Tornadoes, Extreme Weather And Climate Change, Revisited

The big tornado outbreak, including a monster Oklahoma twister, have people asking again about a possible link to climate change. I’ll review the science in this post.

“The news helicopter from caught this image of the shocking near-total destruction of a huge area of Moore, Oklahoma, on May 20, 2013.” Via Masters.

Tom Karl, the director of the National Climatic Data Center, explained in a 2011 email:

What we can say with confidence is that heavy and extreme precipitation events often associated with thunderstorms and convection are increasing and have been linked to human-induced changes in atmospheric composition.

Insured losses due to thunderstorms and tornadoes in the U.S. in 2012 dollars. Data and image from Property Claims Service, Munich Re.

Tornadoes “come from certain thunderstorms, usually super-cell thunderstorms,” explained climatologist Dr. Kevin Trenberth in an email today, but you need “a wind shear environment that promotes rotation.” Global warming may decrease the wind shear and that may counterbalance the impact on tornado generation from the increase in thunderstorm intensity.

Trenberth, the former head of the Climate Analysis Section of the National Center for Atmospheric Research, notes:

The main climate change connection is via the basic instability of the low level air that creates the convection and thunderstorms in the first place. Warmer and moister conditions are the key for unstable air.

The climate change effect is probably only a 5 to 10% effect in terms of the instability and subsequent rainfall, but it translates into up to a 32% effect in terms of damage. (It is highly nonlinear). So there is a chain of events and climate change mainly affects the first link: the basic buoyancy of the air is increased.  Whether that translates into a super-cell storm and one with a tornado is largely chance weather.

After April 2011 saw records set for most tornadoes in a month and in 24 hours — “The Katrina of tornado outbreaks“ — I examined the climate/tornado link in great detail here, looking at the data, the literature, and expert analysis. That piece concluded:

  1. When discussing extreme weather and climate, tornadoes should not be conflated with the other extreme weather events for which the connection is considerably more straightforward and better documented, including deluges, droughts, and heat waves.
  2. Just because the tornado-warming link is more tenuous doesn’t mean that the subject of global warming should be avoided entirely when talking about tornadoes.

Early March 2012 saw what was likely “the most prolific five-day period of tornado activity on record for so early in the year,” as meteorologist Dr. Jeff Masters put it.

Then we had an unusually long “tornado drought” from May 2012 to April 2013, which has now come to a stunning end, punctuated by the devastating Moore, Oklahoma tornado yesterday:

A massive, mile-wide supercell tornado ripped through the suburbs of Oklahoma City, destroying homes, schools and other buildings. The tornado was on the ground for some 40 minutes, according to the National Weather Service (NWS), and police reported that an occupied elementary school was in the path of the cyclone. Early estimates had winds on the ground near 200 mph, which would have made the cyclone an F4 or higher. Witnesses said the damage was like something out of an atomic bomb strike, and there are at least 24 people dead, including many young children, with a toll that could eventually be far higher.

Masters says “the Moore tornado likely to be one of the five most damaging tornadoes in history,” which is particularly tragic because Moore had “previously experienced the 4th costliest tornado in world history, the notorious May 3, 1999 Bridgecreek-Moore EF-5 tornado.”

You can donate to the American Red Cross disaster relief here.

Below is an extended review of the scientific literature along with some analyses from this year and last year by leading experts.


For decades, scientists have predicted that if we kept pouring increasing amounts of heat-trapping greenhouse gases into the atmosphere, we would change the climate. They specifically predicted that that many key aspects of the weather would become more extreme — more extreme heat waves, more intense droughts, and stronger deluges.

As far back as 1995, analysis by NOAA’s National Climatic Data Center (led by Tom Karl) showed that over the course of the 20th century, the United States had suffered a statistically significant increase in a variety of extreme weather events, the very ones you would expect from global warming, such as more — and more intense –precipitation. That analysis concluded the chances were only “5 to 10 percent” this increase was due to factors other than global warming, such as “natural climate variability.” And since 1995, the climate has gotten measurably more extreme.

Multiple scientific studies find that indeed the weather has become more extreme, as expected, and that it is extremely likely that humans are a contributing cause (see “Two seminal Nature papers join growing body of evidence that human emissions fuel extreme weather, flooding that harm humans and the environment” and links therein).

Beyond that, as Dr. Kevin Trenberth, former head of the Climate Analysis Section of the National Center for Atmospheric Research, explained here in 2010: “There is a systematic influence on all of these weather events now-a-days because of the fact that there is this extra water vapor lurking around in the atmosphere than there used to be say 30 years ago. It’s about a 4% extra amount, it invigorates the storms, it provides plenty of moisture for these storms.” He told the NY Times, “It’s not the right question to ask if this storm or that storm is due to global warming, or is it natural variability. Nowadays, there’s always an element of both.”

Trenberth elaborated on this in a must-read 2012 article, “How To Relate Climate Extremes to Climate Change.”

And the latest science says that shockingly rapid loss of Arctic ice is also likely driving an increase in extreme weather.

Munich Re, one of the world’s leading reinsurers, issued a news release in 2010, “large number of weather extremes as strong indication of climate change,” which noted:

Munich Re’s natural catastrophe database, the most comprehensive of its kind in the world, shows a marked increase in the number of weather-related events. For instance, globally there has been a more than threefold increase in loss-related floods since 1980 and more than double the number of windstorm natural catastrophes, with particularly heavy losses as a result of Atlantic hurricanes.

The rise in natural catastrophe losses is primarily due to socio-economic factors. In many countries, populations are rising, and more and more people moving into exposed areas. At the same time, greater prosperity is leading to higher property values. Nevertheless, it would seem that the only plausible explanation for the rise in weather-related catastrophes is climate change. The view that weather extremes are more frequent and intense due to global warming coincides with the current state of scientific knowledge as set out in the Fourth IPCC Assessment Report.

Here is their data:

I have tended to focus on the extreme weather events for which the causal chain is clearest and which will do the most damage to the most people in the foreseeable future. Dust-Bowlification is probably at the top of that list (see “We’re Already Topping Dust Bowl Temperatures — Imagine What’ll Happen If We Fail To Stop 10°F Warming“).

But tornadoes are among the most visibly and viscerally destructive events — though I do find it interesting how much media coverage these tornadoes have gotten compared to, say, Tennessee’s 1000-year deluge aka Nashville’s ‘Katrina’. So it was inevitable that scientists would be asked the obvious question of whether recent remarkable outbreaks are connected to human-caused climate change — and they were indeed remarkable.

Many scientists have weighed in on the climate-tornado link. Two of the people who have done more research and publication on extreme weather and climate change than most are Trenberth and Karl, now director of NCDC. I emailed Karl for his thoughts and here is what he wrote me in 2011:

Best info we have on the relationship between a warmer world and severe convective storms that can produce tornadoes is in the 2008 Synthesis and Assessment Report of the US Global Change Research Program. Chapter three of that Weather and Climate Extremes Assessment indicates that several studies do show that environmental conditions favorable for convection are more likely with more greenhouse gases, but results are not conclusive.

We now have improved resolution models running at our Oak Ridge Supercomputer thanks to the Stimulus funding. We may be able to make more definitive statements (one way or the other) after these get analyzed over the next few years. Meanwhile, we know that La Nina years tend to have a greater chance of severe outbreaks. So as usual, there are natural factors that have to be considered, and any human made factors would be confounded within these naturally occurring events making our attribution much more difficult.

Joe, what we can say with confidence is that heavy and extreme precipitation events often associated with thunderstorms and convection are increasing and have been linked to human induced changes in atmospheric composition.

You can find that 2008 Report here. I wrote about it here (see Sorry, deniers & delayers, Even Bush Administration says human emissions are changing the climate).

Trenberth made clear to me in an extended 2010 interview that he was dismayed by the media coverage of extreme weather, especially extreme deluges, that made no mention whatsoever of global warming:

I find it systematically tends to get underplayed and it often gets underplayed by my fellow scientists. Because one of the opening statements, which I’m sure you’ve probably heard is “Well you can’t attribute a single event to climate change.” But there is a systematic influence on all of these weather events now-a-days because of the fact that there is this extra water vapor lurking around in the atmosphere than there used to be say 30 years ago. It’s about a 4% extra amount, it invigorates the storms, it provides plenty of moisture for these storms and it’s unfortunate that the public is not associating these with the fact that this is one manifestation of climate change. And the prospects are that these kinds of things will only get bigger and worse in the future.

I emailed Trenberth in 2011 to check his quote in ThinkProgress. And I again checked with him in 2012. He stands by the quote with the clarification he had added of the context:

It is irresponsible not to mention climate change in stories that presume to say something about why all these storms and tornadoes are happening.

The environment in which all of these storms and the tornadoes are occurring has changed from human influences (global warming). Tornadoes come from thunderstorms in a wind shear environment. This occurs east of the Rockies more than anywhere else in the world. The wind shear is from southerly (SE, S or SW) flow from the Gulf overlaid by westerlies aloft that have come over the Rockies. That wind shear can be converted to rotation. The basic driver of thunderstorms is the instability in the atmosphere: warm moist air at low levels with drier air aloft. With global warming the low level air is warm and moister and there is more energy available to fuel all of these storms and increase the buoyancy of the air so that thunderstorms are strong. There is no clear research on changes in shear related to global warming. On average the low level air is 1 deg F and 4 percent moister than in the 1970s.

Just because attribution is difficult doesn’t mean that the subject of global warming should be avoided entirely when talking about tornadoes. Equally important, when discussing extreme weather and climate, tornadoes should not be conflated with the other extreme weather events for which the connection is considerably more straightforward and better documented:

There are lots of posts from 2011 on this. Stu Ostro, Weather Channel Senior Meteorologist, wrote in his May 2011 post, “The Katrina of tornado outbreaks“:

The atmosphere was explosively unstable with summerlike heat and humidity, interacting with a classic wind shear setup as a strong jet stream and upper-level trough crashed overhead”….

The atmosphere is extraordinarily complex, and ultimately what’s happened the past month is probably a combination of influences, including La Nina, other natural variability, and anthropogenic global warming.

Here is how meteorologist and former hurricane hunter Dr. Jeff Masters put it May 31, 2011:

In summary, this year’s incredibly violent tornado season is not part of a trend. It is either a fluke, the start of a new trend, or an early warning symptom that the climate is growing unstable and is transitioning to a new, higher energy state with the potential to create unprecedented weather and climate events. All are reasonable explanations, but we don’t have a long enough history of good tornado data to judge which is most likely to be correct.

Here is how Masters put it in 2012:

Last year’s tornado season was incredibly severe, and we are off to one of the worst early-season starts to tornado season on record now in 2012. However, it is too soon to ring the alarm bells on climate change being responsible for this. The tornado data base going back to 1950 doesn’t show an increasing trend in strong tornadoes in recent decades. While climate change could potentially lead to an increase in tornadoes, by increasing instability, it could also decrease them, by decreasing wind shear. I’d need to see a lot more bad tornado years before blaming climate change for the severe tornado seasons of the past two years. One thing that climate change may be doing, though, is shifting the season earlier in the year. The 5-day total of tornadoes from February 28 – March 3 will probably break the record of 131 set in 1999 for the largest tornado outbreak so early in the year. Warmer winters, and an earlier arrival of spring due to a warming climate, will allow tornado season to start earlier–and end earlier. This year’s early start to tornado season is consistent with what we would expect from a warming climate. I have a more extensive article on this subject that has just been published by Weatherwise magazine….

Then we have the study The Effect of Climate Change on Tornado Frequency and Magnitude:

There is an obvious increase in tornado frequency between 1950-1999. This could be due to increased detection. Also this could be due to changing climatic conditions. Looking at the raw data we have seen that there are generally less tornadoes in El Nino years compared to La Nina Years. But, since we were unable to get climate data, we were unable to see if the change in the frequency was due to climate factors.

Our data has failed to show a strong correlation in increase in tornado frequency and magnitude during El Nino and La Nina events.

The jury is out.

A NYT blogger directed us to this chart:

tornado trends

There is no apparent trend in the strongest tornadoes (F5 is the most destructive). The NYT blogger quotes Harold Brooks of the National Severe Storms Laboratory:

The primary changes appear to occur ~1975, most likely as a result of the retrospective rating process that assigned ratings to tornadoes prior to the near-real-time ratings that began when the [National Weather Service] adopted the F-scale operationally in the mid-1970s, and ~2000, for reasons that aren’t completely clear, but are likely due to an increased emphasis on examining construction details and policies that changed the nature in how the ratings are created for the strongest tornadoes. Both have lead to a decrease in probability of a tornado being very strong, given that it’s strong. It’s possible that there’s a meteorological component, but the reporting practice changes are large enough that I don’t think we can pull a physical signal out, even if it’s fairly large.

So it may simply be that the data is simply is too confused by the reporting practices for analysis to draw any strong conclusions. That doesn’t mean the question shouldn’t be asked or that scientists shouldn’t give their best answer.

In general I do think it’s best to avoid statements like “global warming is to blame for” or “global warming caused” or “this is evidence of global warming,” especially in regards tornadoes.

Finally, while tornadoes will continue to grab the headlines wherever they flatten cities and take lives, it is virtually certain that other extreme events — and ultimately the permanently changed climate — will cause the greatest harm attributable to human emissions of greeenhouse gases.

The population hasn’t even acclimatized to the climate change we’ve had already — in part because the GOP and the fossil-fuel-funded disinformation campaign have obfuscated efforts to inform the public.

We’ve only warmed about a degree Fahrenheit in the past half-century. If we keep listening to the disinformers, we are on track to warm nearly 10 times that this century (see literature review here). In short, we ain’t seen nothing yet. Or, as one commenter put it:

“Mother nature is only warming up.”

And yes we need to improve housing for those in tornado alley. That’s a great thing for blogs that don’t focus on climate to write about. Just as obviously we need an aggressive strategy for reducing carbon pollution. that also supports real adaptation.

26 Responses to Tornadoes, Extreme Weather And Climate Change, Revisited

  1. Lawrence says:

    The Climate Desk website notes that the past decade has seen the 6 least active and 4 most active tornado years on record. So the average hasn’t shown a change, but the standard deviation is much greater. One wonders if this could be related to a changing jet stream pattern influenced by the Arctic…

  2. I am not 100% sure if you can completely blame this tornado on climate change, but I think its could be partially to blame.

    One role I am surprised Joe you have not mentioned is the roll urban sprawl can play in the chances of more deadlier tornadoes as cities spread out more into rural areas, like Moore OK/Oklahoma City. I think that is another thing city planners need to take into consideration is start considering limiting urban sprawl. Then you would have less risk of these disasters.

  3. prokaryotes says:

    When a storm occurs today, it is in a more favorable environment, could move slower thus become bigger.

    Climate change caused by greenhouse-gas emissions from fossil-fuel burning does not mean uniform global warming – in the Arctic, the relative increase of temperatures, amplified by the loss of snow and ice, is higher than on average. This in turn reduces the temperature difference between the Arctic and, for example, Europe, yet temperature differences are a main driver of air flow.

    Additionally, continents generally warm and cool more readily than the oceans. “These two factors are crucial for the mechanism we detected,” says Petoukhov. “They result in an unnatural pattern of the mid-latitude air flow, so that for extended periods the slow synoptic waves get trapped.”

    Will update the post later with more data and Joe’s article from today.

  4. Robert in New Orleans says:

    CO2 is like testosterone to the body of the Earth and what we are seeing is planet wide roid rage.

  5. prokaryotes says:

    Doppler radar measures 318 mph wind in tornado

    Scientists measured the fastest wind speed ever recorded, 318 mph, in one of the tornadoes that hit the suburbs of Oklahoma City on May 3, 1999.

    The record-setting wind occurred about 7 p.m. near Moore, where the tornado killed four people and destroyed about 250 houses.

  6. Joan Savage says:

    The LA Times interviewed tornado expert Kevin Knupp.

    Knupp said that he and his graduate student watched the tornado in real time, tracking its path using radar. He said he thought that another dissipating storm in the area may have influenced the megatwister’s formation, creating a surge in momentum in low-level air that might have contributed to the circulation of the larger storm.,0,1385189.story

  7. Merrelyn Emery says:

    It sounds like the only predictable thing in our future will be that it will become more unpredictable, ME

  8. Paul Klinkman says:

    Officially the Moore tornado was an EF5 and 1.3 miles wide. There has never been an EF5 ever recorded that was anywhere near that wide, and few enough tornadoes over a mile wide.

    Furthermore, the tornado was moving more slowly than most, giving it time to drill down on anything below.

    Events on the ground, or just a few feet above above ground level as it were, suggest that we’re seeing a consistent pattern of super-tornadoes, twisters far more intense and damaging than would be predicted by current climate change factors. We can wonder whether this effect of climate change might continue and intensify in the coming decades.

  9. prokaryotes says:

    I think we witness the beginning of a new atmospheric regime which creates more violent weather extremes.

    Katrina and Sandy type of hurricanes, Joplin and Moore tornadoes are the new normal. And then you have more extreme Microburst too, which affect aviation negatively.

  10. Rabid Doomsayer says:

    Regardless of climate change these sort of disasters have happened before and will happen again. Houses can be built to withstand much more than they do and at very little extra cost.

    Take a look at the Monolithic Domes website. That style of construction has the advantage of being very energy efficient. Yes I know some domes are really ugly but each to his own

  11. Mulga Mumblebrain says:

    With the huge added energy being trapped in the atmosphere, how could the weather extremes not become more violent? I’m just waiting for the first incident where a plane is torn apart by some extraordinary atmospheric perturbation.

  12. Geoff Beacon says:

    Don’t worry. Climate change not so bad after all.

    Has anyone any comments on the source of this story Energy budget constraints on climate response in Nature Geoscience by Otto et. al.? They say

    The energy budget of the most recent decade does, however, indicate a lower range of values for the more policy-relevant transient climate response (the temperature increase at the point of doubling of the atmospheric CO2 concentration following a linear ramp of increasing greenhouse gas forcing) than the range obtained by either analysing the energy budget of earlier decades or current climate model simulations.

    How does this fit with Foster and Ramsdorf’s Global temperature evolution 1979–2010?

    Otto seems to maintain that there has been a slowing of global temperature rise. Foster and Ramsdorf seems to tell a different story.

    Does Otto et. al. mean these extreme weather events won’t increase in frequency by much?

  13. Merrelyn Emery says:

    The schools that were hit did not even have minimal protection despite this being a known high risk area. Surely, this is a result of the neglect of the communal interest because of the ideology in favour of small govt, ME

  14. Raul M. says:

    Monolithic dome homes may be built underground as some show.

  15. MarkF says:

    “In general I do think it’s best to avoid statements like “global warming is to blame for” or “global warming caused” or “this is evidence of global warming,” especially in regards tornadoes.”

    Why is that best?

    I expect many non scientists do relate this storm to global warming. I know I do.

    The public discussion of this is a nasty PR campaign, not a lecture in a university.

    And one side is fighting with one hand tied behind its back.

  16. Joe Romm says:

    Already posted on this.

  17. Joan Savage says:

    Joe is warning us against giving sole attribution.

    Global warming is not the sole cause of events. Even a phenomenon as ‘clean’ as rising ocean level is not purely from swelling water and melting ice, because humans divert other water to the ocean.

    Large-scale factors including the tides and the coriolus effect are still part of the complex reality of how the world works.

    Tornadoes are complex and it is not proven how – or if – global warming affects them. As an ecologist, I’m used to systems thinking so it would not surprise me if some relationship is identified at some point, but we don’t have the evidence now.

    Let us stick with verifiable statements about global warming, the truth is tough enough.

  18. Robert in New Orleans says:

    Wait until the other shoe drops in terms of increased rates for homeowners and business insurance. This is of course assuming that the insurance companies just don’t abandon the market altogether.

  19. prokaryotes says:

    It doesn’t even require that much. During the starting/landing phase downward bursts of 30-40 mph can cause a plance to crash.There have been like 36 plane crashes like this in the USA from like the aprox 50’s-90’s. Though i do not have the current numbers.

    Explained here at around 1:43:00

  20. Rabid Doomsayer says:

    One trend that seems to happening is the greater clustering of tornado events, less tornadoes in low tornado periods more tornadoes in high tornado periods. I have not seen any stats so it could be my imagination.

  21. prokaryotes says:

    James B. Elsner, an atmospheric scientist at Florida State University:

    “Climate change increases the available energy for tornadoes through a warmer and moister atmosphere. Wind shear decreases in the global mean, but this might be irrelevant locally when the jet stream dives southward like it did last weekend across the Plains.

    “I believe there is evidence that the strongest tornadoes are getting stronger. They are certainly getting longer and wider.”

  22. Mulga Mumblebrain says:

    I do recall tragedies caused by wind shear in the USA, but I was thinking of increased turbulence in mid-flight, high up. Clear air turbulence I know a very little of, but I wonder if the general turbulence of the upper atmosphere, and the extremes, will worsen as the quantum of energy in the atmosphere inexorably grows.

  23. MarkF says:

    Thanks, and thanks for all of the information you provide at this site.

  24. Harold Brooks says:

    There have been 1.3 mile wide EF5 tornadoes before (e.g. Greensburg, KS), but that’s misleading. The difference between F4 and F5 assessment is small (it depends on what gets hit) and the reported width data is subject to a lot of issues and shouldn’t be treated as exact truth. As more effort is put into surveying storms, the width (which is the maximum width at any point along the track) and, to a lesser extent, the length, will become bigger. This is a big problem when you look at long time series of width and length. Basically, they are records of reported length and width, not necessarily actual length and width.

    In the case of the 2013 Moore tornado, an important note is that the maximum width was reported relatively early in the path (EF2 damage) and the single building rated EF5 was where the path width was ~3/4 of a mile.