Historic Tornado Outbreak

We have witnessed a tragic tornado outbreak this week, one for the record books. The fatality total as of this morning now definitely appears to have topped the April 1974 Super Tornado outbreak. The question many people have been asking is why this outbreak was so deadly. As a tornado impacts researcher, the important question is whether the death toll was due to this being a meteorological outlier event, or due to heightened vulnerability. The southeastern U.S. is vulnerable to tornado casualties; meaning that controlling for the strength, timing, and tornado path characteristics, tornadoes in the Southeast kill and injure more people than tornadoes elsewhere. In my recent book on tornado impacts with Kevin Simmons, we estimated state fixed effects for twenty eight states, and constructed a casualty index based on the fixed effects from various specifications. The indexes are based on tornadoes through 2007. A higher value of the index indicates that tornadoes in the state result in more casualties, everything else equal. Here are the index values and ranks (out of 28 states) for the states with fatalities in Wednesday's outbreak:

Alabama 1.63 23
Georgia 2.65 26
Mississippi 0.91 13
Tennseessee 1.78 24
Virginia 0.72 9

The tornado outbreak struck states which historically have been vulnerable. The Southeastern vulnerability is also associated with the off-season, night time, and mobile home vulnerabilities. While we are waiting for the individual tornado paths to be identified and rated, it is hard to draw conclusions about this event. But it does not fit the standard Southeastern vulnerability, as it occurred during the prime tornado season, and during the late afternoon hours. Many of the deaths in Tuscaloosa and Birmingham appear to have been in permanent homes. It appears to have been more of a problem of multiple long track, violent (F4 or F5) tornadoes.

Many observers have been comparing this outbreak with the 1974 Super Outbreak. As a social scientist who studies tornado impacts, it seems that the more significant historical parallel is the 1925 Tri-State Tornado. The Tuscaloosa - Birmingham super cell thunderstorm had a tornado warning on it continuously for over 300 miles. It hasn't been determined yet how long this tornado was more or less continuously on the ground, but I have seen references to a 200 mile long damage path. The Tri-State Tornado had a damage path of over 200 miles, and yet stands out there seemingly as this extreme outlier event, tracking across 3 states and definitely producing F5 damage across Illinois and Indiana. Most meterologists suspect it wasn't on the ground the entire 200+ mile distance, but it clearly was on the ground for most of this length. It was definitely an outlier event, but the question is how much of an outlier - is that a 100, 500, or 1,000 year event. If we do end up with something approaching a 200 mile damage path, this would really indicate that the Tri-State Tornado was perhaps more like a 100 year event, and creates the real possibility of a similar length storm at some point in the future.