Evidence to Date Does Not Show Clear Link Between Tornadoes and Climate Change

, senior climate scientist | May 22, 2013, 2:15 pm EDT
Bookmark and Share

In the wake of extreme weather events, people often ask scientists if they can be linked to climate change. Naturally, questions are being asked about tornadoes following the tragic losses suffered in the region of Moore, Oklahoma on May 20, 2013 after an EF5 Tornado.

This historic force of nature was met by heroic stories of lives saved through quick actions by teachers and neighbors to help others seek shelter, which are important to be told in the immediate wake of tragedy. Nonetheless, questions are naturally being asked and stories told about this tornado and connections to climate change.

The short answer is that scientists don’t see a clear link between climate change and the number or intensity of tornadoes over the past several decades. The Intergovernmental Panel on Climate Change special report on extreme events expressed “low confidence,” mainly due to inadequacies in monitoring systems.

Severe Tornado Occurrence in the US since 1954

I created this image to help dispel some myths in the news lately. Statement based on data from NOAA (see figure below). Photo from NOAA’s National Severe Storms Laboratory.

While a warmer planet could theoretically affect tornadoes, such as warmer and moister atmosphere, jet stream location changes, and equator to polar temperature gradient changes, we just don’t yet have enough data over the long-term to draw firm conclusions.

Tornadoes are rare, short-lived, and difficult to measure, and the record also has to account for a change in the tornado classification system. By contrast, scientists have many ways to measure and model heat waves, coastal flooding, and changes to precipitation patterns — and consequently have been able to draw definitive links between those phenomena and climate change.

Extreme Weather and Climate Change

The strongest links between historic extreme weather and climate change are for heat waves, coastal flooding, and changes in precipitation. Links to tornadoes and hurricanes are much less clear.

Thankfully, collecting more data on tornadoes has many benefits, especially for people who live in areas at risk for them. In fact, people can report tornadoes directly to NOAA’s National Severe Storms Laboratory and University of Oklahoma through a smartphone once they are safely sheltered.

Better data collection can help further improve warning systems for tornadoes. Minutes matter. The extra few minutes of warning compared to historical tornado warnings are now possible with advances in National Weather Service monitoring and tracking. Extra minutes to get into a shelter can mean more lives saved in a “Weather-Ready Nation.

Fast Facts about U.S. Tornado Trends

Our understanding of how a warmer planet affects storms that spawn tornadoes, as well as wind patterns, suggests more atmospheric energy could be available for tornado systems. On the other hand, tornadoes are difficult to model since specifics for individual tornado formation are still an area of active scientific research. Furthermore, tornadoes are tiny compared to the average grid spacing in most climate models.

Based on the data we have to date, there are no clear trends for tornado frequency or intensity. Tornado damage estimates depend in part on how many people and how much property is in its way.

The number of severe U.S. tornadoes EF3 or greater (those with gusts of greater than 136 miles per hour), has not changed much between 1954 and 2012. The most severe tornadoes, rated on damage, typically occur in KS, AR, TX, OK, TN and MO. As more people occupy areas affected by tornadoes, we have more potential damage, but we also have more eyes watching the sky and more ways to report and track tornadoes, resulting in more data than ever before.

Strong to Violent Tornado EF3-EF5 for the United States

U.S. Tornado Figures from NOAA

Update May 24, 2013 in response to comment:  Below is the figure for the annual count of US tornadoes EF1 or greater recorded in the US.

NOAA Tornado count US 1954 through 2012

Annual Count of U.S. Tornadoes EF1 or greater. Data from National Weather Service Storm Prediction Center at NOAA>


Feature image: NOAA

Posted in: Global Warming Tags: , ,

Support from UCS members make work like this possible. Will you join us? Help UCS advance independent science for a healthy environment and a safer world.

Show Comments

Comment Policy

UCS welcomes comments that foster civil conversation and debate. To help maintain a healthy, respectful discussion, please focus comments on the issues, topics, and facts at hand, and refrain from personal attacks. Posts that are commercial, self-promotional, obscene, rude, or disruptive will be removed.

Please note that comments are open for two weeks following each blog post. UCS respects your privacy and will not display, lend, or sell your email address for any reason.

  • Nathalie Andrews

    It is somewhat dated, but there was a really good article in Physics Today, March 2012, on this topic called Predicting and Managing Extreme Weather Events, which talks about the NOAA’s research on unusual weather and climate patterns. Its very informative about the scientific aspects of our current situation and the need for greater public understanding.

    • Thank you Nathalie for pointing out the article by Jane Lubchenco and Thomas R. Karl in the March 2012 issue (http://bit.ly/133V0Kw) that points in order to save lives, protect property and have a more “weather ready nation” takes support for a vast array of information. As they put it, “One needs to understand phenomena ranging from local vertical wind profiles for predicting tornadoes, to globe-spanning weather patterns for predicting heat waves, cold waves, and drought. Doing so calls for diverse observations from networks of weather balloons, radars, satellites, soil moisture sensors, ocean temperature sensors, and other monitoring devices. … We think that one essential key to meeting those challenges is critical environmental intelligence. Just like the intelligence of the security world, intelligence in the environmental arena combines data, analysis, modeling, and assessment.”

  • Roger Coppock

    Looking at only F3 to F5 tornadoes, and ignoring all others, is cherry picking the data set. This error alters one’s conclusion. When one looks at all the the data, the NOAA Storm Prediction Center’s dataset has a 15 +- 1 Tornados per year growth in the ‘lower 48’ states in the years 1950 to 2012. (R^2=0.69, P-value=2E-17)

    There is very little doubt about it. Observed Tornados have been increasing in the “Lower 48” States since 1950. Some, but certainly not all, of this increase is due to better observation and record keeping. Let’s be scientists, and leave the obvious cherry picking of data to the other side, OK.

    • Roger, thank you for suggesting we look at the full spectrum of tornado counts. I added the figure for the annual count of US tornadoes of EF1 or greater to update the post.

      • Roger Coppock

        Why are you stopping at 1954? The NOAA Storm Prediction Center’s data begin in 1950. Here are the yearly counts of unique tornadoes. (Unique “om” fields.)

        1950 201
        1951 260
        1952 240
        1953 422
        1954 550
        1955 593
        1956 504
        1957 858
        1958 564
        1959 604
        1960 616
        1961 697
        1962 657
        1963 463
        1964 704
        1965 897
        1966 585
        1967 926
        1968 660
        1969 608
        1970 653
        1971 889
        1972 741
        1973 1102
        1974 945
        1975 919
        1976 834
        1977 852
        1978 789
        1979 855
        1980 866
        1981 782
        1982 1047
        1983 931
        1984 907
        1985 684
        1986 765
        1987 656
        1988 702
        1989 856
        1990 1133
        1991 1132
        1992 1297
        1993 1173
        1994 1082
        1995 1235
        1996 1173
        1997 1148
        1998 1424
        1999 1339
        2000 1075
        2001 1214
        2002 933
        2003 1374
        2004 1817
        2005 1265
        2006 1103
        2007 1098
        2008 1692
        2009 1156
        2010 1282
        2011 1692
        2012 940

        Estimate Std. Error t value Pr(>|t|)
        (Intercept) -29612.654 2584.721 -11.46 <2e-16
        Years 15.406 1.305 11.81 2.1e-17

        Residual standard error: 188.3 on 61 degrees of freedom
        Multiple R-squared: 0.6957, Adjusted R-squared: 0.6907
        F-statistic: 139.4 on 1 and 61 DF, p-value: 2.1e-17


      • Thank you for pointing out the increase smaller tornadoes being reported which are a higher proportion of annual counts in the U.S. compared to the 1950s when there was less population, structures and scientific monitoring capabilities. The reasons I chose to display the bar charts from the NOAA/NWS Storm Prediction Center for EF1 or greater tornado counts annually is stated on this National Climatic Data Center site for U.S. tornado climatology (http://1.usa.gov/10ECAhU):
        “With increased national Doppler radar coverage, increasing population, and greater attention to tornado reporting, there has been an increase in the number of tornado reports over the past several decades. This can create a misleading appearance of an increasing trend in tornado frequency. To better understand the variability and trend in tornado frequency in the U.S., the total number EF1 and stronger, as well as strong to violent tornadoes (EF3 to EF5 category on the Enhanced Fujita scale) can be analyzed. These are the tornadoes that would have likely been reported even during the decades before Doppler radar use became widespread and practices resulted in increasing tornado reports.”

  • Al Bore

    Was 28 years of needless CO2 panic a war crime or just laughable Reefer Madness on steroids? “I see the signs of change….we cannot deny the change….”
    Science has never said in 28 years and to this day still refuses to say that their climate crisis from Human CO2 is as certain and as real and as eventual as inevitable as they themselves like to say comet hits are.
    So there you have it folks; Science believes a climate crisis is not as real as they can say a comet hit is.
    Get up to date: *Occupywallstreet now does not even mention CO2 in its list of demands because of the bank-funded and corporate run carbon trading stock markets ruled by politicians.

    • Tornadoes are similar to the geologic hazard of earthquakes in that we do not have long lead times to predict when and where they occur. So in many earthquake-prone regions, building codes exist to encourage engineers to design buildings to better withstand the shakes and rolls of the ground in order to save lives. Tornado shelters in historically prone regions can also help save lives as extra minutes are added to warning systems with better monitoring. Evidence is mounting with climate change that has already occurred, that more extreme events such as heat waves, coastal flooding, extreme precipitation and drought are testing the resilience of communities designed for the climate of the past. As Dr. Sullivan, Acting Under Secretary of Commerce for Oceans and Atmosphere and Acting NOAA Administrator said yesterday during a White House Summit (http://1.usa.gov/16f3iFt) “Past is not prologue” when it comes to planning for future risks associated with climate change.