Here we are three days after the Dixie Tornado Outbreak, and the death tolls keep climbing.
While I still urge us all not to get fixated on specific numbers, especially since they still are in a state of flux, the toll is 342 as I type, fourth highest known (and subject to further change). This tally soon should slow its agonizing climb (I hope, anyway!) and settle horizontally. It is certainly an order of magnitude greater in terms of fatalities than any tornado outbreak since 3 April 1974; and indeed, already surpasses the 3 April 1974 “Super Outbreak” in for worst human toll in the modern era of 1950-present.
This utterly horrible development has taken place despite far more advanced weather prediction services that gave up to 5 days’ notice of at least some severe potential, leading to a tornado-based “HIGH Risk” outlook the night before, outstanding “Particularly Dangerous Situation” watches, and timely warnings. And yet somehow, over 300 folks are no longer with us who, when the morning of 27 April 2011 dawned, had no idea that the atmosphere had condemned them to their final hours.
The scale of human casualty (death and injury) has astounded almost everyone in the meteorological community. Let’s not mince words. This was carnage. The question is why?
We don’t know the answers. We won’t for some time. Rushing to certain conclusions is premature and ill-served. As I noted before, I suspect the bulk of the answers lie in the sociological and preparedness realms.
Until more thorough investigation is done through internal and external (to NWS) surveys, all we can deduce at this early stage of assessment is through secondhand (media) accounts, which are neither systematic, nor necessarily representative. In reading them, in gauging situations described by survivors and relatives of the victims in those accounts, ten themes do arise clearly so far, situations that are known to have appeared in prior killer-tornado events:
1. Some folks did what they were supposed to, and died anyway. Even in an EF4 or EF5 rated tornado, only a small percentage of the path will have top-end damage. The tornado, at their location, was in that terrible, unlucky minority of the most intense damage area–winds too strong even for their well-built and soundly anchored homes. The only survivable places in those rare but real scenarios are underground, or out of the path.
2. “Hit without warning” stories due to no inbound communication of warnings that did exist (power outages, no radio and TV for those without cell phones, lack of properly set, battery-operated weather radio).
3. Complacency: “Hit without warning” stories of the type that usually means, “I could have, but didn’t bother to, make myself weather-aware that day”. In those cases, yes, there absolutely was a warning, but also, lack of effort to tune in to the threat.
4. Sheltering in absolutely unsuitable structures such as mobile homes, clapboard houses over crawlspace foundations, agricultural edifices, and sheds, and/or physical inability to get to available sturdy shelters (handicapped, elderly).
5. Deaths in multi-tenant housing such as apartments where no communal shelter was available.
6. Vehicular destruction.
7. Non-trust in warnings received–desensitization via “cry-wolf” syndrome, i.e., “We get tornado warnings all the time and nothing happens.”
8. Faith in bogus tornado myths (“That mountain/river/Indian legend always protected us.”)
9. Temporal myopia/denial (“It’s never happened here before.”)
10. Fatalism (“If it’s your time, it’s your time.”)
Don’t shoot the messenger. I am summarizing what I have pulled from dozens of local and national accounts. It’s likely that ALL of these contributed. To what extent, and in what proportions? Too soon to say.
Matt Biddle ascertained many of the same factors in his study with David Legates for a violent tornado 13 years ago, in the same geographic area. What reason is there to believe that any of that has changed since? Anecdotally, none.
Still, we should not depend on, nor settle for, mere anecdotes here. Atmospheric and social sciences need to join forces and get to the bottom of this outrageous casualty toll in a tangible, systematic, and unambiguous way! I’m glad Matt is headed to the destruction zone to perform first-hand assessment, and hope others from the various atmospheric/social linkages here (i.e., WAS*IS affiliated folks) can get there soon too. I wish I could join Matt, but have shift responsibilities here to attend to.
I can say fairly safely that a major contributor here clearly was population density. Even though 3 April 1974 affected a few decent sized towns (Xenia, Brandenburg) and the suburbs of one big metro (Sayler Park, for Cincinnati), a greater coverage area of heavy developed land seemed to be in the way of this day’s tornadoes. How much? I’d like to see a land-use comparison. Harold Brooks and Chuck Doswell have discussed the phenomenon of sprawl as a factor in future tornado death risk and the possible nadir of fatalities having been reached…and the future seems to have arrived.
The amount of land area covered by sprawl is not decreasing in the South or Midwest. The nation’s population keeps growing too, cramming more people into these areas that have been built up since the last time a large, violent tornado roared through. We don’t know precisely to what extent sprawl has made the populace more vulnerable to tornado casualty; but it follows logically that when you put more and more people in a tornado’s path, risk goes up. Anyone can plot numerous tornado paths from the early-middle 1900s that would cause enormously greater damage in the same places today. This sprawl-tornado relationship is great fodder for scientifically rigorous examination!