The image on the left is a lightning threat forecast valid up to the time at the top of the image. The colors correspond to the lightning threat for the forecast period. Red indicates a high threat of lightning, yellow indicates a moderate threat, and green indicates a low threat. The image on the right is the "verification" of the forecast, or the lightning observed during the forecast period. The colors in this image represent the accumulation of strikes per grid box within the previous 20 minutes.

The Lightning Threat Algorithm is designed to predict areas of potential lightning up to 30 minutes in advance.

NSSL Lightning Threat Algorithm

by Amy Wyatt

Scientists at NSSL are involved in research on all types of severe weather: thunderstorms, flash floods, winter storms, tornadoes and lightning. Their work constitutes both basic research into the processes that produce severe weather, and development of algorithms and display systems to predict and monitor severe weather. With these systems, NSSL is providing improved warning capabilities to such entities as the NWS and the FAA. NSSL recently broke new ground by creating a specialized algorithm for a customer in the private sector. The Lightning Threat Algorithm was developed as part of a Cooperative Research and Development Agreement with WeatherData, Inc., a commercial meteorological firm whose clients include railroads, utility companies, trucking companies and the media. WeatherDataÕs clients are sensitive to all types of severe weather, but especially to lightning, a threat WeatherData previously had not been able to address specifically. Following brainstorming sessions with NSSL lightning specialists Ron Holle, Ken Howard, Don MacGorman and Raul Lopez, the Lightning Threat Algorithm was developed by Kurt Hondl, Kim Elmore, Amy Wyatt, V. Lakshmanan, Venkat Ganti, and Kevin Thomas. The Lightning Threat Algorithm is designed to predict areas of potential lightning up to 30 minutes in advance. Lightning threat is classified as high, moderate or low. Areas where the potential for lightning exists within the next forecast interval are also identified. The algorithm functions by first determining storm motion utilizing a technique developed by Lincoln Laboratory, producing forecast images of the radar and lightning products, and finally combining the forecast images to determine the lightning threat. Testing at NSSL showed that the Lightning Threat Algorithm has a high degree of skill. For 30 minute forecasts, over 90 percent of the lightning strikes were observed within 4 miles of where the algorithm predicted they might occur. For 10- and 20 minute forecasts, the observed correspondence was as high as 97 and 95 percent respectively.

For more information contact Mike Eilts at: eilts@nssl.noaa.gov