Computer system predicts dangerous clouds
Scientists say they have developed technology for helping pilots avoid those pesky pop-up thunderstorms. They’ll throw in turbulence prediction at no extra charge.
The computer system developed by a team of researchers at the University of Alabama in Huntsville uses data from the National Oceanic and Atmospheric Administration’s GOES weather satellites to provide 15-minute to one-hour warnings of thunderstorms by tracking changes in cloud temperature and water vapor. Researchers say this is the first time forecasters have had a tool to predict storms locally. Doppler radar, for instance, only tracks rain after it starts to fall.
“The radar tells you what’s happening, but not what’s going to happen,” said researcher Wayne MacKenzie.
The team has determined that one of the most important factors in predicting thunderstorms is temperature change. If the top of a cloud cools by 4 degrees C (7.2 degrees F) or more in 15 minutes, it means the cloud is growing quickly and that there is an increasing probability of rain beginning within 30 minutes to an hour.
John Mecikalski, an assistant professor of atmospheric science at the university, got the idea for the system in 2001. He was looking for a way to determine which of the thousands of cumulus clouds on any given summer day would become convective. Called satellite convection analysis and tracking system (SATCASTS), it has been accurate in its storm forecasts between 65 and 75 percent of the time.
SATCASTS has been in place for three years and is operated by university scientists for the National Weather Service’s forecast office in Huntsville. Later this summer a version of the weather program will begin forecasting storms in Central America, southern New Mexico, and the Dominican Republic. The FAA’s air traffic control center in New York City is testing the system. If successful, it could be rolled out worldwide.
Researchers say the system is relatively inexpensive and easy to install because it uses freely distributed weather data from existing satellite sensors.
July 9, 2008