A Fly in the Fun House

We in severe weather prediction often get questions like, “What kind of tornado season will we have?” or “Will the El Nino year give us more tornadoes in Kentucky”? I can’t give a definitive answer to questions like that, and anyone who claims they can is full of buffalo cookies. The reason is that tornadoes — or severe local storms of any kind — happen on time scales of seconds, minutes and hours. The question is being asked about weeks and months. In between: way too many things we just don’t know yet.

The key with tornadoes is that they are, indeed, local phenomena, usually lasting less than 10 minutes. Global weather patterns shift on scales of weeks, and even short-term climate like the ENSO (El Nino / Southern Oscillation) on the scales of months and years. There are even multidecadal cycles that influence climate, as well as global mean thermal cycles occurring over the span of decades, centuries and even millennia. They overlap, sometimes cancel, sometimes magnify, and many of their effects are not known well because we haven’t been around as a science long enough to sample their many possible juxtapositions.

In general: First comes the thunderstorm, then the severe storm, then the tornado. Let’s start with the overwhelming majority of strong and violent tornadoes. They come from supercells. To get supercells, you need four ingredients:

  • Moisture

  • Lift

  • Wind shear

  • Instability

    Processes acting on scales ranging from weeks to seconds create moisture, instability, shear and lift. Some are large in scale and involve hemispheric flow patterns, but the connections are very, very loose between something as big as hemispheric scale flow on the time scale of weeks, and a tornado in minutes or seconds.

    The closer in time, the more direct the connection. A boundary laid down by last night’s thunderstorms, for example, is known to play a far bigger role in tornado potential than equatorial Pacific water temperature. Several studies (to which I can direct one offline if desired) have found no statistically meaningful connection between El Nino and tornadoes; yet we often see storms spawn tornadoes upon interacting in certain ways with an old boundary.

    Remember those four ingredients — moisture, lift, instability and shear — because they are the keys that open the door to the “fun house” of violent thunderstorms and tornadoes. And even with those four seeming to be in place, even more has to happen on the scale of individual storms and their environments. To get a tornado, a bunch of dominoes have to fall very precisely in one part of one thunderstorm — far removed from all that planetary scale stuff.

    There are a lot of distorted images and mirrors in that meteorological fun house that can deceive the forecaster or storm observer. The tornado process is a buzzing fly hidden somewhere in those halls of the fun house and moves from day to day, sometimes hour to hour. My greatest challenge, whether as a professional forecaster or as a storm observer when off duty, is to predict the location of that fly. Which of many rooms is it buzzing through today, or is it even active? It doesn’t matter whether the house is shifted around a little some by slow outside forces. That doesn’t help me find the fly within and has no provable influence on where the fly chooses to land.

    In short, don’t look to climate for the answers to foretelling tornadoes. There’s just too much going on in between global trends and local tornadoes to make a connection. We have enough trouble predicting tornadoes in a couple of hours. It will be many lifetimes before we can predict them next month or next year.


    Leave a Reply

    You must be logged in to post a comment.