Abstract
Sferic pulse-train amplitudes and arrival times were used to locate lightning flashes and correlate cloud-to-ground (C–G) lightning with radar reflectivity in two typical Oklahoma storm systems. One system was a squall line lasting about 6 hr and producing detailed azimuthal displays of at least six individual storms with C–G lightning. The second system was a group of weaker air mass thunderstorms lasting about 3 hr. Although lightning counts per unit time varied widely between storms and within the same storm, the correlation of lightning counts with the radial extent of radar reflectivity factors ≥550 mm6 m−3 was fair. A rough correlation existed between C–G lightning counts and the amount of rainfall. The overall average rate of one cloud-to-ground flash per 1.6 × 1010 gm of rainfall found in a limited number of storms compares with Battan's report of one flash per 3 × 1010 gm. Even though there was some uncertainty about the rainfall measurements, the Oklahoma results support the opinion that in a developing thunderstorm the number of C–G flashes increases more rapidly than the rainfall.
