AE31B-0435
Performance Study of Earth Networks Total Lightning Network using Rocket-Triggered Lightning Data in 2014

Wednesday, 16 December 2015
Poster Hall (Moscone South)
Stan Heckman, Organization Not Listed, Washington, DC, United States
Abstract:
Modern lightning locating systems (LLS) provide real-time monitoring and early warning of lightning
activities. In addition, LLS provide valuable data for statistical analysis in lightning research. It is
important to know the performance of such LLS. In the present study, the performance of the Earth
Networks Total Lightning Network (ENTLN) is studied using rocket-triggered lightning data acquired at
the International Center for Lightning Research and Testing (ICLRT), Camp Blanding, Florida.
In the present study, 18 flashes triggered at ICLRT in 2014 were analyzed and they comprise of 78
negative cloud-to-ground return strokes. The geometric mean, median, minimum, and maximum for the
peak currents of the 78 return strokes are 13.4 kA, 13.6 kA, 3.7 kA, and 38.4 kA, respectively. The peak
currents represent typical subsequent return strokes in natural cloud-to-ground lightning.
Earth Networks has developed a new data processor to improve the performance of their network. In
this study, results are presented for the ENTLN data using the old processor (originally reported in 2014)
and the ENTLN data simulated using the new processor. The flash detection efficiency, stroke detection
efficiency, percentage of misclassification, median location error, median peak current estimation error,
and median absolute peak current estimation error for the originally reported data from old processor
are 100%, 94%, 49%, 271 m, 5%, and 13%, respectively, and those for the simulated data using the new
processor are 100%, 99%, 9%, 280 m, 11%, and 15%, respectively. The use of new processor resulted in
higher stroke detection efficiency and lower percentage of misclassification. It is worth noting that the
slight differences in median location error, median peak current estimation error, and median absolute
peak current estimation error for the two processors are due to the fact that the new processor
detected more number of return strokes than the old processor.