As the US Southeast continues to endure an onslaught of severe thunderstorms and tornadoes, a recent study of tornado patterns in the US points to Smith County, Miss., as the tornado-risk capital of the United States.
For generations raised on the notion of a "Tornado Alley" in the US – running roughly from northern Texas, through the eastern Great Plains, and into the western Great Lakes states – the results may come as something of a surprise.
National Weather Service forecasters have long warned that while some areas have higher risks of tornadoes than others, every state in the union experiences twisters. States within Tornado Alley, along with Florida, have been viewed as the areas with the highest risk.
In the new analysis, the five counties in the eastern US facing the highest tornado risk – in this case, a high probability that in any two-year period, three tornadoes will track through the same 20-mile radius – include counties in Nebraska, Oklahoma, and Arkansas.
But Smith County topped the list.
Beyond the risk any individual county may face, perhaps the biggest surprise was the sheer breadth of the country at an elevated risk for tornadoes, says Grady Dixon, an assistant professor of meteorology and climatology at Mississippi State University.
"I knew going in that the southeast was going to have a generally elevated amount of risk," he says. "What I was surprised about was the east-west span of this elevated risk, stretching all the way to the Carolinas."
In effect, Tornado Alley in this study becomes a patchy blotch over much of the central and southeastern US.
"When you say 'tornado alley,' people think Great Plains, and rightfully so," he says. But from a hazard-assessment perspective, "we don't think it's wise to consider the Southeast as a separate entity or some sort of afterthought."
The aim of the study was to focus less on the location of the five or so most significant "bullseye" counties and more on where the areas of relatively higher and lower risk are in the broader regions the study covers.
Still, the magnitude of the risk in the area around Smith County was perhaps the biggest surprise.
Looking at tornado data with other techniques might shift the location of that high-risk location somewhat, he says. But the results indicate that the area should experience on average "more than one tornado every single year in any given 20-mile radius."
Professor Dixon says the study, published in the Bulletin of the American Meteorological Society in December, grew out of an interest in trying to define what had come to be known as Dixie Alley. It's a kind of southern analogue to Tornado Alley, but one that had received far less scientific scrutiny than its more-famous counterpart.
If his study is right, it suggests that Dixie Alley is actually just an extension of the classic Tornado Alley. Dixie Alley's assumed individuality stems from a region of relatively low tornado risk between them, near the Ozark and Ouachita Mountains.
But long-term tornado data are notoriously hard to work with, he and other weather researchers say. Numbers of tornadoes appear to have risen, an artifact of more people living in once-rural areas where few tornadoes were once spotted or reported.
In addition information on a tornado's full track is a relatively recent addition to the data pool. Typically, studies were based on noting where a tornado first touches down.
The team adjusted for the changes in tornado numbers by including the number of "tornado days" – which show no trend up or down over the study period. And they built their approach on looking at the full paths of tornadoes and how often those paths passed relatively close to each other.
Some tornado researchers aren't quite ready to sign off on the approach Dixon's team has taken.
The team's analytical approach may involve too few tornadoes in any one area to yield meaningful statistics, notes Harold Brooks, a tornado researcher with the National Severe Storms Laboratory, also in Norman, Okla.
Moreover, the classic Tornado Alley is distinctive because displays a stronger seasonal feature – most of its tornadoes are spawned in violent spring storms – than Dixie Alley, Dr. Brooks notes. The Southeast has about the same number of tornadoes, but they are more evenly spread out over a year.
Yet Dixon's team maintains that the density and lengths of tornado paths are at least as important for risk assessment as the number of twisters and the times of year they appear.
Given rapid population growth in the Southeast, and the generally longer paths tornadoes there etch across the landscape, the study can provide additional details for assessing tornado hazards in the US in general, and the Southeast in particular, the research team suggests.
"It would not hurt my feelings if in two years or five years, with better data or more advanced statistics, we can show this paper is not valid anymore," Dixon says.
In the meantime, what may be just as intriguing are the areas, within the wider region of elevated risk, where the risk is still relatively low. "I have yet to hear a reasonable explanation for why [those] particular areas" have few tornadoes, Dixon says.