Total Lightning and Radar Characteristics of Severe Thunderstorms in the Mid-Atlantic Region

Numerous studies over the past decades have examined the relationship between thunderstorm lightning activity and associated severe weather (large hail, damaging winds and tornadoes), with somewhat ambiguous results. More recently total lightning data, i.e., cloud-to-ground (CG) and intra-cloud (IC), have become available, and the IC flash rate has been found to be several times larger than the CG flash rate. Researchers have also found that severe weather events are often preceded by a rapid increase (“jump”) in total lightning activity. Unfortunately, whereas CG lightning detection is well established over the entire CONUS, total lightning detection systems have been available only for limited areas, e.g., northern Alabama, north-central Texas, and the Washington, D.C. area. Most studies of total lightning have focused on the southern U.S. or the Great Plains.

Within the past couple of years however, a new total lightning network has been developed by Earth Networks ™ which has the potential to greatly increase the areal coverage for total lightning detection. While this network is still under development, the greatest density of sensors at this time is over the mid-Atlantic region, roughly from Washington, D.C. through Philadelphia to New York City. The estimated total lightning detection efficiency from the new network over this area is 80 to 90 percent. This new data set was made available to select National Weather Service (NWS) forecast offices beginning in the summer of 2011.

This study seeks to determine the potential utility of total lightning data for improvement of severe weather warnings in the NWS forecast office operational environment. Total lightning data for selected severe thunderstorms over New Jersey and vicinity during summer 2011 were obtained from a six-month Earth Networks ™ online archive. These data were examined in Google Earth ™ and sorted by type (CG vs. IC), time and location to subjectively eliminate lightning from nearby storms. High resolution time plots of CG, IC and total lightning for each storm were compared with official reports of severe weather from that storm. Also, NEXRAD Level II radar data were obtained from NCDC for the KDIX or KDOX radar, from which the height of selected echo intensity (dBZ) thresholds was determined, along with VIL, VIL density, and other parameters, for each radar volume scan.

Peak total lightning flash rates varied greatly storm to storm, from less than 50 per minute to over 500 per minute. However, of the 15 or so storms examined, about three-quarters exhibited a rapid increase in the total lightning flash rate (lightning jump) within 30 minutes prior to occurrence of severe weather. Comparison with radar-based parameters strongly suggests that total lightning can often give additional support and confidence for warning issuance decisions and in some cases may provide additional warning lead time. Given the very large number of IC lightning flashes compared to CGs (roughly 10 times greater), it will be important for operational warning applications to create a real-time contoured analysis of total lightning flash density, rather than simply plotting individual flashes.