Results

Starting at 23:19:48.535 UT, the electric field system was triggered in 3 consecutive time intervals. The broadband electric field data, sampled at 500 kHz, is shown in Figure 4.4. A -CG field change precedes a +CG field change with a large slow tail. The slow tail is followed by a large and isolated slow field change with a waveshape very similar to that produced by energetic nighttime sprites (Reising et al., 1999; Brook et al., 1997; Cummer and Stanley, 1999; Cummer et al., 1998). The +CG had a peak current of 70.9 kA and occurred in the MCS over southern Texas (see Figure 4.3) while the -CG was produced by an unrelated small storm in central Wyoming. The onset of the sprite field change occurred 13.2 ms after the +CG onset.

Figure 4.4: VLF radiation from a +CG return stroke followed by a large slow tail indicative of a large continuing current. More than 13 ms after the return stroke, a large slow field change occurred which is not immediately preceded by VLF radiation. The slow field change was likely produced by an energetic sprite event.

The charge moment change of the 23:19:48 UT parent +CG discharge and sprite event are shown in Figure 4.5. The charge moment change of the +CG at the onset of the sprite field change was 6100 C$\cdot$km. In contrast, the onset of nighttime sprite ELF signatures occurred at +CG charge moment changes well below 1100 C$\cdot$km (Cummer and Stanley, 1999). Charge moment changes exceeding 6000 C$\cdot$km may be exceptionally rare since Huang et al. (1999) found no charge moment changes larger than $\simeq$5000 C$\cdot$km across the entire globe during an $\simeq$18 day observation period in the summer of 1997.

Figure 4.5: The cumulative charge moment changes of the five +CGs which produced the largest such changes (within 20 ms of the return stroke) during the data acquisition period. The three most energetic discharges produced sprite events having charge moment changes of 910-2800 C$\cdot$km, as shown by the separation between the total charge moments (solid lines) and the estimated parent discharge components (dashed lines).

The charge transferred to ground by the parent +CG can be estimated by assuming that the average charge height was no lower than $\simeq$4 km, which is the base altitude of a significant positive charge layer in stratiform regions (Marshall et al., 1996), and no higher than the maximum cloud top height of $\simeq$15 km indicated by infrared satellite measurements of cloud top temperature (Figure 4.2). From the charge moment change, the total charge transferred to ground would therefore have been between 410 C and 1500 C prior to the onset of the sprite field change. The largest directly measured total charge transfer that the authors are aware of is $\simeq$450 C (also associated with a +CG) (Berger, 1972), which further attests to the highly energetic and unusual nature of this event.

The sprite event added $\simeq$2800 C$\cdot$km to the total charge moment change, as indicated in Figure 4.5 by the separation between the total charge moment change (solid line) and the estimated parent discharge charge moment change component (dashed line) for the 23:19:48 UT event. To put this into perspective, the largest sprite charge moment change found by Cummer and Stanley (1999) for 11 nighttime sprite events was $\sim$840 C$\cdot$km (see Section 5.3.3) while Reising et al. (1999) found no sprite charge moment changes larger than $\simeq$380 C$\cdot$km for 81 nighttime sprite events. Since nighttime sprite charge moment changes are known to be linearly correlated with spatially-integrated sprite brightness Reising et al. (1999), it is likely that this daytime sprite event was unusually bright.

Five minutes later at 23:24:55.658 UT, another +CG occurred within the same region of the storm as the first (Figure 4.3) and produced the second sprite field change. The onset of the sprite field change occurred after 11.0 ms had elapsed when the parent discharge charge moment change had reached 4300 C$\cdot$km (Figure 4.5). Although this was less than the threshold for the previous sprite ELF signature onset, it is still far in excess of that observed for nighttime sprites. The sprite event added $\simeq$1200 C$\cdot$km to the total charge moment change.

At 23:29:03.129 UT, the third consecutive +CG to strike within the same general region as the previous sprite-producing discharges produced the third (and final) sprite field change. Figure 4.5 shows that the charge moment change of the parent discharge for the onset of this sprite field change was 3900 C$\cdot$km after 12.3 ms had elapsed. The sprite event added $\simeq$910 C$\cdot$km to the total charge moment change.

There were other +CGs during the 35 min data acquisition period which also produced large slow tails but were not accompanied by isolated slow field changes. The two such discharges with the largest charge moment changes occurred at 23:10:40.964 UT and at 23:29:03.532 UT, and their charge moment changes are shown in Figure 4.5. The larger of these +CGs produced a charge moment change of 2200 C$\cdot$km in 20 ms, which was either insufficient to initiate a sprite or, more specifically, a sprite with a detectable ELF signature.