Experimental evaluation of an arrival time difference lightning positioning system

Results from a five-station LPATS lightning positioning system covering the Netherlands are analysed and compared with observer reports and with registrations from lightning counters.
It is demonstrated that the median location accuracy of individual flashes is about 2 km. The reliability of detection is diminished by the occurrence of false alarms and the possibility of system saturation during widespread thunderstorms. Contrary to expectations the system may respond to intracloud discharges.
Despite these difficulties the system seems quite acceptable to replace the even more imperfect human observations for warning purposes and for thunderstorm statistics. In addition the system offers possibilities to collect statistics on lightning parameters like peak currents and lightning discharge density.
A preliminary statistical analysis shows that the establishment of preferred lightning subregions in the Netherlands would require at least ten years of measurements, while small-scale so-called lightning spots will remain masked by the limited location accuracy.