Analysis of charge transport during lightning using balloon‐borne electric field sensors and Lightning Mapping Array

Abstract

Recently, wide band measurements of the electric field near a lightning flash have been obtained by a balloon‐borne electric field sonde or Esonde. This paper develops new techniques for analyzing lightning‐associated charge transport in a thundercloud by using both the Esonde data and simultaneous Lightning Mapping Array (LMA) measurements of VHF pulses emitted during lightning breakdown processes. Innovations in this paper include the following: (1) A filtering procedure is developed to separate the background field associated with instrument rotation and cloud charging processes from the lightning‐induced electric field change. Because of the abrupt change in the signal caused by lightning, standard filtering techniques do not apply. A new mathematical procedure is developed to estimate the background electric field that would have existed if the lightning had not occurred. The estimated background field is subtracted from the measured field to obtain the lightning‐induced field change. (2) Techniques are developed to estimate the charge transport due to lightning. At any instant of time during a cloud‐to‐ground (CG) flash, we estimate the charge transport by a monopole. During an intracloud (IC) flash, we estimate the charge transport by a dipole. Since the location of the monopole and dipole changes with time, they are referred to as a dynamic monopole and a dynamic dipole. The following physical constraints are used to achieve a unique fit: charge conservation during an IC flash, separation (distance between the CG monopole charge center and the ground and separation between IC dipole charge centers both exceed a minimum threshold), location (charge is placed on lightning channel), and likelihood (after a statistical analysis based on instrument uncertainty, highly unlikely charge locations are excluded). To implement the constraint that the charge is located on the lightning channel, we develop a mathematical object called the “pulse graph.” Vertices in the graph are pulse locations obtained from the Lightning Mapping Array. Edges in the graph (that is, the pairs of vertices which are connected by line segments) are obtained by joining, in a systematic way, neighboring vertices. One CG and two IC flashes observed on 18 August 2004 near Langmuir Laboratory are analyzed. In the CG flash, initial strokes drained 12 C charge from an altitude of 5 km, while an intermediate stroke discharged 12 C from a higher charge center at 8 km. For the IC flashes, the current flow lagged behind the channel formation by time intervals on the order of 0.1 s, roughly the same time delay observed for lightning optical signals detected by NASA's Lightning Imaging Sensor.