Sunday, May 1, 2011

Ignorance is not bliss

The 27 April 2011 outbreak of tornadoes in the southeastern US clearly has achieved historic proportions. Its death toll has exceeded the so-call Super Outbreak of 03 April 1974. It remains to be seen what the final count of tornadoes and violent (EF4-5) tornadoes will be. It's possible that the "Super Outbreak" has been exceeded. Whatever the final tally turns out to show, it seems obvious that the 27 April 2011 event has achieved its place amidst the most infamous outbreaks in US history.

The media seem to be focused on the "freak" nature of this event, as usual. For those of us who have spent years studying the record of US tornado events, this outbreak is simply a reflection of what's possible as seen in the historical record and, in fact, what's most likely. Some of us have predicted this vulnerability. Events of the sort like 27 April 2011 occur roughly every 20 years or so -- the frequency is not at regular intervals, of course, and each event is different. But the long history of tornadoes in the US makes it clear that large fatality counts result from the infrequent coincidence of violent, long-track tornadoes with populated areas. Although outbreaks of long-track violent tornadoes in the southeastern US are relatively infrequent, the southeastern US is particularly vulnerable: a large fraction of mobile home owners, a relatively high density of communities at risk, a growing population at risk, a relatively low fraction of homes with basements, and so on -- all resulting in a high vulnerability whenever long-track, violent tornadoes occur. Thus, the tragic events of 27 April 2011 were quite predictable, although we couldn't know exactly when or where this would happen.

By studying the historical record, it's possible to anticipate events such as 27 April 2011, although we can't know precisely when they will occur. Despite advances in technology and in the infrastructure devoted to tornado warnings, when violent tornadoes interact with population centers, people will die. The infrastructure dedicated to saving lives in situations involving tornadoes has resulted in roughly a factor of 10 decrease since 1950 in average annual tornado fatalities. But the increase of population at risk has begun to offset this, in part associated with an increasing fraction of people living in mobile homes (where their risk is substantially higher than those living in "stick-built" homes). We've reached the point where the majority of tornado fatalities are those living in mobile homes, and there's a higher percentage of such in the southeastern US than elsewhere.

Thus, as I suggested in the earlier essay, the southeastern US was the most likely location for an outbreak resulting in a large number of fatalies. At times, being right is not a very satisfying situation. But it does underscore the reality that the US remains vulnerable to tornadoes, despite the new technologies and the new warning infrastructure. Even more devastating events are possible in the future. We aren't invulnerable to tornadoes and not likely to be anytime soon.

No comments: