Lightning, supercells and sprites

During the decade-long series of summer sprite observation campaigns at the Yucca Ridge Field Station (YRFS), events above nocturnal High Plains convection were routinely detected above a variety of convective regimes generating positive polarity cloud-to-ground (+CG) flashes. However, sprites were rarely observed above nocturnal supercells, even those producing substantial numbers of +CGs. The one exception would be from some +CGs occurring during the dying phases of the storm, when an expanding stratiform precipitation region of the order of 10,000 km2 would evolve. The summer 2000 Severe Thunderstorm Electrification and Precipitation Study (STEPS) monitored several supercells utilizing low-light cameras, the 3-D Lightning Mapping Array (LMA), the NLDN, radar, GOES satellite and remote ELF/VLF transient signatures. A key question addressed during STEPS was why do some +CGs produce sprites, while most do not? A companion issue was why supercell +CGs were even less prone to generate sprites? Research has strongly suggested that NLDN peak currents are not a robust predictor of sprite potential. Rather, the key metric is the charge moment change (Mq) of the stroke. Sprite probabilities rise rapidly to near unity across a threshold between 500 and 1000 C km (over 5-10 ms duration). In supercells, were large peak current CGs (of either polarity) associated with relatively small charge moment changes? On 29 June 2000, a +CG dominated daytime supercell passed through the LMA. Analyses suggested that charge moment changes were uniformly too small to trigger sprites based on the above criteria. A nocturnal supercell on 25 June produced two sprites from the last two +CGs of the storm. A detailed analysis of Mq values confirmed these two events were both associated with large Mq events, with metrics similar to those found in MCS sprite producing systems. But the analysis also found three large Mq events occurred during the middle of the storm. Re-analysis of the video records found 3 very small sprites. These mid-storm events appear to be rarities associated with highly impulsive removal of charge from relatively small volumes of cloud having high charge densities.

Analyses of the lightning metrics (peak current, Mq, estimated charge lowered, etc.) from the 23 June 2003 Aurora, NE record hail stone event (during BAMEX) are also underway. During the large hail phase of this supercell, only 3% of the CGs were positive, with strokes of both polarities having very low average and maximum peak currents. A new ULF/ELF/VLF technique to estimate impulse stroke Mq values (first 2 ms) is being applied. Implications for storm electrification processes as well as sprite potential will be assessed.