Search for a command to run...
The paper constitutes a model of the final jump that occurs during the encounter between a negative downward stepped leader and an upward connecting leader in cloud to ground lightning. The gap is first covered by negative and positive streamers, followed by leader bridging. Voltage collapse yields substantial impulse current determined by both the gap and stepped leader characteristics. A model is formulated for the leader bridging and subsequent current impulse flow, during which the negative downward leader is simulated as a lossy transmission line. Within the range of variables investigated the following findings are reported. With neutral charge point in the cloud at a height of 4500 m, ambient ground field of 8 kV/m, a 10 m ground mast and prospective return stroke current in the range of 10-100 kA, the model resulted in final jump gaps in the range of 7-27 m. The average leader bridging speed is found to vary in the range of 5-16 m/μs. The final jump impulse current is evaluated in the 2-12 kA range corresponding to 10-20% of the subsequent prospective return stroke current. The final jump impulse current is therefore substantially larger than the preceding corona bursts. The input surge impedance of the downward leader at the instant of final jump varies in the range 1.6 kohm to 2.8 kohm. Results with a height of charge cloud point origin Hcs of 2500 m have been added and showed very significant increase in the final jump gap length and impulse current amplitude but a moderate reduction in the leader input surge impedance. The discharge phase of the final jump is therefore established as a distinct prerequisite for the formation of the return stroke. In principle the final jump impulse current can be measured on instrumented towers, discarding prior upward leader current and corona pulses from adjacent branches. It can also be estimated from remote electric field measurement, with due consideration to the specific speed of the resulting signal propagating along the stepped leader. In this paper whenever possible comparison of the model results with field measurement is made and found satisfactory.