Yes, you read that correctly. Despite the recent spate of deadly twisters, including those that tore through the Midwest over the weekend, the scientific evidence shows that strong to violent tornadoes have actually beendecreasing for the past 58 years, and it is possible that the explanation lies with global warming.
That is certainly not the popular impression. Senator Barbara Boxer, Democrat of California, commented in May on the powerful tornadoes that had just hit Oklahoma. “This is climate change,” she said, adding: “You’re going to have terrible storms. You’re going to have tornadoes.” Commenting around the same time, Michael E. Mann, a prominent climatologist, was only slightly more cautious. He said, “If you’re a betting person — or the insurance or reinsurance industry, for that matter — you’d probably go with a prediction of greater frequency and intensity of tornadoes as a result of human-caused climate change.”
But the evidence shows the opposite.
I am not talking about global warming per se, which I am convinced is real and caused by man-made emissions of greenhouse gases. But not everything attributed to global warming has a scientific basis.
Let’s begin by consulting the relevant data from the National Oceanic and Atmospheric Administration. On its website, NOAA has a chart plotting the total number of tornadoes recorded in the United States from 1950 to 2011. At first glance, the increase looks dramatic: The number of tornadoes in 1950 was only 200, and by 2011 it had shot up to 1,700.
But the 200 number is ridiculously small. If it had truly been that low in the past, “The Wizard of Oz” never would have been written. And indeed NOAA accounts for those early small numbers by explaining that originally only the most violent tornadoes were recorded, leading to lower reported totals. By contrast, these days even little storms are logged; backyard dust devils are reported for insurance claims, and Doppler radar makes it hard for any tornado to escape notice.
What about in more recent decades? If you look at the numbers plotted on the NOAA chart for 1976, you’ll see that only about 800 tornadoes were recorded in the United States — around half of today’s numbers. This increase also looks dramatic.
But here, too, the increase is a function of enhanced reporting. The scientists at NOAA note explicitly that the rise in the numbers has come almost exclusively from the improved documentation of so-called category EF0 tornadoes — the ones that are so weak that their legacy consists only of broken branches and damaged billboards. As another NOAA website warns, “This can create a misleading appearance of an increasing trend in tornado frequency.”
Misleading indeed. If you count all tornadoes except the EF0 tornadoes, then the number has remained fairly steady from 1954 to now.
But enhanced reporting might be distorting more than just the numbers of the very weakest storms. So let’s consider only the most violent tornadoes, the ones in categories EF3 to EF5. A tornado of EF3 is “severe,” with winds of 136 to 165 miles per hour. Roofs of strong houses are ripped off; trains are overturned. EF4 and EF5 are worse, hurling cars and destroying even well-built homes. These devastating tornadoes are always reported.
The NOAA chart shows that the number of these storms has been significantly decreasing over the past 58 years, from over 50 per year in the first half to under 40 per year in the second. The statistical significance of this decrease is extremely high: well above 99 percent confidence.
How can this be? What about all the recent horrible tornadoes? What about the fact that a tornado in Oklahoma in May set a width record of 2.6 miles?
It is wise to be cautious about the panic that sets in when a storm kills a large number of people. People search for reasons to believe the storms are worse than in the past, even if the numbers contradict them. Victims naturally wish to explain why loved ones died and they look for a villain — and they can find one in global warming.
But global warming does not obviously lead to increased or more violent tornadoes. It is possible, for instance, that the increased energy brought by the higher temperatures of global warming is less significant than global warming’s reduction in the north-south temperature difference (the poles warm more than the Equator). The latter could reduce the kind of hot-cold weather fronts that generate severe storms. The current climate models are simply unable to make a clear prediction, and reduced tornadoes from global warming are just as plausible as increased ones.
One thing is clear, however: The number of severe tornadoes has gone down. That is not a scientific hypothesis, but a scientific conclusion based on observation. Regardless of the limitations of climate theory, we can take some comfort in that fact.