpactWX, which supports organizations seeking better scientific awareness and public understanding of severe weather events. Kopp says the team chose a goal that they believed was important, but that was unlikely to be funded by traditional grants. To gather evidence, they use an arsenal of technology, from drones to satellites to aerial photography taken from airplanes.

“It’s kind of expensive,” Kopp explains.

Without an accurate baseline of how many tornadoes are occurring and how violent they are, it would be impossible to accurately determine if changes are underway — or why.

Where are Canada’s tornado hotspots?

Before he came to the Northern Tornadoes Project, David Sills, then a scientist at Environment Canada, co-authored a 2013 study that statistically modelled the probability of tornado occurrence across Canada. The model accounted for higher tornado reports in more densely populated areas and the predictive relationship between cloud-to-ground lightning flashes and twisters.

The model indicated that Canada’s tornado bulls-eye should be in south-central Saskatchewan and southern Manitoba — an extension, the scientists suggested, of the American tornado alley just to the south.

But when Sills came to the Northern Tornadoes Project in 2019, this bulls-eye did not match up with what he was seeing on the ground.

Ottawa in particular has been getting “pummelled” in recent years, Sills notes. Tornadoes, including big ones, aren’t unheard of in Ontario — a violent storm in 1985 spawned an F4 that killed eight people in Barrie — but the province seemed to be picking up more than its expected quota.

“There’s been quite a concentration of tornado activity in the eastern part of the country. And we’re expecting the big bulls-eye in southern Saskatchewan, we’re not seeing it,” Sills says.

“It does seem like changes are afoot.”

Sills turned to Canada’s national tornado database, now updated with the Northern Tornadoes Project’s more recent, robust data. He and his colleagues compared tornado activity over the last 30 years to an older, overlapping 30-year period.

Their analysis found that the overall number of tornadoes had decreased slightly. But the most striking finding was where they were hitting: in the older era, Saskatchewan had the most tornadoes, just as Sills’s model predicted. But in the more recent period, Ontario had the most.

The researchers presented the findings at an American Meteorological Society conference in 2022. While still preliminary, they mirror published studies from several research groups in the U.S., who have all independently tracked the same concerning trend.

The new ‘Dixie Alley’

As a researcher specializing in extreme weather, Victor Gensini, a professor at Northern Illinois University, spends a lot of time doing field work: “storm chasing, essentially.”

Historically, that meant tracking tornadoes in Great Plains states that are part of Tornado Alley, like Oklahoma and Kansas. But then, “we had four or five, maybe six years in a row where there just wasn’t a lot in the Great Plains,” Gensini says.

The U.S. national tornado database underestimates the true number in that country, too, so Gensini wanted to analyze the question a different way. He looked at the “Significant Tornado Parameter” — an index that weather forecasters rely on to determine whether all the right ingredients for a severe tornado are present — and ground-truthed it against the reports in the database.

The total number of tornadoes across the U.S. has not changed, he found. But Gensini found a “robust” drop in tornado-producing conditions in the central and southern Great Plains, and an upward trend in parts of the Southeast, Midwest and Northeast. The research was published in the journal Climate and Atmospheric Science in 2018. A small handful of studies from other groups have come to similar conclusions using different methods.

The phenomenon is well-documented enough that this new hotspot has its own name: Dixie Alley.

Gensini’s paper warned that the shift could have disastrous consequences. States such as Tennessee and Alabama are more densely populated and have weaker housing stock, including lots of mobile homes. They also have more trees compared to the open plains, making it tougher to see a tornado coming.

Since that paper was published, two of the costliest tornadoes in American history have hit the Southeast U.S., including a massive series of storms in the Southeast and Midwest over two days this spring that killed 27 people.

Gensini emphasizes it’s not that the Great Plains have no tornadoes anymore — the region still gets plenty. It’s just that the uptick in the Southeast is coinciding with a more vulnerable population.

That trend is important to watch in Canada, too, he suggests. Gensini actually extended the tornado environment analysis to Canada, and observed the same shift: a decrease in activity in Alberta and Saskatchewan, and an increase in Manitoba, Ontario and Quebec. But because he didn’t have a good database of Canadian tornado reports to check the environmental findings against — he hadn’t yet met Sills, who he now collaborates with — he clipped off the study at the U.S. border.

“We’ve had reported tornadoes in places like Ontario before,” says Gensini. “But what would it mean if you doubled or tripled or quadrupled the frequency of tornadoes in those areas? You’re talking about a much more significant risk in dealing with a peril that people just haven’t really dealt with before.”

Is it climate change?

Gensini, Sills and other researchers say the next task is to figure out why this eastward shift might be happening, and whether it is tied to climate change or just a natural — and possibly temporary — variation.

The eastward trend tracks with modelling Gensini has worked on that shows climate change is likely to drive an increase in severe, significant storms in the northeastern U.S., Great Lakes and southeast Canada. Tornadoes are tricky to model specifically though, Sills says, because they occur over such small areas.

Climate change may also be shifting the so-called “dry line” in the U.S., the climatic boundary that separates the more arid western U.S. from the more humid east. A 2018 study found that climate change had pushed this barrier, which typically runs north-south across parts of the Great Plains, about 225 kilometres eastward since 1980. This could be pushing parts of this historically tornado-heavy area into a drier climate — one less conducive to tornadoes.

Sills says we will need another 10 years of data to confirm the trend. “But right now, it’s looking like the observations are matching what the climate modelling is saying.”

Whether or not climate change is behind shifts in tornado activity, these storms are hitting places where many people — and the buildings that protect them — are unprepared.

In addition to whether or not the public has an accurate perception of tornado risk and knows what to do in a storm, when the Northern Tornadoes Project assessed the performance of Environment Canada’s tornado warning system from 2019 to 2021, it gave it a failing grade. That score improved last year, but still fell well below the department’s own targets.

When Adam Hunwicks’s phone alerted him to a nearby tornado, he was already standing in the wreckage of his and his neighbours’ homes. The teenager wouldn’t be able to return home for another nine months; for a long time afterward, he felt more comfortable sleeping in the basement, his mother says.

“I’m amazed no one was killed,” Kopp says of the Barrie twister.

“I feel like we get lucky all the time.”

Kate Allen is a Toronto-based reporter covering climate change for the Star. Follow her on Twitter: @katecallen

SHARE: