Storm development history

Figures 3.1-3.2 show the formation, maturation, and dissipation of a mesoscale convective system (MCS) over a time span of 4 hours between 23 UT on June 21, and 3 UT on June 22, 1997 (7-11 pm Eastern Daylight Time on June 21). The data points are 3D lightning VHF source locations determined by LDAR. These points are rainbow-colored with respect to time such that blue corresponds to VHF sources near the beginning, yellow is near the middle, and red is near the end of each 20 minute time interval. The $\triangle$ and $+$ symbols correspond to -CG and +CG stroke locations, respectively, as determined by the National Lightning Detection Network (NLDN) (Cummins et al., 1998).

In Figure 3.1a, a couple of isolated thunderstorms produced lightning throughout the 20 minute time interval. Towards the end of the time interval, a couple of new storms each produced their first intracloud discharge, as denoted by the clusters of red data points between and to the south of the active storms. Note the significant radial errors manifested by the spreading of the points along a radial line to the LDAR stations ($\Box$ symbols).

In Figure 3.1b, additional storm cells became electrified and these storms began to organize into a line. In the next time interval, the cell on the northern edge became inactive while the CG flash rate from the other cells intensified. In Figure 3.1d, new electrically active cells were evident on the southern edge. The MCS flash rate appeared to have reached a maximum between 00-01:00 UT (Figure 3.1d-f).

Figure 3.1: The formation of a mesoscale convective system (MCS) is shown in 20 minute intervals between 23 UT and 1 UT on June 21st and June 22nd, 1997. The Lightning Detection and Ranging (LDAR) data points are rainbow-colored from blue to red with respect to elapsed time within each interval. -CG stroke ($\triangle$) and +CG stroke (+) locations were obtained from the National Lightning Detection Network (NLDN). The observation site (*) corresponds to $x\!=\!0$, $y\!=\!0$. The LDAR station locations ($\Box$) are also shown.

Figure 3.2: A continuation of Figure 3.1 for 1-3 UT on June 22nd, 1997. The MCS continued its southward advancement as new storm cells formed on the southern flank while older cells decayed on the northern flank. Sprites were associated with some of the +CGs produced by horizontally extensive discharges in panels c-e.

In Figure 3.2a, some discharges began to propagate into the region ``behind'' the northern half of the convective line. The horizontal discharges became more extensive with time, as evident in panel b). This pattern of increasing horizontal discharge dimensions was always readily evident 20 minutes or more prior to sprites, and turned out to be a good diagnostic for predicting where sprites might appear. In contrast, the CG strike locations indicated by NLDN did not provide any predictive lead time as to when and where sprites might occur. However, it should be noted that increasing horizontal discharge extent was not always followed by sprites, particularly if the discharges were somewhat 1-dimensional with a limited total area.

A large number of NLDN-indicated +CGs were associated with the active cell on the southern end of the storm line in Figures 3.2a-b. Almost all of these ``+CGs'' had NLDN signal strengths below 100 LLP units ($<18.5$ kA peak current). An earlier unpublished study (by the author) of weak ``+CG'' sferic records indicated that such events were probably ICs, though this has not yet been quantified. A few NLDN-indicated +CGs were evident in the region of horizontally extensive discharges. All of these ``+CGs'' were also weak at under 150 LLP strength and may also have been ICs. The northermost ``+CG'' is known to be a false detection which followed less than 0.2 ms after a real -CG from the northernmost active convective cell.

Sprites were initiated by many of the +CGs associated with horizontally extensive discharges in Figure 3.2c. The sprite events were coincident on video with the +CGs at 01:42:57.172, 01:47:52.711, 01:51:02.762, and 01:55:47.956 UT. These sprite events and their parent discharges are examined in detail later in this chapter. Additional sprite events occurred at 02:09:16.9, 02:09:17.1, and 02:27:21.3 UT in Figures 3.2d-e. Unfortunately, the interferometer system failed to trigger for these events.

The storm system continued to advance southward as new cells formed at the southern edge. However, it is clear that the convective line lost its cohesion in Figure 3.2. The storm complex eventually subsided within a few hours and no sprites were observed after the 02:27 UT event.