A tornado happened where? Really?

Take a look at this tower cam video from Springfield, MA from Wednesday afternoon. The camera clearly shows what was likely, at that point, an EF1 tornado developing over downtown Springfield, crossing the Connecticut River and continuing east. On radar, the storm showed a classic “debris ball” signature (radar actually picking up debris thrown upward by the tornado) for many miles beyond Springfield. At its strongest, the tornado may have strengthened to EF2 or even EF3 intensity once it moved east of Springfield.

Many seemed shocked when they heard of the tornado’s location. It seemed surprising to see a tornado “there.” But is it really? Though most common in tornado alley, the plains are not exclusive holders of the rights to all tornadoes across the United States. There are several tornado “hot spots” beyond our area that are just as likely to see them, just during different times of the year. Take a look at this map:

The map above (provided by UALR) shows areas at most risk of seeing a tornado. This is calculated based on the number of tornadoes “per square mile” over the last few decades. It’s no surprise that “tornado alley,” composed of Texas, Oklahoma, Kansas and Nebraska is in the “highest risk” category. But look around. The area around Tampa, Florida has a high risk. Jackson, MS to Chattanooga, TN has as high of a risk as Tornado Alley, as does central Arkansas (more on this below) & Indiana. Notice any other areas? Yep! Northwest Connecticut, southern New York and western Massachusetts. The area is a “micro tornado alley.” Cool dry air is readily available from down-sloping winds off the Appalachian Mountains and warm humid air is easily moved northward from the Gulf Stream (which runs just off the coast). The counties in this area in fact, see as many “tornadoes per square mile” as some counties in central Oklahoma. There aren’t more tornadoes in southern England. It’s that the few that do happen per year (2-4), happen in the same concentrated area giving it a high “tornado per square mile” count. So though the Springfield tornado was rare in how well it was captured live on camera, it was not all that rare for the region that it occurred in.

So what about that “bullseye” over central Arkansas? It’s oriented SW to NE, by the way, because that’s how most tornadic storms in Arkansas move, from southwest to north east, seemingly paralleling I-30 (a coincidence in the alignment of the interstate with the most common track of a tornadic storm in Arkansas). Does central Arkansas really see that many more tornadoes than the rest of the state? The simple answer is yes. There may however be more to this than meets the eye.

Click on the map to the right and look at it closely (another great work by the folks at UALR). The map shows two pieces of information together. First, tornadoes per county. The counties in the brightest red have had the highest number of tornadoes from 1950-1998. The green dots show areas of greatest population density. Here are my thoughts on this:

1.) Yes, these may be something to the fact that more tornadoes occur at the intersection point of the hills and the delta, much like in southern New England. The counties in brightest red are in that “transitional” region. There may however be another factor in play.

2.) That factor may be the population itself. The NWS issues tornado warnings based on radar information all the time. Just because radar is saying there is a tornado doesn’t necessarily mean that one is actually occurring. Many times, in fact, the radar sees rotation but a tornado is never produced. A tornado is not counted unless it is seen by human eyes. Reports of damage from local law enforcement are later surveyed by a team from the NWS and labeled as tornado damage based on what they see. The NWS however will not go out to survey an area unless they actually get a report. It may not be a coincidence that there are more reports of tornadoes in places where there are more people, in counties that have higher population densities (indicated by the green, the more green, the more people live in that county). There may very well have been, especially prior to the 1980s, many locations in rural areas that saw tornadoes but b/c the storm was never seen by a human or the damage was never found by anyone, the storms were never counted. This may also mean, that as populations grow and expand into more rural areas, the number of reported tornadoes will go up, not because there are actually more tornadoes happening, but because we are simply more likely to see them or their damage.

Cities have expanded outward quite a bit over the last few decades. For example, look at these two satellite pictures taken of the Atlanta, Ga metro in 1973 and again in 1993. The city has basically doubled in 20 years and has grown even more since 1993. The chances of a tornado tracking through Atanta in 1973 were smaller because the city was smaller. We’ve basically made the cities, the targets of tornadoes, larger and thus easier to hit. In fact, it’s easy to be bold and predict that Atlanta, Georgia will see more tornadoes move through the metro in the next 30 years than the previous 30 years. That’s just based on pure chance. Though once rare, storms like the Oklahoma City tornado in 1999, Atlanta in 2008 (image right), Joplin in 2011 and Salt Lake city in 1999 will not only happen again but will happen in other cities. Birmingham, Memphis, Dallas, Kansas City and yes, even Little Rock, have a much higher chance of seeing tornadoes move through their downtown cores than ever before, simply because those “downtown cores” are now much larger in size.

Keep this in mind as you watch some of the “hyped up,” “outbreak ridden” tornado coverage on some media outlets in the future.

–Greg “Weather Dork” Dee