Physical Mechanism of Initial Breakdown Pulses in Lightning Discharges

Abstract:

The initial breakdown stage of a lightning flash encompasses its first several to tens of milliseconds and it is characterized by a sequence of pulses typically detected with electric field change sensors on the ground [e.g., Villanueva et al., JGR, 99, D7, 1994]. A typical (referred to as "classical") initial breakdown pulse (IBP) has duration of tens of microseconds and it is one of the largest pulses at the beginning of a lightning flash, but a wide range of pulse durations and amplitudes also occur [e.g., Nag et al., Atmos. Res., 91, 316, 2009]. Recent results by Marshall et al. [JGR, 119, 445, 2014] suggest that IBPs should be observable in all lightning discharges. Complementarily, Stolzenburg et al. [JGR, 118, 2918, 2013] correlated individual IBPs to bursts of light that appear to be illumination of a lightning leader channel and Karunarathne et al. [JGR, 118, 7129, 2013] have determined that as a flash evolves the location of IBP sources inside the cloud coincide with the position of negative leaders as determined by a VHF lightning mapping system. In view of the above listed properties of IBPs, we have developed a new numerical model to investigate the electromagnetic signatures associated with these events and to relate them to the initial lightning leader development. The model is built on a bidirectional (zero-net-charge) lightning leader concept [e.g., Mazur and Ruhnke, JGR, 103, D18, 1998]. We simulate a finite-length finite-conductivity leader elongating in the thunderstorm electric field and we solve a set of integro-differential equations to retrieve the full dynamics of charges and currents induced in it. Our proposed approach is a generalization of the transmission-line [e.g., Nag and Rakov, JGR, 115, D20102, 2010] and electrostatic [e.g., Pasko, GRL, 41, 179, 2014] approximations used for analysis of in-cloud discharge processes. We also allow for different propagation mechanisms at the different polarity leader extremities, i.e., continuous (VHF-dark) elongation of positive tip and impulsive (through streamer corona flashes) elongation of negative leader tip [e.g., Williams and Heckman, J. Aerospace Lab, 5, 2012]. We demonstrate that a wide range of observed IBP morphologies can be effectively explained by the variation of only two parameters in the developing leader: step length and conductivity.