Tornado warnings: Too much of a good thing?

Unbeknownst to you, an F5 tornado -- the kind that can obliterate a house in a heartbeat -- is barreling directly toward yours: Would you prefer 60 seconds warning or 20 minutes?

Silly question, you say? Yet research by a pair of Texas economists suggests that your choice may not matter in survival terms (that you'd have 19 minutes more to worry with option two is indisputable). In fact, the study raises the question of whether that extra time might actually do more harm than good.

But they're economists.

More surprisingly, the chief technology officer for the country's most extensive severe weather warning network says he would "bet there is some truth to their study" -- up to a point, and with funnel cloud-sized caveats that we'll get to in a moment.

The discussion began last week with this headline -- Early Tornado Warnings Not Always Helpful -- that appeared on the Web site LiveScience.

The researchers analyzed data from more than 18,000 tornadoes in the United States between 1986 and 2002. Overall, they found that early warning is very helpful: On average it reduced expected injuries by about 32 percent.

But when the researchers examined data from the most severe cases - the 300 out of 18,000 tornadoes in which people died - the effects of advanced warning were less clear. Overall, when people were notified of a tornado up to about 15 minutes ahead of time, deaths decreased. However, lead times greater than 15 minutes seemed to increase fatalities compared with no warning.

What could possibly account for such a counter-intuitive outcome? The researchers acknowledge that their data is insufficient for firm conclusions, but here's what they suspect:

"The concern is that longer lead times would encourage dangerous behavior," said Kevin Simmons, an economist at Austin College in Texas. "There is anecdotal evidence that came out of the tornadoes in Oklahoma and Missouri last week. Out of the 23 fatalities, eight were people in cars. I don't know if those people were trying to outrun the storm, or if they just happened to be in their cars."

Being from Massachusetts, where tornadoes are rare, I can claim little first-hand knowledge of such situations (aside from that vacation scare in Minnesota where the family watched on TV as a tornado skittered by about a half-town away from our rental.)

But I know a guy who knows: Chris Sloop, CTO at Weatherbug, which operates a network of more than 8,000 tracking stations and 1,000 cameras in schools, public safety buildings and TV stations. I made Sloop's acquaintance in 2005 when he accused me of callous disregard for the health and well being of my fellow man after I wrote what struck me then as a rather sympathetic column about his company being at Interop. Since that dustup, we've become e-mail buddies.

Here's some of what Sloop had to say about the economists' research:

I bet there is some truth to their study. When people don't know what to do in a dangerous situation, many times they do the wrong thing.

I am surprised the authors didn't take that approach. It is almost like they are saying that advance warning is a bad thing, when in reality it is a GREAT thing; it's just that people are not well educated enough to know how to respond. Of course, in cases like the most severe tornados, there may not be much you can even do that will save your life.

Another factor to consider here is that technology has changed dramatically since the era of the data sample used by the economists: 1986 to 2002.

People have a cell phone with them now wherever they go. The alerts that people get now are very localized and direct. So, there is much less worry about the crying wolf syndrome where people just ignore warnings ... and with our system, people can be alerted to just knowing that severe weather conditions are heading their way. ... The key though, is education and preparation. You need to know exactly what you are going to do when faced with a dangerous situation.

Work and recreational habits have changed as well, he notes.

The final thing I would like to point out is that all of these articles always talk about NOAA Weather Radio and television/radio like they are the only ways to get weather warnings. Who watches TV any more? People are on computers more than TV; I know I am and that is why products such as WeatherBug Desktop or WeatherBug Alert are very important. Not to mention all the cell phone/SMS capabilities.

Coincidentally, as I started to write this post, I received an alert about potentially damaging thunderstorms with marble-sized hail moving through my area. No, it wasn't from TV, radio or Weatherbug, rather it came via old-fashioned e-mail from a friend who for years has provided the service as a hobby. It works ... but like I said earlier, we don't get the killer storms in these parts very often.

In the event our luck does run out, please put me down for as much warning as possible.

Welcome regulars and passersby. Here are a few more recent Buzzblog items. And, if you'd like to receive Buzzblog via e-mail newsletter, here's where to sign up.

Earth to Vint Cerf: What's up with the lousy video?

This Year's 25 Geekiest 25th Anniversaries.

Firefox looks to set an official world record.

The REAL sticking point between Microsoft and Yahoo!

Google News, McCain and The Mystery Blonde.

Worst of the lot for two years running: PCMall and PCConnection.

Times breaks out xkcd-to-English translator.

YouTube's down, everybody panic.

Top 10 Buzzblog posts for '07: Verizon's there, of course, along with Gates, Wikipedia and the guy who lost a girlfriend to Blackberry's blackout.

8 can't-miss tech predictions ... for 1998