Data shows fewer tornado days in U.S. but more per event over past couple decades

Image: U.S. National Oceanic and Atmospheric Administration, via Wikimedia

A trio of researches with the U.S.'s National Oceanic and Atmospheric Administration (NOAA) has found that though there are fewer total days per year when tornados occur in the U.S., the number that occur on days when there are tornados has increased over the past couple of decades. In their paper published in the journal Science, Harold Brooks, Gregory Carbin and Patrick Marsh describe how they studied weather data over the past half century and what they found when looking for trends.

Tornados happen in many places, but because of its unique geography, the U.S. has more than any other country—mainly due to the lack of a large mountain dividing east and west. There has been speculation recently, that is causing more tornados to occur—though it has also been suggested it only seems that way because of how quickly information about tornadic events disseminates in the modern era. The trio at NOAA decided to let hard facts tell the story. They collected from the national storm database, which goes back to 1954, to see if they could coax out any patterns (they only included tornados at least as strong as an F1).

As it turns out, the trio did find a pattern, they say the data shows very clearly that the U.S. actually has a trend of having fewer days in which there is a tornado over the past two decades—that's the good news. The bad news is that on days when there is a tornado, there are more than there used to be. The data shows that back in the 1970's there were just .6 days a year that had 30 or more tornados—after the turn of the century, that number had risen to 3 days per year. Curiously, the numbers suggest that the country still experiences on average, the same number of tornadoes each year, approximately 1,200—they're just spread out differently. They also noted that the beginning and end of the tornado "season" in recent years has fluctuated more wildly than the years prior to that.

The researchers cannot say of course why the spread of tornados has changed in the U.S., though some might suggest it's due to global warming or even changes in atmospheric conditions in parts of the country due to pollution or other unknown factors. What is clear, is that something is causing a change, and there is now evidence of it, providing a path for moving forward for better understanding what is really going on.

Explore further

Magazine reporting below average numbers of tornados in 2013

More information: Science 17 October 2014: Vol. 346 no. 6207 pp. 349-352. DOI: 10.1126/science.1257460
Journal information: Science

© 2014

Citation: Data shows fewer tornado days in U.S. but more per event over past couple decades (2014, October 17) retrieved 20 September 2019 from
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.

Feedback to editors

User comments

Oct 17, 2014
An eminent display of "scientific" techniques of deceit.
Again, no one wants to mention that the Oklahoma Climatological Survey chart, which is only available online in a size as small as a postage stamp, shows tornado numbers were a constant 180 a year before about 1951. Note, no chart provided today goes before 1950. Liars claims the change in numbers of reports is because there are more people in the U.S. today. Since 1950, the population only tripled, while the number of tornadoes is seven times what it was in 1950! Where did more than 1000 tornadoes hide during those years?
With the flat land in most of "Tornado Alley" tornadoes could be seen from miles away, so you didn't have to be right next to them, in a densely populated area to see them!

Oct 17, 2014
More demonstrations of "scientific" lie techniques.
Talking about the "average" being "about" a certain amount is patently at least potentially misleading. Many think you can only take an average of a mostly constant set of numbers. You can take an average even if numbers are exploding. And, if they vary significantly with respect to a constant low value, the "average" can stay constant for awhile. But only awhile. So you need to know how large a set the average is over! The article doesn't mention that! Also, to say a value stays "about" the same means it varies by no more than, say, 25 percent. So, for awhile, the averages can increase by 25 percent every year and still deceitfully be called "constant". Also, averages diminish individual elements, so, if there were 3000 tornadoes this year, the average would stay the same!

Oct 17, 2014
Note, too, such fraudulent "measures" as looking only at a limited range of tornado strengths. The article only looks at F1-F5. But F0 represents winds of 45 to 73 miles per hour. You can have 100,000 of these a year, which could flatten the entire country, but it wouldn't show up in the article! It's like looking at the number of tsetse flies in Virginia and concluding the number of houseflies is almost non existent, too! Most "scientific" scamming on tornadoes has looked at F3 to F5 tornadoes and so concluded tornado numbers are not increasing! But that lie seems to be wearing thin.

Oct 17, 2014
And, consider the scam of "tornado days"! Tornado days says almost nothing! If you have 1 or one million tornadoes in a day, it's still called "one tornado day! You can have one billion tornadoes in one day and proudly declare the nation only had "1 tornado day! And, note, there can be no more than 366 tornado days in any year, which establishes an artificial, and illegitimate, maximum to this tornadic activity!
And the claim about "better reporting techniques is equally as much a lie as the claim about the population increasing. There was almost exactly the same type of equipment as today during the 1990's and tornado numbers have shot up at least 25 percent since then!

Oct 17, 2014
Why does God inflict all of the world's nastiest tornadoes on the Bible Belt?

Oct 17, 2014
You make lots of valid points. But until someone actually looks at the data and can verify what you are saying is actually happening, you're just drawing straws from thin air about what could potentially be happening.

I wish data like this weren't constrained to being behind pay walls, but please don't spread misinformation about things that might not actually be happening.

Oct 17, 2014
"If you have 1 or one million tornadoes in a day, it's still called "one tornado day!"

No. Three in one day equals three tornado days.

Oct 17, 2014
One data point against the Water_Prophet, if JuianPenrod is incorrect, and the article is right.

If more extreme single incidents would imply the ability of CO2 to lessen gradients until they burst through.

Not strong evidence, obviously there could be other factors, but it should be enough for the people on the site to get some licks in on me. Maybe I'll get some intelligent comments back.

I've been following the US midwest a bit, and have been astounded by the difference in tornadoes and tornado force winds, there has been a lack of distinction. I wonder if that is reflected here.

Oct 18, 2014
It is difficult to compare modern tornado counts with pre-1950 counts because the technology used to track tornadoes has vastly improved. Before weather radar, the only ways to track tornadoes were the human eyeball or by analyzing the path of damage that they left. Many were missed completely.

Today we have modern Doppler radar, storm chasers, smart phones with cameras, and a population density that has almost doubled since 1950. It's hard to miss a tornado in much of the Midwest and Eastern USA.

We also have a better understanding of tornadoes, so that a storm cell that drops multiple tornadoes along its path will have each one counted separately. Before 1950 it was often assumed that one tornado did all of the damage in the path of a cell.

Oct 19, 2014
gkam proves that the rating system on PhysOrg is nothing but a laughingstock.
I placed legitimate and non contestable points, but received only one 5 ranking and a collection of 1's intended, as the Hate Mafia working here hopes, to dilute the ranking of what I said. In contrast, gkam says three tornadoes in one day is three tornado days and they receive two 5 rankings! A tornado day is a day in which there were tornadoes. Three tornadoes in one day would be one tornado day. The article itself states that! It says, "fewer tornado days in U.S. but more per event". The "event" being referred to is a "tornado day". You aren't having more tornado says per tornado day. One tornado day is one tornado day. So the only "more per event" the article speaks of is more tornadoes per tornado day! That means that you can have any number of tornadoes in one day and still call it a tornado day!

Oct 20, 2014
"1" it's the new "5."
What's worse is some individuals hold multiple accounts.

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more