Surviving tornadoes mostly depends on a lot of luck and the right attitude

The pathway to Oz has been open a lot this spring. Tornadoes have pummeled much of the U.S. Southeast and Midwest, with many more victims than the Wicked Witch of the East. 2011 is already one of the deadliest tornado seasons in U.S. history.

Even meteorologists are taken aback. Nearly 1,500 tornadoes have killed at least 536 people, notably in Alabama in April and Joplin, Mo., in May. The last year this many Americans died from tornadoes, Franklin Roosevelt was president.

Many have tried to link this outbreak to climate change, a connection that may exist but simply can’t be drawn yet given the limited historical record for tornadoes.

Disaster rubberneckers would do better to step back and put the 2011 toll in greater context. For one thing, tornado death rates have dropped dramatically in the past century, thanks to radar and other warning technologies. Just installing a nationwide Doppler radar system in 1988 meant that tornado deaths dropped by 45 percent from what would be expected without it.

Still, big tornado outbreaks do happen every 20 to 40 years or so. And sometimes they happen where lots of people live in lots of poorly constructed homes. “When these big fatality counts occur, it’s mostly a matter of bad luck,” writes Chuck Doswell, a longtime research meteorologist now at the University of Oklahoma in Norman.

So what can people do to survive tornadoes, other than watching more Helen Hunt movies? In part, they can understand their own minds.

As far back as 1972, scientists tried to link tornado death rates in Illinois and Alabama to the difference between regional outlooks on life. John Sims, a psychiatrist at the University of Chicago, and Duane Baumann, a geographer at Southern Illinois University in Carbondale, found that southerners place more weight than northerners on the role of an external agent (such as God) in their lives. Related feelings of fatalism and passivity might “constitute a weak defense against the terrible strike of the tornado,” Sims and Baumann wrote in Science.

Such broad-brush attitudes clearly play only a small role, if any, in surviving a tornado, as the scientists were careful to point out. But more recent research has also found ways in which people’s thinking about twisters affects responses to the danger.

Sheer exposure may be one factor. Bob Drost, a graduate student at Michigan State University, has explored how personal experience with tornadoes affects people’s reactions to severe weather warnings. He found 23 volunteers who had experienced a tornado and 26 who had only read about twisters. Before and after watching a five-minute slide show on tornadoes, the participants filled out a questionnaire about how they would react to a tornado warning (listen to NOAA weather radio, seek shelter, wait for sirens, and so on).

Surprisingly, he found that the students who had personally experienced tornadoes were more blasé about reacting to a warning than those who hadn’t. Yet after watching the slide show —and perhaps being scared into remembering how bad things could get — both groups became more careful in their decision making. Drost reported the findings last month in Oklahoma City at a conference on broadcast meteorology.

Of course, paying attention to tornado warnings will get you only so far. In a review of storms between 1986 and 2007, scientists found that nearly one-quarter of all tornado fatalities came during storms for which no warning had been  issued, according to a paper in the January Weather, Climate, and Society. And warnings seem to work best when issued about 10 or 15 minutes before the actual twister; any longer than that, and people apparently conclude that the warning is a false alarm.

So longer warning times might actually work against saving lives. A survey of 320 visitors to the National Weather Center in Norman found that if given a one-hour warning before a tornado was expected to strike, people said taking shelter was less of a priority, and they felt less threatened, than if given a 15-minute warning. Stephanie Hoekstra of the University of Oklahoma and colleagues describe those results in a paper to appear in Weather, Climate, and Society. The National Weather Service has considered issuing warnings an hour or two ahead of tornadoes, based on computer forecasts rather than actual sighting of a tornado — but Hoekstra’s work suggests that such well-meaning goals may backfire.

In the end, the biggest barrier to tornado survival may be people themselves.

SN Prime | July 1, 2011 | Vol. 1, No. 4