Are Extreme Tornados a Harbinger of Global Warming?

Not so far.


Moore 2013 Tornado
Credit: Wikimedia

"This is climate change. We were warned about extreme weather: Not just hot weather, but extreme weather," Sen. Barbara Boxer (D-Calif.) declared on the floor of the U.S. Senate last week. "You're going to have tornados and all the rest." The senator's impassioned outburst of climatological alarm was provoked by a mile-wide tornado that had just struck Moore, Oklahoma. The same twister provoked a Washington Post piece to speculate that record insurance losses in the past five years may be the result of tornados and other storms "possibly ramped up by climate change."

The May 20 tornado has been classified as an EF-5, with peak winds at around 210 miles per hour'"a storm in which even well-built houses are leveled off of their foundations and automobile-sized projectiles fly more than 100 yards through the air. It killed 24 people, injured nearly 400 others, destroyed between 4,500 and 5,000 structures, and damaged another 15,000 to 20,000. Overall, it may have inflicted as much as $5 billion in property damage and other losses. The level of destruction was so great because the storm stayed on the ground for 40 minutes and traveled 17 miles.

Are extreme tornados becoming more frequent, as Sen. Boxer's remarks suggest? The National Climate Data Center does say that there's been "an increase in the number of tornado reports over the past several decades." But it also notes that the increase is likely a spurious result attributable to more extensive Doppler radar coverage, increasing population, and greater focus on tornado reporting. 

In particular, Doppler radar has made it much easier to detect EF-0 tornados (with wind speeds of 65 to 85 mph), resulting in a dramatic rise in the numbers reported. If you exclude those, the agency reports, "There has been little trend in the frequency of the stronger tornadoes over the past 55 years." Similarly, the United Nations' Intergovernmental Panel on Climate Change's 2012 report on trends on weather extremes concludes that there is no evidence for either an increase or a decrease in tornado frequency or intensity.

While the number of strong tornadoes has stayed steady, the rate of deaths from tornados has fallen about 90 percent in the last 90 years, from just under 2 per million in the 1920s to 0.2 per million recently. This is true even taking into account the tragic toll of 2011, in which 553 Americans were killed by the storms. The biggest tragedies in 2011 occurred when tornados ripped through Joplin, Missouri, and Tuscaloosa, Alabama, killing 158 and 64 people, respectively. The fact that warning times have increased from less than 5 minutes to 13 minutes since the 1970s accounts for much of the trend toward a lower death rate. In addition, Americans living in tornado-prone areas take twister warnings more seriously and engage in protective measures more actively.

What about the insurance losses cited by the Washington Post? Do they suggest anything about trends in tornados? Again, no.

In 2012, the University of Colorado political scientist Roger Pielke Jr., the Austin College economist Kevin Simmons, and the Troy University economist Daniel Sutter normalized U.S. tornado damage from 1950 to 2011. Their analysis published in the journal Environmental Hazards, provides an estimate of the damage that might be expected if past tornados or tornado seasons occurred today. Normalization takes changes in population, wealth, and housing stock into account, along with inflation.

In another analysis'"this one for the Geneva Association, an international insurance think tank'"the three researchers compared the tornado damage in the standout years of 1953, 1965, 1974, and 2011. In 1953, 519 people died and normalized damages exceeded $20 billion, in current dollars; in 1965, 301 people were killed and damages were greater than $20 billion; in 1974, 366 people died and losses were more than $10 billion; in 2011, 553 people were killed and damages totaled about $26 billion. But the losses in those years are way above the general trend: Overall, there was a 63 percent decline in normalized damages across the six-decade period. 

While not dispositive, their analysis suggests that this normalized decline in damages may mean that damaging tornados are becoming fewer. They add, though, that the tornado damage incurred in 2011 could indicate that "maximum damage levels have the potential to increase should societal change lead to increasing exposure of wealth and property." In other words, a richer and more populous America could provide more targets for future tornados to hit. 

So as Americans become wealthier, tornado damage may get worse. On the other hand, casualties will likely fall, as technological advances increase warning times, enabling people to get out of way or take shelter sooner. Sen. Boxer is right about this much: Tornados are a kind of extreme weather. But on the evidence so far, she is wrong to suggest that the Moore tornado is related to climate change.