A new analysis of tornado ratings on the Enhanced Fujita Scale is shining fresh light on a weird 12-year absence of EF5-rated U.S. tornadoes. Rather than natural variability – or a consequence of longer-term climate change – the absence of top-rated twisters may simply be an unintended consequence of the way that house destruction is now accounted for.
Before the Fujita scale was updated in 2007, a well-built house swept clean of its foundation was one of the hallmark indicators of an F5 tornado. But in the enhanced scale, the average value of winds expected to produce a clean sweep is at the very top of the EF4 range instead of reaching EF5. The result: Since the adoption of the enhanced scale, at least 13 high-end EF4 tornadoes have fallen short of EF5 ratings, despite each of them having swept site-built, well-constructed houses off their foundations.
“Where Have the EF5s Gone? A Closer Look at the ‘Drought’ of the Most Violent Tornadoes in the United States” is now in early release at the Bulletin of the American Meteorological Society. The three authors – Tony Lyza, Harold Brooks, and Makenzie Krocak – are all affiliated with the NOAA National Severe Storms Laboratory, and Brooks and Krocak are also at the University of Oklahoma.
The authors argue that a slight downward adjustment in the EF4/EF5 wind boundary would better harmonize the original and enhanced scales and produce a more consistent long-term record of top-tier tornadoes.
Unveiled in 1971 by eminent tornado scientist Theodore “Ted” Fujita with meteorologist Allen Pearson, the original Fujita scale was revolutionary. It divided a huge spectrum of winds – from the edge of minimal hurricane strength to the speed of sound – into 12 groupings, then used the first five of those groupings as the F1 to F5 ranges, topping out at 318 mph (512 km/h). Using the limited evidence available at the time, Fujita and Pearson then related the winds in each F range to the levels of damage one would expect.
After the scale debuted, there was a collaborative effort to analyze and rate past twisters in the National Weather Service database extending back to 1950, and meteorologist Tom Grazulis led a massive project to identify and document all F2 tornadoes going back to 1880. The Fujita scale soon became familiar to storm-savvy Americans, especially after the 1996 blockbuster film “Twister.”
Despite the Fujita scale’s success, scientists quickly recognized a few issues. For one, engineering studies soon found that peak tornado winds were unlikely to exceed much above 250 mph (400 km/h). Moreover, it was clear that winds needn’t be that strong to rip a house apart – especially a poorly built one. Several tornadoes rated F5 in the late 1990s could have produced the observed damage to homes even with winds well below the F5 range. In one famous example, a highly destructive tornado that struck La Plata, Maryland, in 2002 was initially rated EF5 but later “demoted” to EF4.
From this extended angst came a years-long process that led to the Enhanced Fujita Scale. One goal was to bring the wind ranges into more physically plausible territory while preserving an apples-to-apples relationship between the old and new scales so that the damage from an EF3 or EF4 would be essentially the same as from an F3 or F4. The enhanced scale is also much more comprehensive than the original version, with 28 different damage indicators, or DIs, each with its own scale of damage severity that corresponds to estimated wind speeds. The highest estimated wind from the various indicators is typically used to peg that tornado’s EF rating.
It was a couple of crucial choices in developing the EF scale – plus a rounding issue – that led to the apparent EF5 “drought,” according to the new study.
Based on engineering studies, the highest level of damage – “destruction of engineered and/or well-constructed residence; slab swept clean” – on the home-based damage indicator (“one- and two-family residences” of 1,000 to 5,000 square feet) was set to correspond to an expected peak gust of 200 mph (322 km/h), much lower than in the original scale. That would have been at the bottom of the EF5 range, except that the EF range brackets were rounded into 5-mph increments as the scale was finalized, so the originally calculated EF4 range shifted from 168-199 mph to 166-200 mph. The result: The expected wind gust of 200 mph for a well-built home swept off its slab was now at the very top end of the EF4 range.
As the study notes: “Under the strictest application of the EF scale, to attain an EF5 rating from a single-family home being swept off its foundation, the home must technically be built above building code, which is a fundamental break from the F scale and will inherently reduce the number of EF5 DIs found in surveys.”
How a more consistent EF5 record might look
The current stretch of nearly 12 years without an EF5 is unprecedented in the record going all the way back to 1880. There’s less than a 1% chance that natural variability would produce this long of a stretch without a single EF5 year, the authors calculated.
The most recent U.S. tornado to earn an EF5 rating struck Moore, Oklahoma, on May 20, 2013. As we pointed out that year, a total of 59 U.S. tornadoes were rated F5 or EF5 in the 63 years from 1950 through 2013. That’s a bit less than one per year on average. However, the distribution is quite uneven: Years can pass without a top-rated twister, and then a major outbreak can produce several on a single day, as with the 1974 and 2011 Super Outbreaks.
| Year | All tornadoes (F/EF0 to F/EF5) | Intense tornadoes (F/EF3 to F/EF5) | Violent tornadoes (F/EF4 to F/EF5) | F/EF5 tornadoes (Totals in asterisks include EF5 “candidates” from Lyza et al., 2025) |
| 1990 | 1133 | 53 | 15 | 3 |
| 1991 | 1132 | 46 | 7 | 1 |
| 1992 | 1313 | 58 | 14 | 1 |
| 1993 | 1173 | 36 | 6 | 0 |
| 1994 | 1082 | 35 | 5 | 0 |
| 1995 | 1235 | 31 | 11 | 0 |
| 1996 | 1173 | 23 | 3 | 1 |
| 1997 | 1148 | 39 | 10 | 1 |
| 1998 | 1424 | 43 | 8 | 2 |
| 1999 | 1339 | 64 | 13 | 1 |
| 2000 | 1075 | 23 | 3 | 0 |
| 2001 | 1215 | 29 | 6 | 0 |
| 2002 | 934 | 31 | 5 | 0 |
| 2003 | 1374 | 35 | 8 | 0 |
| 2004 | 1817 | 28 | 5 | 0 |
| 2005 | 1265 | 21 | 1 | 0 |
| 2006 | 1103 | 32 | 2 | 0 |
| 2007 | 1096 | 32 | 5 | 1 |
| 2008 | 1692 | 59 | 10 | 1 |
| 2009 | 1159 | 22 | 2 | 0 |
| 2010 | 1282 | 45 | 13 | 0 |
| 2011 | 1705 | 85 | 23 | 6 (*13) |
| 2012 | 939 | 30 | 4 | 0 |
| 2013 | 916 | 28 | 9 | 1 (*3) |
| 2014 | 929 | 27 | 7 | 0 (*2) |
| 2015 | 1178 | 21 | 3 | 0 (*1) |
| 2016 | 974 | 28 | 2 | 0 |
| 2017 | 1418 | 15 | 2 | 0 |
| 2018 | 1121 | 12 | 0 | 0 |
| 2019 | 1529 | 36 | 3 | 0 |
| 2020 | 1086 | 24 | 6 | 0 (*1) |
| 2021 | 1313 | 24 | 3 | 0 (*1) |
| 2022 | 1176 | 24 | 4 | 0 |
| 2023 | 1378 | 31 | 2 | 0 (*1) |
| 2024 | 1780 | 47 | 4 | 0 |
| Average (1990-2021) | 1227 | 35 | 6.7 | 0.6 |
What if the EF5 threshold were to be brought down from 200 to 190 mph? Under this definition, Lyza and colleagues found 13 tornadoes since 2008 that would be “upgraded” from EF4 to EF5 based on homes swept from foundations, and two others that would get similar promotions based on damage indicators other than homes.
With these upgrades in the mix (see asterisks in Fig. 3), the EF5 frequency over the past 17 years suddenly jumps above the 145-year average. However, most of this is due to the devastating Super Outbreak of April 26-28, 2011, which caused more than 300 deaths. All by itself, that outbreak produced four confirmed EF5s as well as five others that merit an upgrade, according to Lyza and colleagues.
Overall, as shown in Fig. 3, the upgrades would yield EF5 gaps since 2007 that are one to five years long – much more in keeping with the 145-year record.
The authors of the new study use their findings as a springboard for some provocative questions, including how to think about the very purpose of an EF5 rating, how best to incorporate new data on powerful near-surface winds detected by Doppler radar, and “perhaps most radically, should tornado ratings be more reflective of total impact, and not solely tied to wind speed estimates?”
What about climate change?
Unlike some well-established byproducts of human-caused climate change – such as intensified rains and hotter, more fire-supportive droughts – there aren’t any flashing red lights in the recent tornado record suggesting “more and more.” If anything, the frequency of intense or violent tornadoes has dropped just a hair over the last 20 years or so. However, the clumpy nature of the biggest outbreaks makes it tough to separate any such brief trends from natural variability. And small changes in how the most violent twisters are assessed can yield big differences, as the new study makes clear.
There are other ways in which tornadoes do seem to reflect a changing atmosphere, as we cover in detail in a Climate Explained post.
- Tornado seasons are getting more variable: Months and even years are swinging more dramatically from extreme quietude to extreme activity.
- Tornado outbreaks are getting larger and more frequent: Twisters appear to be clumping even more than usual, with the biggest outbreaks growing more prolific and longer, quieter periods stretching in between.
- Winter may be getting more tornado-dangerous: The unprecedented twin outbreaks of December 2021 are just one sign that warmer winters may already allow for more tornado-favorable weather setups.
- Tornado Alley is shifting eastward: Tornado-favorable weather is increasing across the vulnerable lower and mid-Mississippi Valley and declining slightly across the Southern Plains.
As the century unfolds, it’s possible other tornado trends will emerge, sometimes in surprising ways. The wind shear crucial for supercell thunderstorms is expected to diminish overall. But several recent studies – including this one from 2023 using global climate simulations and embedded high-resolution weather models – find that pockets of continued strong wind shear may coincide with increased instability to boost the odds of tornado-favorable supercells toward the eastern U.S., especially later this century and for higher-end emissions scenarios.
In the meantime, it’s all too clear that U.S. tornado vulnerability is on the rise right now. A sobering 2024 analysis in npj Natural Hazards led by Stephen Strader of Villanova University examined the traditional and emerging tornado hotbeds of the Southern Plains and mid-South to see how meteorological and sociological trends have intersected.
Results indicate that escalating vulnerability and exposure have outweighed the effects of spatially changing risk. However, the combination of increasing risk and exposure has led to a threefold increase in Mid-South housing exposure since 1980. Though Southern Plains tornado risk has decreased since 1980, amplifying exposure has led to more than a 50% increase in mean annual tornado-housing impact potential across the region. Stakeholders should use these findings to develop more holistic mitigation and resilience-building strategies that consider a dynamically changing tornado disaster landscape.
What Strader and colleagues have issued is a clarion call for boosting tornado resiliency – one worth keeping in mind as we head into peak tornado season with a National Weather Service that’s under immense, sudden, and unprecedented stress from mass firings and early retirements.
Jeff Masters contributed to this post.
!function(f,b,e,v,n,t,s)
{if(f.fbq)return;n=f.fbq=function(){n.callMethod?
n.callMethod.apply(n,arguments):n.queue.push(arguments)};
if(!f._fbq)f._fbq=n;n.push=n;n.loaded=!0;n.version=’2.0′;
n.queue=[];t=b.createElement(e);t.async=!0;
t.src=v;s=b.getElementsByTagName(e)[0];
s.parentNode.insertBefore(t,s)}(window, document,’script’,
‘https://connect.facebook.net/en_US/fbevents.js’);
fbq(‘init’, ‘568493883318626’);
fbq(‘track’, ‘PageView’, {“page_title”:”New study reveals potential cause of a ‘drought’ in violent EF5 tornadoes”,”user_role”:”guest”,”event_url”:”https://yaleclimateconnections.org/2025/03/new-study-reveals-potential-cause-of-a-drought-in-violent-ef5-tornadoes”,”post_type”:”post”,”post_id”:127811});
