Tornadoes are becoming more frequent in populated parts of the United States and are often occurring as damaging clusters — a development seen in recent deadly outbreaks from Alabama to Michigan.
The number, damage and deadliness of individual tornadoes has held roughly steady over the past 50 years, federal experts with the National Oceanographic and Atmospheric Administration told The Hill.
But broad shifts in the patterns of how tornadoes occur will pose serious challenges to policymakers and emergency managers across the South and Midwest — even as risks remain in the traditional heart of Tornado Alley.
The role to which climate change is a factor in these shifts is unknown, and the changes in tornado behavior overall represent a major meteorological mystery.
Warm winter means a powerful start to storm season — but not much else
The first quarter of 2023 represented a powerful start to the year’s tornado season, with each month offering a number of tornadoes substantially above the historical average, according to federal data.
“This past winter and our early spring has been as active as I can remember in years,” said William Bunting of the national Storm Prediction Center. “The number of events, many of the same areas being affected has been remarkable.”
This January had more than three times as many tornadoes as the historical average; February had half again as many, and March had twice as many.
Many of these storms have been both dramatically large and deadly. Tornadoes also have strayed far outside their usual domain — including touchdowns in Delaware and the suburbs of Los Angeles.
In March alone, for example, a four-day outbreak of 31 tornadoes killed 22 people in a belt across the Southeast. That was followed a few days later by a line of storms that drove 66 tornado touchdowns from Alabama to Indiana, killing 27.
The increased winter tornado activity — a recurring pattern over the last several years — is unusual, Bunting added.
On the most basic level tornadoes require two factors to form. There has to be hot, moist air rising to create movement of energy through the atmosphere, and there has to be powerful circular wind shear — or vertical changes in wind speed and direction — to bring a funnel cloud to earth.
Southeastern tornadoes usually require warm, moist and unstable air moving up from the Gulf of Mexico — weather that is usually blocked by cold air moving down from the northern latitudes.
But several years of warm winters have changed that pattern, Bunting said — allowing tornado conditions to penetrate as far north as Michigan — far in advance of the usual start of tornado season.
Does that mean it’s going to be a brutal year for tornadoes? Like almost everything else about tornadoes, that’s hard to say, Bunting said.
“An early start to a tornado season doesn’t have any predictive power on how the year will turn out,” he told The Hill, pointing to several historic years where a deadly winter gave way to a quiet spring.
Two illusions at center of issue
Recent reporting about increasing tornadoes conceals an important fact: as far as we can tell, tornado numbers have been pretty consistent as long as we’ve been keeping records, Harold Brooks, a senior research scientist at NOAA’s Severe Storms Laboratory, told The Hill.
Major tornadoes — those that merit rating on the Fujita scale — have been “a relatively consistent 500 per year, and we don’t see any evidence that the intensity has changed,” Brooks said.
The apparent increase in tornado frequency — and damage — turns out to rely on eccentricities and changes in the way American meteorologists gather data as much as any change in tornado behavior itself.
As scientists get better at tracking or counting tornadoes, an increase in the number of tornadoes observed reflects improved science as much as it does more frequent cyclones.
Then there is the question of damage caused by tornadoes. A simple look at the data shows that both fatalities and amounts of damage in dollars have increased over the past century — which seems to suggest that tornadoes are getting worse.
But this is a mirage too, Brooks said: an illusion caused by the sheer fact that America in 2023 has three times as many people, and “more stuff in the way,” than America in 1923.
“There are more televisions in your house than great grandparents did, and in your kitchen compared to great grandma’s kitchen, the appliances cost more,” he said. “So if your kitchen is destroyed, it will cost more than it would for your great-grandma — but that doesn’t mean the tornado is worse.”
With these factors taken into account, the deadliest tornadoes in U.S. history happened more than a century ago: an 1890 tornado that hit Louisville, Ky. and an 1896 one that devastated St. Louis, Mo. in 1896, Brooks said.
“The old big [tornado] days and the current big days look a lot alike,” Brooks concluded.
Tornadoes are expanding in space and time
What has changed is how the cyclones are distributed across both the country and the year — a phenomenon scientists still struggle to explain.
Over the past 30 years tornadoes have begun to cluster, with far more months with either zero tornadoes — and far more punctuated by series of devastating storm complexes which may drop dozens.
In the last 20 years, the U.S. has set 8 monthly records for the most tornadoes per month — and 7 for the fewest.
“On days when tornadoes occur, more of them per event — and they’re occurring more often further east,” said Bunting of the Storm Prediction Center.
The distribution of those tornadoes has also changed, shifting broadly east of the I-35 corridor, which is a rough biogeographical line separating the historic boundary between the arid, sparsely populated plains and the relatively wet, densely populated eastern U.S.
Bunting emphasized that the rising risk in the East doesn’t mean that risk has disappeared for the historically tornado-prone Plains — though the region has had a few quiet years.
Rather, “tornado alley is expanding,” he said.
Scientists aren’t sure why this is happening. One common culprit is climate change: a January study in the Bulletin of the American Meteorological Society found that supercell thunderstorms — the spinning, unusually stable weather patterns that give rise to 70 percent of tornadoes — would move broadly east as the climate warmed and the Western Plains gave way to desert.
That could lead to an accompanying increase in tornadoes, by increasing the availability of warm, wet air — or decrease them, as higher temperatures suppress wind shear, Brooks said.
While it’s tempting to blame any change since the 1970s on a hotter climate, “we don’t have the physical links that go from ‘If the planet warms, oh, that means we should be having more variability,’” Brooks said.
Other theories attribute the change in tornado patterns to deeper, less understood patterns in the Earth’s ocean and atmosphere. Scientists have proposed that the change could be driven by shifts in decades-long cycles in the atmospheric and oceanic currents; in shifts in atmospheric waves; or the increased ‘wobbliness’ of the jet stream.
Policymakers — and residents — need to be alert
Brooks emphasized that the change in tornado patterns don’t yet rise to the level where individuals should change their migration decisions because of them.
“The numbers we’re talking about over 40 years are basically 10 percent changes,” he said. “It’s gone from being a woman once in a decade to being one every nine years, or one more or one fewer over the course of a lifetime. Not enough to change your life.”
But in policy terms, these numbers begin to look much more significant, he said — particularly when the differing population patterns of those areas come into play.
For emergency managers in, say, central Tennessee, a 10 percent increase in tornadoes — while insignificant on the individual scale — is a substantially increased threat, leading to substantially increased costs, Brooks said.
That threat is magnified by the fact that while Western cities like Topeka, Kansas, or Amarillo, Texas may look roughly like their Eastern counterparts, the rural hinterlands of eastern states are far, far more densely populated.
While a tornado 10 miles outside of Lubbock, Texas may do little damage, a tornado outside similarly-sized Birmingham, Ala., risks devastating many smaller communities.
That danger is magnified by widespread poverty in those areas — exemplified by the difference between waiting out a tornado in a fortified shelter versus in a mobile home.
The unpredictability of tornadoes has a silver lining: it means that while summer is usually the most active part of the year, it’s just as possible that the deadliest part of the season is already behind us, Bunting said.
“In a year like this where it’s just been so incredibly active early on, but it could be another 2012 where things get very quiet as we get into April, May, June. We just don’t know. It doesn’t look particularly active over the next seven to 10 days there, but we all need to be paying attention.”
To residents across the Eastern U.S. — including the unusually quiet parts of Tornado Alley — Bunting urged caution.
The year’s active start is “a reminder to do the things that we preach year in year out. Which is: Make sure you have a severe weather plan. Make sure you have multiple ways of receiving warnings. Make sure you pay close attention when severe thunderstorms are forecast in your area. And then be ready to take action when the warnings are issued for your area.”