People in the path of the tornado that tore through Moore, Okla., yesterday had about 16 minutes warning to find shelter. A lot of money is going into research to improve that lead time and predict where tornadoes might touch down.
Scientists have been trying to figure out how to forecast tornadoes in the U.S. since the late 1800s.
"You might think by now we would actually understand how tornadoes form and why tornadoes form, but that's still a very open area of research," explains Glen Romine, who studies tornadoes at the federally funded National Center for Atmospheric Research. "We're still actively trying to go out and look at what's going on in those storms that do and don't produce a tornado and figure out what's different about them."
Romine is working with researchers who are flying over the Great Plains this spring to sample the atmosphere where storms might form. Their goal: better forecasts up to a day in advance.
Other researchers are trying to give people more lead-time to get out of a tornado's path. They know where to start -- building thousands more Doppler RADAR stations.
Scientist David McLaughlin, with the Collaborative Adapting Sensing of the Atmosphere (CASA), says existing stations are too far apart, and forecasters can't get information fast enough. "You can't really tell which streets and individual neighborhoods those vortices are going down, unless it happens to be a well-behaved, mile-wide storm that everybody can see."
CASA is developing a new network of shorter-range RADARs that would give more accurate and specific information. "There is this need for higher resolution, better observations close to ground, now the question is who has a couple billion dollars to deploy this across the entire country," McLaughlin says.
A grant from the National Science Foundation paid for testing of the system in Oklahoma. Now the prototype is set up in the Dallas-Fort Worth, where cities are helping to pay for it.