The characteristics of multiple stroke negative cloud to ground (CG) lightning flashes in Padang, Indonesia based on electric field recording were investigated. 100 negative CG lightning flashes containing 623 strokes, recorded from August to October 2014, were examined. It was observed that all first return strokes (RS) negative CG lightning flashes were preceded by preliminary breakdown pulses (PBPs). However, only 40% of the detecable PBP trains could be analyzed statistically on PBP train duration, PBP-RS separation, and PBP/RS ratio, while the detectable leader pulses was 69%. In addition, it was also found that the arithmetic mean (AM) and maximum number of strokes were 5.2 and 18, respectively. The AM and geometric mean (GM) interstroke interval were 55.34 and 34.71 ms, respectively. Furthermore, 2% of the subsequent return stroke (SRS) peak electric fields were larger than those of the first return stroke (RS). The AM and GM ratios of SRS/RS were 0.36 and 0.3, respectively.Keywords: negative cloud to ground, preliminary breakdown, leader, return stroke, interstroke interval, peak electric field ratio
IntroductionNegative cloud to ground (CG) lightning flashes is the most common type of CG discharges and has thereafter studied by many authors. A CG lightning flash usually contains preliminary breakdown pulses (PBPs) in the cloud and downward stepped leader (SL) propagation followed by the first return stroke (RS). The characteristics of a negative CG could largely be determined by the magnitude of the so called lower positive charge regions (LPCR) [1][2][3][4]. A larger LPCR causes the direction of negative leader propagation to be predominantly horizontal, even though finally the stepped leader propagates toward the ground. Meanwhile, a smaller LPCR causes negative stepped leader propagation to be predominantly vertical [5][6]. The time separation between PBPs and RS is approximately several milliseconds to hundreds of milliseconds. The generated electric field signature differentiates these processes. Nag and Rakov [5] stated that in negative CG lightning flashes, PBPs may not be detected on the first return stroke when the magnitude of the LPCR is very small. They claimed that CG lightning flashes vary in detectable PBP percentage according to latitude. Higher latitudes have a higher percentage of detectable PBPs than lower latitudes. In contrast, many researchers have argued that the first return stroke always involves PBPs in all negative CG flashes [7][8][9]. Their studies revealed that the detectable PBP percentage does not depend on latitude. A PBP can be detected if the amplitude of the PBP is above the noise level of the environment and electric field change sensors. However, Marshall et al. [9] have stated that they "cannot be certain that detection of PBPs always equals 100%" for all latitudes. In a more recent study, Zhu et al. [10] found that PBP detectability is not only affected by the sensor noise level but also by storm type, RS peak current and observation distance. The diffe...