P 2 . 6 Lightning , Supercells and Sprites

The serendipitous discovery of red sprites by the late Prof. Jack Winckler, while testing a low-light television camera (LLTV) for a sounding rocket mission in 1989, changed forever our view of the middle atmosphere (Franz et al. 1990). Once thought to be dynamically and electrically quiescent, the stratosphere and mesosphere are increasingly understood to be regions of intense gravity waves (much generated by tropospheric convection) and a growing variety of lightning-related electrical discharges and intense transient electric fields (Lyons and Armstrong 2004). The discovery of literal cloud-to-stratosphere electrical discharges from intense thunderstorm tops, including blue jets, giant jets and true upward lightning (Lyons et al. 2004a) continues to engender the need for intensive investigations of this region. However, the relative scarcity and apparently random nature of cloud top discharge events makes them difficulty to study in a systematic manner. Red sprites, by contrast, are increasingly well understood and predictable. During the summer of 2000, a major field program, the Severe Thunderstorm, Electrification and Precipitation Study (STEPS) was conducted on the U.S. High Plains. While its focus was on supercell convection, especially low-precipitation events, the experimental design also allowed for detailed investigations of mesoscale convective systems (MCSs). Sprites are known to frequently occur in association with positive cloud-to-ground (+CG) strokes in MCSs (Lyons 1996), although even in the most productive storms rarely do more than 1 in 5 +CGs trigger a sprite. Theoretical research into red sprite production has seen the proposal and disposal of a number of theories. At the current time, sprites are generally agreed to be the result of conventional dielectric breakdown at approximately 70-75 km height, the result of a strong transient electrical field resulting from the removal to ground of large amounts of electrical charge in a CG flash (Pasko et al. 1996). Though this theory is not polarity dependant, the vast majority of sprite parent CGs are positive (SP+CGs), with only two documented –CG events on record (Barrington Leigh et al. 1999). While the peak current of SP+CGs is typically 50% larger than the other +CGs in the same storm (Lyons et al. 2003b; Lyons 1996), the peak current by itself is not a good predictor of sprite formation.