Man standing in tornado-stricken site

A twister destroyed this structure in Fairdale, Illinois, last year.

Matt Marton/AP Images for The Weather Channel

Extreme tornado outbreaks are getting worse, but why?

The number of tornadoes pounding the United States during the most extreme outbreaks has roughly doubled over the past 50 years, a new analysis shows. But the study also yields a big surprise: The increased severity of such tornado outbreaks, at least at first glance, doesn’t seem to be related to climate change.

“Either the recent increases are not due to a warming climate, or a warming climate has implications for tornado activity that we don’t understand,” says study author Michael Tippett, a climate scientist at Columbia University.

When Tippett and his colleagues pored through tornado statistics from 1965 through 2015, they identified 435 “extreme outbreaks”—clusters of a dozen or more twisters rated strong enough to have caused at least moderate damage to structures. (Because tornado numbers vary so much from year to year, the team grouped their data into 5-year intervals.) Across that half-century, the number of such outbreaks didn’t change very much, Tippett says. However, the total number of tornadoes in those outbreaks jumped dramatically: In 1965, the worst outbreak included about 40 tornadoes, but in 2015 that number had statistically grown to include nearly 80 ground-scouring twisters, the researchers report online today in Science.

The increase doesn’t seem to be related to better reporting in recent years, says Tippett, so the trends seem to be genuine. The team then looked to weather data for a scientific explanation.

One parameter meteorologists use to help identify regions at increased risk of severe weather is called convective available potential energy (CAPE). It’s a measure of the tendency for warm air at Earth’s surface to rise, and most climate models suggest that CAPE will increase as the world’s climate warms. But when Tippett and his colleagues looked at day-to-day estimates for CAPE over the contiguous United States between 1979 (the earliest data available) and 2015, they didn’t see any long-term increase. “This is an unexpected finding,” he notes.

But the team did find a substantial increase in another known risk factor for tornado formation: a parameter called storm relative helicity (SRH), which is related to the differences in wind speed and direction at various altitudes between ground level and 3 kilometers. Current models don’t suggest that this factor will increase as the climate warms, but the new study shows that it has in recent years. That, in turn, hints that scientists may not fully understand the link between SRH and climate.

“I think the work is well done and intriguing,” says Harold Brooks, a meteorologist at the National Severe Storms Laboratory in Norman, Oklahoma, who was not involved in the study. Overall, the team’s findings provide at least a partial answer to the disturbing trend of the last 50 years, he says. The link between the changes in severe tornado outbreaks and an increase in storm relative helicity is unsurprising but difficult to explain, Brooks says. “We don’t really have a good conceptual model for why SRH should increase as the planet warms.”

The next big question, adds Brooks, is to sort out whether long-term climate cycles such as the Atlantic Multidecadal Oscillation are playing a role in those tornado trends. The Atlantic Multidecadal Oscillation is related to sea surface temperatures in the North Atlantic Ocean and, like El Niño, it affects weather far away. Alternatively, researchers might find via further analyses that the increase in SRH is truly to blame. “My money is on the latter,” says Brooks, “but I want essentially even odds and small amounts being wagered.”