An above-average number of tornadoes may be ripping through — and ripping up — Canada this summer, scientists say.

Just this past Sunday, a tornado near Guelph, Ont., smacked down trees and powerlines and caused "significant damage" to two houses. Last week, a 'monster' tornado whipped through southwestern Manitoba, uprooting trees and smashing fences and buildings for nearly three hours. A week earlier, a tornado touched down southwest of Calgary amid a thunderstorm, funnel clouds and golf ball-sized hail.

In fact, a string of summer twisters have been reported in Canada since the beginning of June.

Canada may see more than the usual number of tornadoes this year because it's an El Nino year, and that may give tornado activity in Canada a boost, says John Allen, postdoctoral research scientist at the International Research Institute for Climate and Society at Columbia University in Palisades, N.Y.

"El Nino seems to produce the conditions that we would expect to foster more tornadoes in Canada, similar to what is seen in Florida in El Nino years," he told CBC News in a phone interview.

El Nino is a warm spot that appears in some years in the Pacific Ocean that can influence weather patterns around the globe.

Canada is one of the world's tornado hotspots, second only to the U.S, with about 60 reported each year, typically during a season that lasts from June through August, according to Environment Canada. Of those, 43 would be expected in the Prairies and 17 in Ontario and Quebec.

Tornadoes in Canada are usually too small to do much damage, but can sometimes cause massive destruction — one in southern Ontario in 2005 caused $500 million damage. Another in Edmonton in 1987 caused $181 million in damage, killed 27 people and injured 600 others.

While there are seasonal predictions for other destructive storms, such as hurricanes, scientists know surprisingly little about tornado patterns and how to predict tornadoes more than a few days ahead of time.

1st seasonal forecast

Allen is among scientists now working on that problem. He was the lead author of a study published earlier this year that included the first ever seasonal tornado forecast for part of the U.S.

That prediction, that the central southern U.S. would have fewer tornadoes than usual between March and May because of El Nino, turned out to be correct, said Michael Tippett, who also co-authored the study, published in Nature Geoscience in March.

EDMONTON TORNADO

Tornado victim Clint Keates stands beside what was his trailer in the Evergreen Mobile Home Park in Edmonton's northeast end three days after the July 1987 tornado. It caused $181 million in damage, killed 27 people and injured 600 others. (Ray Giguere/Canadian Press)

But more research needs to be done to figure out what the effect of El Nino would be in other parts of North America.

Allen cautions that while El Nino is expected to boost tornadoes in Canada this year, that isn't yet backed up by a rigorous, scientific data analysis.

El Nino is predicted to gain strength from fall into spring, but Allen isn't sure what that will do to next year's Canadian tornado season.

What makes tornadoes so hard to track and predict is that they generally are only a few metres wide — tiny compared to thunderstorms, which are about eight to 40 kilometres wide, or hurricanes, which are on average 160 kilometres wide.

"You can't use satellites to see them," says Tippett. "And they don't last a long time." Typically they touch down for just a few minutes to half an hour.

Because of that, they're tracked using eyewitness reports from people who see them, meaning their numbers are probably underestimated, especially in sparsely populated areas.

150 tornadoes a year?

Vincent Cheng, a postdoctoral researcher at the University of Toronto Scarborough's Climate Lab, estimates there are actually 150 tornadoes in Canada per year – more than double the 60 reported on average.

Tornadoes aren't just hard to count, they're also hard to measure, he says: "You can't put an instrument in front of it."

To estimate the wind speed of a tornado, scientists extrapolate after it has passed based on the damage it caused.

Cheng estimated the number of unreported tornadoes in Canada using a mathematical model he's been developing with colleagues at Environment Canada. The goal is to be able to generate monthly and seasonal tornado forecasts for different parts of Canada. 

The model predicts how different conditions in the atmosphere during a thunderstorm will affect the risk of a tornado. For example, the risk is higher when conditions make it easy for air to rise quickly and there is a big difference in the speed and direction of winds at different heights above the ground. The team published their study in Nature Communications earlier this year.

Forecasts could be used in disaster planning or to figure out how tornado-resistant buildings such as hospitals need to be in different regions.

Vincent Cheng

Vincent Cheng, a postdoctoral researcher at the University of Toronto Scarborough's Climate Lab, estimates there are actually 150 tornadoes in Canada per year – more than double the 60 reported on average. (Ken Jones/University of Toronto Scarborough)

Cheng and the Columbia University researchers hope to eventually predict how tornadoes will be affected by climate change.

But, Cheng says, "We have to be able to predict the present first."

Allen says even when scientists figure out the role of different factors such as El Nino in increasing or decreasing the number of tornadoes in a given year, that says nothing about any single twister.

"An individual event cannot really be attributed to this climate source alone," he said. "Even quiet years can produce deadly tornadoes."