Tornado-Chasing Project Aims to Improve Forecasts

When a tornado is about to cut a devastating swath through an American town, those in its path get a warning lead time of 13 minutes on average to try to reach shelter.

"If you live in a trailer community, is 13 minutes enough to wake your family and get them bundled up and outside?" asks tornado researcher Joshua Wurman, head of the nonprofit Center for Severe Weather Research. And "if you are elderly or handicapped, you're going to have a hard time getting to a shelter in 13 minutes," he said.

And that's the average; many times people are warned about six or seven minutes earlier. That is because although scientists know that certain kinds of "supercell" thunderclouds can spin off tornadoes, they know very little about the exact conditions that indicate a tornado will occur and whether it will be a mild twister or a violent killer.

In the mid-1990s, a two-year study called Vortex had a phalanx of scientists chasing tornadoes around the Great Plains, inspiring much public fascination, daredevil amateur tornado-chasers and the 1996 movie "Twister."

Vortex (which stands for Verification of the Origins of Rotation in Tornadoes Experiment) resulted in significant advances, including the revelation that tornadoes can occur on smaller time and space scales than previously thought and that sometimes they do not show up on radar. Knowledge gained from the study led to an increase in average warning times, but it did not unlock the secrets of exactly when and why tornadoes form.

As a result, predictions about tornado occurrence are successful only about a quarter of the time.

"Sometimes people will choose not to take shelter even if they're told to," said Yvette P. Richardson, a meteorology professor at Pennsylvania State University. "In general, the more we can reduce the false-alarm rate, the more seriously the public will take warnings."

Now comes Vortex2, a five-week tornado-chasing project beginning next month that scientists hope will finally provide the knowledge to accurately predict when and where a tornado will develop.

"Ultimately we'd like to get to the point where we can put sufficient data into our models so we know when a tornado will happen," said Stephan P. Nelson, a program director in the atmospheric sciences division of the National Science Foundation, which, along with the National Oceanic and Atmospheric Administration, provided the $12 million funding for Vortex2. "Then you can get first responders to be better prepared -- police, fire, medical personnel, even power companies. Now, that's not even remotely possible."

As part of Vortex2, about 80 veteran scientists and graduate students will chase storms across a wide swath from South Dakota to Texas and from eastern Colorado to Iowa and Minnesota, with their nerve center in Norman, Okla.

They will be armed with a host of tools, including lasers that measure raindrops, Doppler radar mounted on trucks, high-tech balloons, unmanned aircraft and instruments on tripods anchored in the tornadoes' path.

"We're throwing everything but the kitchen sink at it," said Wurman, who has chased 141 tornadoes over 14 years. "We'll have a whole potpourri of instruments surrounding the storm, all measuring different things in different ways."