When more than 70 tornadoes raged across central Oklahoma May 3, the National Weather Service credited advances in technology for helping to warn people well in advance and for saving lives, according to agency officials testifying at a congressional hearing this month. The lead time of tornado war
When more than 70 tornadoes raged across central Oklahoma May 3, the National Weather Service credited advances in technology for helping to warn people well in advance and for saving lives, according to agency officials testifying at a congressional hearing this month.
The lead time of tornado warnings has increased over the past 10 years from an average of 4 minutes to more than 11 minutes. In addition, the accuracy of the warnings - the percent of tornadoes that actually occur following a warning - has improved from about 30 percent to around 65 percent, said Dennis McCarthy, the meteorologist in charge of the Norman Weather Service's Forecast Office.
Technology, including Next-Generation Weather Radar (Nexrad) and the Advanced Weather Interactive Processing System (AWIPS), helped the Norman, Okla., office warn people about 18 minutes ahead of the tornado outbreak that hit the central part of the state on May 3, McCarthy said. Although 42 people died during that outbreak, the number of fatalities could have been much higher, he said.
"Even though a large and extremely violent tornado moved through densely populated sections of the southern Oklahoma City metropolitan area, the excellent warnings provided by the National Weather Service can be credited with the extremely low loss of life," he said this month before the House Science Committee's Energy and Environment Subcommittee and Basic Research Subcommittee.
According to estimates, the advanced technologies and preparedness programs saved more than 600 lives in Oklahoma that day, said Rep. Nick Smith (R-Mich.), chairman of the Basic Research Subcommittee.
Several tools played an important role in forecasting the tornadoes, McCarthy said. For example, Nexrad shows the motion of air in thunderstorms, which can be precursors to tornadoes. It enables NWS to predict tornadoes and where they are headed, McCarthy said. "In the past all we could do was report tornadoes and guess where they were going," he said.
AWIPS, which integrates data from satellites, Nexrad, and other observations in one workstation, is just as important for forecasting tornadoes, McCarthy said. Forecasters typically had to look at many data sources to develop forecasts and warnings. "I don't know what we would have done without AWIPS," he said. NWS expects to finish deploying AWIPS this month.
In addition, the Norman office is testing a new technology called the Warning Decision Support System (WDSS), which provides access to more Nexrad data than is available on existing systems. It shows more detail and ranks a storm based on its intensity and its severity. "During the May 3 event, changeable menus and tables in WDSS, along with radar displays in AWIPS, allowed us to maintain an overview of the outbreak," McCarthy said.
It is expected that future tornado lead times will improve to as much as three to six hours in some cases, said Morris Weisman, a scientist at the National Center for Atmospheric Research. "How far can we extend such lead times? The research models that can answer this very question are just coming online," he testified at the hearing.
For example, the experimental Advanced Regional Prediction System model helped identify the Oklahoma City region and much of north-central Oklahoma as potential targets for developing storms on May 3. Conventional forecasts showed only a broad region of precipitation mainly over northern Texas.
Using the ARPS model required a computer running at 12 Gflops of sustained performance. However, nationwide coverage will require more than 100 times the computer power currently available, Weisman said.
"Along with increases in computer power, expanded observing systems as well as further basic research will be necessary to take the next step forward," he said.