Skip to content

Tornado Stats

July 12, 2014

By Paul Homewood




Traditionally, US tornado counts peak in early June. It is encouraging then to be able to report that, so far, the tornado season has been another quiet one.




While not as low as last year’s record low count, it continues the trend of below average years, starting in 2012.


Many more tornadoes get to be reported these days, because of changes in observation procedures and technology, such as Doppler radar. To allow for this, the Storm Prediction Center produce “inflation adjusted” statistics of tornado counts dating back to 1954.

This method shows that, so far this year, tornado counts are well within the bottom 25th percentile.





  1. July 12, 2014 6:25 pm

    I’m just waiting for someone to claim that this year is 24% up on last year!

  2. July 12, 2014 7:38 pm

    Did you know that they changed the Fujita Scale back in 2008 — modified it? It’s called the Enhanced Fujita Scale. In Ontario, one guy was telling me how tornadoes increased three-fold from 2008 to present and this was proof on Climate Change. So I went and did some research and found that they changed the F scale in 2007 and it was put into action fully in 2008. Suddenly, Ontario went from 10 or 11 tornadoes (F1 to F3) per year to 27, 31, and so on. But if you look at the ratings for most of them, they’re classified as “F0”.

    In other words, what used to be classified as just a big storm is now classified as an F0. I can’t help but wonder if this was to further the ‘more frequent and severe storms’ predictions of the AGW crowd.

    • July 13, 2014 8:47 pm

      The original Fujita scale (and the modified one) are based on damage, from which you can estimate wind speeds. F0s, the weakest ones, are going to be the most common and cause the least damage, sometimes virtually indistinguishable from downdrafts/microbursts in gust fronts.

      Since we can sometimes see these now and couldn’t in the past (though they’ve always been there), you can either add them in and estimate numbers in the past or you can restrict your analysis to the strongest ones (F3+/EF3+) for which reporting bias is going to still happen but be lessened (you will still have survivor bias in that areas with no survivors can’t report and areas with no humans can’t report).

  3. Brian H permalink
    July 17, 2014 9:23 pm

    F0s are tornadoes with no funnel?

    • Brian H permalink
      July 17, 2014 9:24 pm

      F0s are …

      • July 31, 2014 7:32 pm

        EF0 tornadoes have a funnel. It is typically small, has the weakest winds and has a short life time.

Comments are closed.

%d bloggers like this: