More Tornadoes in the Biggest U.S. Outbreaks--for an Unexpected Reason

The largest U.S. tornado outbreaks have been spitting out an ever-increasing number of twisters, according to a study published Thursday in the journal Science. The new paper, led by Michael Tippett (Columbia University), reinforces prior work showing that U.S. twisters are increasingly concentrated in big outbreaks, with the quiet periods becoming even quieter. Tippett and colleagues also threw in a noteworthy curve ball. It appears the growing number of tornadoes in the most extreme outbreaks is not because of greater instability--something that long-term climate change is expected to produce--but instead as a result of increased storm-relative helicity, one measure of the “spin” produced by vertical wind shear that translates into rotation within tornadic supercell thunderstorms. As the paper puts it, “Tornado outbreaks with many tornadoes are increasingly frequent, but lack the presently understood meteorological signature of global warming.”


Figure 1. This EF2 tornado south of Dodge City, Kansas, on May 24 was one of 87 twisters observed in a five-day outbreak from May 22 to 26, 2016. Image credit: Bob Henson.

Signs from the worst of the worst
The new study examines tornado outbreaks for the period 1965 - 2015 using extreme value analysis. This is a technique especially well suited for delving into very infrequent events, where the outliers of a “long-tail” distribution are the topic of keenest interest. The authors limited their scope to tornadoes of at least F1/EF1 on the Fujita/enhanced Fujita scales, given that the vast increase in the number of people spotting, chasing, or otherwise finding twisters has artificially boosted the number of weak tornadoes (F0/EF0). The trends in outbreaks emerged most clearly when looking at those with at least 12 tornadoes, and when placing those 435 outbreaks into five percentile brackets (e.g., dividing each year’s outbreaks into five groups based on the number of tornadoes per outbreak). This revealed a significant increase in the number of tornadoes per outbreak, especially for the largest events. From the late 1960s to the early 2010s, the number of tornadoes per outbreak in the top fifth of outbreaks rose from around 20 to more than 35. Likewise, the size of the largest outbreak one might expect every five years roughly doubled, from 40 in 1965 to nearly 80 in 2015.

Since there hasn’t been any significant growth in the number of outbreaks themselves, or in reliable tornado reports, it appears that tornado production is increasingly concentrated in the biggest outbreaks, with fewer twisters in between. It’s a conclusion reached in recent years through work by Harold Brooks (National Severe Storms Laboratory) and others. The 2010s have featured increasingly wild swings between hyperactive and relatively tranquil periods. One example: for the period from April 15 to July 31, the U.S. lurched from a near-record total in 2011 (including the catastrophic Super Outbreak and the horrific Joplin tornado) to a record-low total in 2012. Early 2015 was also remarkably quiet on the severe weather front. And the latter half of drought-ridden 2016 has pushed the year’s tornado total close to record lows, even when we factor in the spate of more than 50 twisters reported on Tuesday and Wednesday that led to at least 5 deaths.


Figure 2. Marcia Remick, of Rosalie, Alabama, digs through the debris and aftermath of a damaging tornado on Thursday, December 1, 2016. The EF3 tornado carved a 13-mile path through the area on Wednesday, killing 3 people. Image credit: AP Photo/Brynn Anderson.

Shear bemusement
What’s behind the clustering of tornadoes into ever-more-extreme outbreaks? That’s the most intriguing part of the new paper. It’s long been expected that human-produced climate change would lead to an increase in warm, moist conditions at lower levels of the atmosphere as compared to upper levels. This instability is often expressed in the form of CAPE: convective available potential energy. Global climate models also tend to project a decrease in vertical wind shear over the nation.

Putting two and two together, one might expect that more overall CAPE and less overall shear would work against any increase in tornadoes--but that’s not necessarily the case, according to a landmark study published in 2013 and led by Noah Diffenbaugh (Stanford University). That paper, which drew on the most recent suite of global climate modeling in support of the IPCC process, found that projected reductions in wind shear tend to be concentrated in days that aren’t very unstable to begin with, whereas there may actually be an increase in days with enough CAPE and enough shear to produce tornadic storms. “This is the key result of our 2013 paper: changes in the mean aren’t indicative of the changes in the occurrence of extremes,” Diffenbaugh told me.

What Tippett and colleagues found in looking at outbreaks from the last 50 years is something quite different than the scenario of future tornado activity from Diffenbaugh and colleagues. It turns out that CAPE hasn’t yet increased significantly in the extreme-outbreak settings, but wind shear in these settings, as measured by storm-relative helicity (SRH) in the lowest 3 kilometers of the atmosphere, has increased, with the biggest increases found for the largest outbreaks. “This is an unexpected finding,” Tippett said. “The fact that we don’t see the presently understood meteorological signature of global warming in changing outbreak statistics leaves two possibilities: 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.”

There’s no immediate explanation for why SRH during outbreaks has been on the increase. Could it be related to some form of multidecadal climate variability? The authors didn’t find any strong relationship to signals such as the Atlantic Multidecadal Oscillation, the Pacific Decadal Oscillation, or U.S. temperature trends. “I’m not convinced it's [because of] long-term climate oscillation,” Harold Brooks told me. “I think SRH could be changing because of global warming in ways we don't understand yet.” Tippett said: “We’re definitely raising some questions without answers yet. If global warming is changing tornado activity, then we might expect to see either continued increases in the number of tornadoes per outbreak or at least no return to earlier levels. On the other hand, if multidecadal variability, anthropogenic or natural, is responsible, then a return toward earlier levels might be possible.”

Is it too early to detect a human-produced signal in tornado outbreaks?
Although it might seem so at first glance, the new paper isn’t really in conflict with the longer-term projections from the work of Diffenbaugh and colleagues. The climate models analyzed by that group don’t show an increase in tornado-favorable environments kicking in strongly until decades from now, assuming that global temperatures will by then reach the commonly cited (and feared) 2°C threshold over preindustrial values. “We see some indication of emergence between the present and the 2°C level of global warming, particularly in the spring over the eastern U.S.,” said Diffenbaugh. “It’s really after 2°C that we see very clear emergence across the climate models and in multiple seasons.” For now, he added, it’s important to clearly determine exactly what changes in tornado activity have occurred across recent decades, especially given the complexities of the data. “I think [the Tippett et al. paper] is an important step in this line of research.”

We’ll be back with our next post on Monday.  Have a great weekend, everyone!

Bob Henson


Figure 3. Cunningham Park in Joplin, Missouri, has become a memorial to victims of the tornado of May 22, 2011, which killed more than 150 people in Joplin. The memorial includes a large ring (foreground) celebrating the human spirit. In the background of this photo from April 21, 2012, is St. John’s Regional Medical Center, which was severely damaged by the tornado and eventually demolished. St. John’s reopened in 2015 as Mercy Hospital in a new facility four times larger than its predecessor. Image credit: Stephanie Himango/NBC/NBC NewsWire via Getty Images.