Kamis, 12 Juli 2007

Tornado Prediction

New Problems in Tornado Prediction
Let's face it -- many people in tornado country aren't worried about energy flows within tornadoes. If they think much about twisters, it's mainly to wonder why they can't be predicted more accurately.

Poor predictions can cause two problems:

  • foolish behavior, when people ignore a warning, or
  • hours wasted in shelter, during false alarms.

Tornadoes, unfortunately, are tough to predict. Monster thunderstorms that should produce them don't. Smaller storms that shouldn't, do.

Unfortunately, the tornado-prediction picture seems to be getting worse. The results come from Project VORTEX, the world's largest storm-chasing project.

Storm chasing? I just want to crawl into a culvert...
For better or worse, the only way to get information on these monumental storms is to get up close and personal, says Howard Bluestein. Bluestein is a meteorologist with 19 years storm-chasing experience.

"They're fascinating," says Bluestein, "a violent display of nature that encompass a very small surface area. They're somewhat mysterious...and if you want to know about them, you have to be in the right place at the right time."

Being in the right place at the right time requires a network of vehicles, sensors and radios. It also requires having enough college students willing to work long hours on the slight chance they'll come face-to-face with a twister.

The work is not extremely dangerous, Bluestein says, if you discount hazards caused by winds, rain, hail and lightning. Storm chasers seldom get blown away by twisters. And the only recruitment problem, he says, is "keeping people away."

What the storm chasers found.
The Project VORTEX chasers are finding that some tornadoes form much more rapidly than previously thought. Researchers had thought that it took 20 to 30 minutes for a tornado to form. The new results, however, show that some tornadoes take only five to ten minutes to develop.

That causes problems for tornado researchers and forecasters, people like meteorologist Robert Davies-Jones. Davies-Jones is a scientist at the National Severe Storms Laboratory in Norman, Oklahoma.

The research results have obvious warning implications, he says, since you cannot issue a warning until a tornado has started forming.

More bad news.
A second finding is equally unhelpful to forecasters, he says. Tornadoes can form in small weather patterns that fall between weather stations: "You can get such a small-scale region around a thunderstorm that's favorable for tornado formation, but they may not be detected by the everyday network of detectors."

Further discoveries can be expected as researchers "slog through" data from Project VORTEX, Davies-Jones says, but one thing is already clear: creating the finer-grained, more detailed computer models needed to increase forecast accuracy will depend on "bigger computers and as much data as we can get."