Will we see an EF5 tornado in 2017?

Discussion in 'Introductory weather & chasing' started by Jeff Allen, Mar 12, 2017.

  1. Jeff Allen

    Jeff Allen Noob

    Joined:
    Mar 11, 2017
    Messages:
    1
    Likes Received:
    0
    What do you guys think? I myself think it's a good possibility. With us already having a violent tornado, the EF4 in Perryville MO, this early in the year I think an EF5 might happen in 2017. Heck the last EF5 happened almost 4 years ago. And they usually occur every few years. 59 have happened since 1950. So we're definitely due for one. The only question is where will the next one occur? Maybe in Oklahoma or Kansas, as usual? If you look at past years with a violent tornado happening so early in the season like this year an EF5 usually happened just months later. In early 2013 you had the EF4 that happened in Hattiesburg, Mississippi and then the Moore, OK EF5 struck May 20. And then 11 days after that the largest twister in history was observed near El Reno. Back in 2008 you had early season EF4's that occurred in Arkansas and Tennessee during February and then the Parkersburg, Iowa monster EF5 happened in May. Do you agree?
     
  2. Charles Ryan Mauk

    Joined:
    Jul 20, 2015
    Messages:
    21
    Likes Received:
    1
    It's possible. I'm just hoping for a decent amount in the Dakotas. Rurally...and big would be great yes :)


    Sent from my iPad using Tapatalk
     
  3. Ben Holcomb

    Ben Holcomb Digital Janitor
    Staff Member

    Joined:
    Mar 30, 2008
    Messages:
    1,021
    Likes Received:
    568
    Hard to say, hope we don't have any EF5's.
     
    • Agree Agree x 1
  4. James Hammett

    James Hammett Member

    Joined:
    Sep 25, 2006
    Messages:
    227
    Likes Received:
    31
    I definitely hope not since serious injury and loss of life are inevitable in violent tornadoes. Gambler's fallacy reminds us that the probability of a future event occurring is statistically independent of past occurrences. Trying to discern patterns from early-season events (or lack thereof) doesn't tell us much especially for something as rare and as small-scale as a violent tornado.
     
    • Like Like x 1
  5. Paul Knightley

    Joined:
    Feb 6, 2006
    Messages:
    786
    Likes Received:
    56
    I think the answer is the same as whether you asked this question in any other year: maybe. Beyond that there's nothing to add! Some EF-5 tornadoes occur in situations which look very, very ominous (April 27th, 2011); others occur on days which look a bit less ominous (May 22nd, 2011). But determining whether these days will or will not occur more than a few days out is simply not possible.
     
    • Like Like x 1
  6. Jeff Duda

    Jeff Duda Resident meteorological expert
    Staff Member

    Joined:
    Oct 7, 2008
    Messages:
    2,428
    Likes Received:
    733
    It's hard to take a question like this seriously, as there is really no way whatsoever of predicting such an event. Not only are such events exceedingly rare, but whether or not a tornado gets rated EF5 is dependent on human factors as well as meteorological. On the large scale, the conditions needed for EF5 strength tornadoes occurs on dozens of days from May through September across some part of the US. So some of the necessary ingredients are there pretty frequently. However, there is a lot of detail on the smaller scales (down to the tornado scale) that we neither observe regularly nor can we predict even hours in advance that determine whether a given tornado will attain a rotational velocity to satisfy EF5 criteria, whether a given storm will produce a tornado, or whether or not storms will even develop. Even if you get a tornado with rotational velocity above the EF5 threshold, it has to hit a strong enough damage indicator so that the maximum DOD is not exceeded and the estimated wind speed from said DOD is > 199 mph.

    Everyone here knows El Reno (2013) was really an EF5, but in the official records it was an EF3.
     
    • Like Like x 7
    • Agree Agree x 1
  7. Tim Supinie

    Tim Supinie Member

    Joined:
    Mar 2, 2009
    Messages:
    111
    Likes Received:
    40
    Statistically speaking, we have about a 50% chance to see an EF5 in any given year. The exact number depends on how you do the calculation.

    For instance, in the SPC database, 30 of the 66 years have an EF5 tornado, which is about 45.5%.

    For a slightly different method, there are 60,114 tornadoes in the database, and 59 of them are EF5's. We have about 1000 tornadoes in a year, so that works out to be 1 - (1 - 59/60114)^1000 = ~63%. So the probability that there will be at least 1 EF5 tornado in 1000 tornadoes (about 1 year's worth) is about 60%. The difference between the numbers is because some years, particularly 1974 and 2011, had more than 1 (E)F5 tornado. There are also some caveats to using the SPC database that anyone who has strong opinions on the rating of the 2013 El Reno tornado is acutely aware of, but it's what we have.
     
    • Like Like x 2
  8. Jeff Duda

    Jeff Duda Resident meteorological expert
    Staff Member

    Joined:
    Oct 7, 2008
    Messages:
    2,428
    Likes Received:
    733
    Good thinking, Tim!

    I'll offer my opinion that the exact probability depends more on which question you answer (and ask) rather than how you perform the calculation. I'd tend to believe the first number Tim gave to be more representative since it answers the question "if you sample any year from the database (assuming those years represent a reasonable sample from climatology), what are the odds that year will contain at least one (E)F5 tornado?" That seems to represent the chances of seeing an EF5 in any given year.

    The second calculation answers the question "if you take 1000 random samples of tornadoes from the historical database, what are the odds that at least one is an (E)F5?" This calculation requires you to know (or estimate) the yearly tornado total and ignores any statistical progression throughout the year (i.e.,what if you happened to only sample tornadoes from July and later [there are fewer (E)F5s in the database later in the year]?). Nonetheless, If you have an active year with 1400 tornadoes, for instance, then you could use that formula to arrive at a probability of ~75%. For a lesser active year of 800 tornadoes, the probability becomes ~54%.

    Since the advent of the EF-scale in 2007 (so 10 years), there have been undisputed EF5s in 4 years (2007, 2008, 2011, and 2013), for a rate of 40% over recent times. IMHO, however, if you account for other tornadoes that were rated less than EF5 but might have had EF5 winds in them, the odds are probably a tick higher (closer to 50-60%).
     
    • Like Like x 1
  9. Tim Paitz

    Tim Paitz Member

    Joined:
    Apr 27, 2015
    Messages:
    170
    Likes Received:
    39
    There's no way to know for sure until the tornado surveys are conducted after the storm. The next week or two will be interesting in terms of severe weather.
     
  10. Andy Wehrle

    Andy Wehrle Member

    Joined:
    Oct 10, 2004
    Messages:
    958
    Likes Received:
    48
    One thing I have noticed is there seems to be a sharp dropoff in survivability from the low end of EF4 to the high end. Take the St. Louis tornado of April 22, 2011, Tuscaloosa tornado of April 27, 2011, and the Hattiesburg tornado of 2013. As EF4s, they are all more intense than about 98.5% of all other tornadoes. They all struck in densely populated metro areas. One (St. Louis) even struck after dark and passed near a major airport terminal. Yet it caused no deaths, and neither did Hattiesburg. Yet Tuscaloosa, which occurred in broad daylight on a well-forecast high risk day and was broadcast live on television for nearly an hour before it arrived in the city, killed 64. Just looking at the rating won't tell you the whole story of the tornado's intensity. The St. Louis and Hattiesburg storms did isolated pockets of low-end to moderate EF4 damage and were (correctly) rated based on the single most intense destruction surveyed. The Tuscaloosa tornado, on the other hand, did EF4 to borderline EF5 damage across a much larger portion of its entire path. I have no doubt this tornado was capable of EF5 damage at some point in its life, but the EF4 rating is justifiable since there is no clear-cut instance of EF5 damage that could be found.
     
  11. Randy Jennings

    Joined:
    May 18, 2013
    Messages:
    139
    Likes Received:
    82
    Another way to look at the 56 F5/EF5 tornados in the SPC database from 1950 to 2015:

    median (middle number in a given sequence of numbers when it’s ordered by rank): 0
    mode (most frequent value in a set of data): 0
     

Share This Page