No abstract

Sprites are optical emissions in the mesosphere mainly at altitudes 50-90 km. They are caused by the sudden redistribution of charge due to lightning in the troposphere which can produce electric fields in the mesosphere in excess of the local breakdown field. The resulting optical displays can be spectacular and this has led to research into the physics and chemistry involved. Imaging at faster than 5,000 frames per second has revealed streamer discharges to be an important and very dynamic part of sprites, and this paper will review high-speed observations of sprite streamers. Streamers are initiated in the 65-85 km altitude range and observed to propagate both down and up at velocities normally in the 10 6-5 9 10 7 m/s range. Sprite streamer heads are small, typically less than a few hundreds of meters, but very bright and appear in images much like stars with signals up to that expected of a magnitude-6 star. Many details of streamer formation have been modeled and successfully compared with observations. Streamers frequently split into multiple sub-streamers. The splitting is very fast. To resolve details will require framing rates higher than the maximum 32,000 fps used so far. Sprite streamers are similar to streamers observed in the laboratory and, although many features appear to obey simple scaling laws, recent work indicates that there are limits to the scaling.

The complex dynamics of a sprite discharge are not limited to the propagation of streamers. After the passage of a streamer head, the ionized channel established in its wake develops intricate luminous patterns that evolve on timescales from 1 up to 100 ms. To investigate these patterns, conventionally called beads and glows, we present high‐speed recordings of their onset and decay; our main observation here is that in many cases distant points within a channel decay at the same rate despite considerable differences in the underlying air density. We then show that the properties of beads and glows, including this synchronized decay, are explained by the tendency of electric current within a streamer channel to converge to an uniform value and by an attachment instability of electric discharges in air. However, we also discuss the uncertainty about the chemical reactions that affect the electron density during the sprite decay.

Propagation and splitting of sprite streamers has been observed at high temporal and spatial resolution using two intensified high-speed CMOS cameras recording at 10 000 and 16 000 frames per second. Concurrent video recordings from a remote site provided data for triangulation allowing us to determine accurate altitude scales for the sprites. Diameters and speeds of the sprite streamers were measured from the high-speed images, and the diameters were scaled to the reduced diameters based on the triangulated locations. The sprite streamers with larger reduced diameter move faster than those with smaller diameter; the relation between the reduced diameter and speed is roughly linear. The reduced diameters at ≈65–70 km altitude are larger than streamer diameters measured at ground pressure in laboratory discharges indicating a deviation from the similarity law possibly due to the effects of the photoionization and an expansion of the streamer head along its propagation over a long distance. The reduced diameter and speed of the sprite streamers agree well with the diameter–velocity relation proposed by Naidis (2009 Phys. Rev. E 79 057401), and the peak electric field of the sprite streamers is estimated to be approximately 3–5 times the breakdown threshold field.

[1] High time resolution (10,000 frames per second) images of sprites combined with multistation concurrent video recordings have provided data for triangulation of the altitude of the initial sprite onset. The high-speed images were obtained from the Langmuir Laboratory, New Mexico, during summer campaigns in 2007 and 2008 with video observations from sites at Portales, New Mexico, and Las Vegas, New Mexico. Sprites start with one or more downward-propagating streamer heads. The triangulated onset altitudes of this initial downward streamer vary between 66 and 89 km. In some sprites the downward streamers are followed a little later by upward-propagating streamers. The upward streamers start from a lower altitude and existing luminous sprite structures and their triangulated altitudes vary from 64 to 78 km. The downward streamers create C sprite characteristics, while the upward streamers form the broad diffuse tops of carrot sprites. In the sprites analyzed the higher onset altitudes for the downward-propagating initial streamers were associated with C sprites and the lower with carrot sprites, but our larger data set indicates that this is not generally the case. It appears that the dominant sprite types vary from year to year, indicating that some longer-lasting environmental parameter, such as mesospheric conductivity and composition or thunderstorm cloud dynamics, may play an important role in determining the types of sprites observed.