1. Two Similar CG Lightning Flashes with Vastly Different Associated IC Activity William H. Beasley 1, Douglas M. Jordan 2 and Stephanie A. Weiss 1 1 University of Oklahoma 2 University of Florida

2. Alternative Scholarly-Sounding Titles: A Tale of Two Lightnings or, On the Varieties of Lightning Experience

3.  Yesterday it became clear that in the context of NWP models, “lightning data” can mean a number of different things, including  NLDN or LLDN ground-strike data  LMA, LDAR or other VHF mapping data  Satellite optical data  WWLN ground-strike data

4.  There was some discussion about what is to be gained by incorporation of lightning data of one kind or another in various models.  There was some discussion about the use of different types or aspects of lightning data.

5.  Lightning data are being used in models as proxy for  Location (and timing?) of convection  Strength of convection inferred from flash frequency, flash density, etc.  Lightning data are being used for  Initialization  Nudging or adjustment, typically by addition of latent heat or adjustment of stability?

6.  There was some discussion about what constitutes a “flash”.

7.  To help inform your judgment about the use of lightning data, I am going to show you that two CG flashes with apparently very similar characteristics can have vastly different associated IC activity that could make a difference in how the lightning is registered by different observing systems.

8.  Further, I ask whether such differences could bear any relationship to differences in the parent storms?  In any case, clearly,

9. what you or I see

10. what an LMA/LDAR sees

11. what satellites see

12. What the WWLN sees

13. what the NALDN sees

14. what Alex Fierro saw from his balcony in Miami

15. can be quite different manifestations of lightning. For some purposes, it may not matter what you call a flash.

16. W.H. Beasley But, if you are going to use lightning characteristics such as flash frequency or flash density to modulate model parameters, then what constitutes a “flash”, as observed by any particular type of observing system, could matter.

17.  To show you some cool high-speed images  To show you that CG flashes with very similar characteristics can have vastly different associated IC activity  To leave you with a sense of perspective on relation between, and relative scales of, CG and IC portions of lightning discharges My Goals for Today

18.  For two CG flashes I will show  Ordinary HDTV Video  NLDN data  High-Speed Images  VHF Source Locations  Radar Reflectivity

19. NLDN Ground-Strike points and Camera Location at NWC ~ 4 km Estimated height of visible CG channel

20. HD video, 10:40:15 UT Flash  Ordinary HD camcorder  Two CG Flashes according to NLDN  One at 10:40:15.587 (to left of fov) probably a false positive CG  One of interest at 10:40:15.759

22. NLDN data @10:40:15 Flash Timelat lon I peak est UTCdegdegkA 10:40:15.58735.1159-97.4094+7* 10:40:15.75935.1665-97.4800-6 *false positive CG

23. High Speed Video  Flash at 10:40:15.759951  Obtained with Photron SA1.1  10,000 frames per second  0.6665 seconds of data  Stepped Leader in view for ~ 6 ms  Continuing current in view for ~300 ms  ~13 M-components in ~300 ms  HDTV video shows CC lasted >700 ms

24. 10,000 fps video, 10:40:15 flash

25. VHF LMA data @ 10:40:15  Standard view using xlma  3d projection using Google Earth +  Animations using Google Earth +  Animation provided by Ken Cummins and Jean-Yves Lojou

26. 10:40:15 Flash Plan View Projection of points onto vertical E-W plane time history Projection of points onto vertical N-S plane 40 km 25km25km

27. Flash at 05/07/08 1040:15 UTC: XLMA interpretation  initial negative leader moves upward into positive charge  negative breakdown propagates into the horizontal positive storm charge region centered near 9.5 km  as channel develops westward, it slopes downward to 6 km  positive leader propagates horizontally through negative storm charge centered near 4 km.  negative leader to ground starts near 4 km and propagates horizontally for 6 km before turning toward the ground  suggests localized region of positive storm charge near 3.5km just beneath the main negative storm charge.  Other flashes close in space and time to this flash also have negative leaders moving through this positive storm charge  however, most of these other flashes have leaders moving through only this positive charge region and the negative charge region above it.  negative leader to ground is unusual in this case because it seems to involve a third charge region within the same flash.  possible positive polarity breakdown moving along previously conducting negative channels (from the westernmost end of the positive storm charge region back toward the flash initiation point) propagates close enough to the negative storm charge to cause a large enough electric field to initiate a new negative leader.

28. All VHF source locations 10:40:15 flash

29. ~ 5000 VHF source locations for the flash at 10:40:15, displayed in a 3d rendering. Total duration of the flash is just under one second. Total spatial extent is a significant fraction of 100 km. Animated mapping of VHF sources

30. LMA data overview 10:40:15.759951 flash  Nearly 5000 VHF source locations  Just under 1 second duration  No clear gap in activity between IC and CG portions of discharge

32. VHF sources during 6 ms of stepped leader in Photron FoV

33. Whole cc interval

34. 10:40:15.759951 flash All LMA Points

35. HD video, 22:24:27 UT Flash  Ordinary HD camcorder  Two CG Flashes in fov  One at 22:24:27.669 of interest

37. Another High Speed Video  Trigger at 22:24:27.669368  Obtained with Photron SA1.1  10,000 frames per second  0.6665 seconds of data  Stepped Leader duration ~ 2.3 ms  Continuing current duration ~12 ms  No M-components

38. 10,000 fps video, 22:24:27

39. VHF LMA data

40. 22:24:27 Flash Projection of points onto vertical E-W plane Plan View time history 11 km 12km12km 4 km

41. Similarities and Differences  Stepped Leaders, Branching Similar  Upward-going leaders (attachment) Similar  Return-strokes Similar  Continuing Currents Different  104015 flash had continuing current of extremely long duration (~700 ms) and numerous M components  222427 flash had continuing current of ~ 12 ms duration and no M components

42. How do the Meteorological Situations Compare  Look at Radar Images

43. Radar Only Radar with VHF Sources Radar scan ~10:35 - 10:41 Radar and LMA data,10:40:15 flash, CAPPI at ~ 3 km

44. Radar Only Radar with VHF Sources Radar scan ~22:23 - 22:29 Radar and LMA data, 22:24:29 flash, CAPPI at ~ 3 km

45. 10:40:15 22:24:27 Radar and LMA data comparison

46. 10:40:15 22:24:27 Radar and LMA data comparison

47. Observations and Comments  There were very few VHF radiation sources determined by the LMA to be co-located and contemporaneous with the stepped leader, return stroke and M components.

48.  There were upward-going leaders bef ore both return strokes  In the last 100  s before rs  approximately 50 meters length.

49.  LMA data during photographically identified M-components suggests  discharge channel in cloud expands in at least two directions more or less simultaneously,  pretty much horizontally as suggested by Krehbiel (1979) and Proctor et al. (1988), but with some vertical extent also,

50.  presumably this shows additional charge sources being tapped to keep cc going.  The continuing current for the 10:40:15 flash lasted more than 700 ms, exceeding the longest previously reported cc duration by about 200 ms (to the best of our knowledge).

51.  The horizontal extent of the IC discharge during the extremely long cc is considerable, suggesting charge layer of considerable horizontal extent.  Comparison of radar images at time of the two flashes shows  extensive stratiform* layer associated with long cc in early morning flash  more vertical structure indicative of convection associated with short cc afternoon flash *with apologies to Chuck Doswel

52.  How would these two CG flashes appear to a satellite?  Do long continuing currents occur more often during stratiform storms?  Is large horizontal extent of charge layer necessary for long cc?  Is large horizontal extent of charge layer sufficient for long cc?  Is this possibly relevant to fire weather, since long ccs set fires? Questions

53. Bill Rison, LMA data Don MacGorman, LMA data Phillip Bothwell, NLDN data Ken Cummins and Jean-Yves Lojou, 3d visualization Julia Jordan, HD video support This work has been supported in part by NSF grant ATM-072119 and in part by NASA EPSCoR grant NNX07AV48A/Oklahoma NASA Spacegrant Acknowledgements A Good Example of AGU Motto: “Unselfish Cooperation in Research” We gratefully acknowledge the valuable help of

54. Stay Tuned!