Abstract Title: Comparison of Lightning Channel Luminosity Profiles in the Infrared and Visible Ranges

Abstract Submitted to: ATMOSPHERIC AND SPACE ELECTRICITY

Abstract Text:

Ding et al. (2020) obtained first infrared images of lightning and found that negative stepped leader branches were most heated near their lower ends and connected via cooler sections to the channel trunk. We now present, for the first time, medium-to-far infrared range (3 to 5 µm) luminosity versus time profiles for multiple-stroke and single-stroke lightning flashes and compare them with the corresponding visible-range (0.4 – 0.8 µm) luminosity versus time profiles. The data were obtained using high-speed framing cameras (FLIR with 1-ms interframe interval for infrared and Phantom with 50-µs interframe interval for visible) operating at the Lightning Observatory in Gainesville (LOG), Florida. Channel luminosity variations during 105 time intervals between individual strokes and after 67 last (or the only) strokes in the flash were examined.

We found that the infrared (IR) luminosity always persisted throughout the entire interstroke interval, which is in contrast with the visible luminosity that always decayed to a non-detectable level prior to a subsequent return stroke pulse. After the last (or the only) stroke of the flash, the IR luminosity persisted much (typically a factor of 10) longer than the visible luminosity. The largest IR luminosity duration was about 1 second. Further, the IR luminosity often exhibited a hump at the time the visible luminosity was monotonically decaying. The IR luminosity duration tended to increase with increasing the visible luminosity duration for relatively small values (up to a few tens of milliseconds) of the latter and showed much weaker or no correlation for larger values. Leaving aside long (>40 ms) continuing currents, the observed trend apparently indicates that longer continuing currents, identified by the visible luminosity, lead to longer afterglow periods, identified by the persisting IR luminosity. We also examined other factors, including the number of preceding strokes, the return-stroke peak current, and the occurrence of M-components, that could potentially influence the IR luminosity duration. Additionally, we used the cross-sectional IR luminosity profiles to estimate apparent channel widths at three different luminosity levels. A decrease with time in the IR luminosity was usually, but not always, accompanied by a decrease in the apparent channel width.