Tornado prediction

Print Friendly
Tornado prediction: How good?

The death toll from the May 22, 2011 tornado in Joplin – now 122 — was the latest tragedy of a horrific year for tornadoes. On April 27, twisters in Alabama and nearby states killed 314, the fourth highest in U.S. history. The 480 deaths in 2011 are already the highest number since 1953, and tornado season continues through mid-August.

Joplin, MO after the May 22, 2011 tornado

This slideshow requires JavaScript.

Photos of Joplin,KOMU News: slide 1, slide 2, slide 3

The Why Files asked Jonathan Martin, an expert on the large atmospheric disturbances that form tornadoes, some questions about tornado prediction. We edited the answers of Martin, a professor of atmospheric and oceanic sciences at the University of Wisconsin-Madison, after the interview.

The Why Files: What must we know to make a good tornado prediction?

Jonathan Martin: Tornado prediction is based on understanding the essential ingredients that are coming into play to forecast the storms that can produce tornadoes:

1. A very strong jet stream, which provides the necessary vertical wind shear — an increase of wind speed with height. This wind shear is what starts the funnel rotating.

2. A substantial amount of water vapor, especially in the lower troposphere. When this moisture condenses, it releases most of the energy that drives the storm — acting rather like a steam engine.

3. Warm, dry air at middle altitudes. In Tornado Alley, this air comes off the Mexican plateau and puts a lid on the warm, moist air building in the lower atmosphere. In the Southern plains, solar energy almost literally cooks the water vapor, but the cap prevents gradual release of this energy. Then, suddenly, an explosive thunderstorm occurs out of the blue sky and starts to release this energy, which is the source of power for the convective storms that create thunder, lightning and tornadoes.


Illustration of large cloud; arrows show air flows converging into a twisting funnel

This diagram shows how air flows converge to create a tornado.
The Why Files: Are predictions getting more accurate?

Martin: Yes. The ability to predict the likelihood of tornadoes has improved, especially in the one-two day range. We can say with fair confidence, “This wide area of Iowa is likely to be under the gun for tornadic storms, although they won’t occur everywhere in this area.”


Massive storm cloud and funnel cloud touching down on grayed landscape

Photo: NOAA; OAR/ERL/National Severe Storms Laboratory (NSSL)
This tornado tore through Seymour, Texas on April 19, 1979.

Such two-day forecast were available 10 years ago, but they did not garner much attention, because they were not that good. It’s interesting that most of this year’s major outbreaks have been forecast more than one day in advance.

Once the predicted day arrives, the emphasis shifts to monitoring with satellites and radar. We spent $4 billion networking the country with Doppler radar in the 1980s; this was a fantastic investment that has saved 10,000 lives, at a minimum. Last Sunday, radar is what gave people in Joplin the warning: “You have X minutes to find cover.” Undoubtedly that saved lives; Joplin could have been even worse.

Those three critical elements come in endless varieties and circumstances, and that’s where expertise comes into play: “How will today’s vertical wind shear, heat and humidity, and capping play out in terms of tornadoes?”

For short-term predictions, we are trying to understand exactly how a severe thunderstorm produces tornadoes. We have several viable theories, but they need to be tested more thoroughly. Still, predicting a tornado at a specific location several hours in advance is not something we can do. We may never be able to do this, but it may not be necessary, given the other improvements in prediction and warning.


Bar graph of April 1950 to 2011, 2011 has highest tornado count at about 875

Graphic: NOAA
Preliminary counts show about 875 twisters in April, 2011, the most since 1950. NOAA expects to issue a final count in a couple of months.
The Why Files: Why so much damage and death this year? Is this a result of climate change?

This tornado season is by no means over, and we are already at about 1,200 tornadoes, twice the average for this date. I’d guess we are not running at twice the level of EF 5 [the most intense tornadoes], but we have had the great misfortune that several of the 5s have hit heavily populated areas like Tuscaloosa and Joplin. That’s somewhat unusual, although it may be purely random.

The question we are asked is whether an increase in tornado intensity can be attributed to global warming. For the longest time, I said these are very small-scale disturbances, but I am beginning to think there is a link. Earth is warming, there can be no skepticism about that, and that may have a significant impact on the interaction between tropical circulation and temperate-zone circulation that is likely to form tornadoes in the central United States.

Warm areas near the equator in the western Pacific energize the spring jet stream, which flows to the middle latitudes and influences severe spring weather in Tornado Alley. For Tuscaloosa, Ala. on April 27, there is clear evidence that a precursor disturbance some days ahead in the far western equatorial Pacific had a significant and obvious hand in shaping the jet stream all the way to the Southeast, and was a big ingredient in producing these tornadoes. This is getting us beyond the vague notion that warming must be increasing the number of storms, and allows us to hang our hat on a particular kind of interaction, and test to see if it’s accurate.

— David J. Tenenbaum

Terry Devitt, editor; S.V. Medaris, designer/illustrator; Jenny Seifert, project assistant; David J. Tenenbaum, feature writer; Amy Toburen, content development executive


  1. National Severe Storms Laboratory.
  2. National Climatic Data Center on tornados.
  3. Interviews with NOAA experts on April 2011 tornado outbreak.
  4. 2011 tornado info.
  5. F5 tornados of the U.S.
  6. Joplin, MO Q & A.
  7. Interactive map: deadliest tornado years.
  8. Tornado basics.
  9. Tornado encyclopedia entry.
  10. Animation of 2011 tornado satellite imagery.
  11. Climate change could spawnmore tornados.
  12. Deadliest tornado season, but why?