1. Photograph by William Biscorner The World of TGFs David M. Smith Physics Department and Santa Cruz Institute for Particle Physics University of California, Santa Cruz Special thanks to: Brian Grefenstette and Bryna Hazelton, UCSC Joseph Dwyer, Florida Institute of Technology

2. Photograph by William Biscorner Terrestrial gamma-ray flashes, relativistic runaway, and high-energy radiation in the atmosphere David M. Smith Physics Department and Santa Cruz Institute for Particle Physics University of California, Santa Cruz Special thanks to: Brian Grefenstette and Bryna Hazelton, UCSC Joseph Dwyer, Florida Institute of Technology

3. Photograph by William Biscorner Mechanisms of air breakdown (decreasing E): Cold runaway Any e- goes relativistic Conventional (Townsend) Ionization > attachment Streamer Self-propagating Relativistic runaway Rel. seed electron(s) Leader Thermal ionization Figure by V. Pasko, from tutorial at the NATO summer institute on Sprites, etc., Corte, Corsica, 2004

4. Photograph by William Biscorner Graphic: Canadian Forest Service +CG +IC -CG Types of lightning stroke:

5. Photograph by William Biscorner - - - - - - - - - - - When/where might runaway occur relative to lightning? Before (initiation) During (EMP/Elve) After (Sprite)

6. Photograph by William Biscorner How does lightning trigger? M. Stolzenburg et al., GEOPHYSICAL RESEARCH LETTERS, VOL. 34, L04804 Relativistic runaway? Streamers from enhanced fields near hydrometeors?

7. Photograph by William Biscorner Timeline of relativistic runaway Wilson (1925) predicts runaway electrons Wilson (1925) predicts runaway electrons Many observational attempts – mixed results Many observational attempts – mixed results Gurevich et al. (1992) predict runaway avalanche Gurevich et al. (1992) predict runaway avalanche 1980's to today: modern observations from ground, space, aircraft, balloons – bremsstrahlung means remote sensing 1980's to today: modern observations from ground, space, aircraft, balloons – bremsstrahlung means remote sensing Dwyer (2003/2007/2008) Babich 2005: feedback mechanism Dwyer (2003/2007/2008) Babich 2005: feedback mechanism  counts w/ lightning Geiger counters (10’s of km away) Schonland & Viljoen (1933)  counts, attributed to x- rays from thunderstorms Dosimeters on 500m tower Whitmire (1979)  counts w/ T-storms Scintillation counter - mountain top Shaw (1967) No lightning effect Geiger counters (~2000 km away) Appleton & Bowen (1933) No high energy electrons Photographic plates on balloon Macky (1934) No high energy electrons Ionization chamber under T-storms Schonland (1930) No high energy electrons e - sensitive emulsions - 300m tower Hill (1963)  counts w/ T-storms Ionization chambers & counters Clay et al (1952)  counts w/ lightning Cloud chamber synch. w/ lightning Halliday (1934) ResultExperimentPaper Thanks to B. Hazelton

8. Photograph by William Biscorner THREE Requirements for RR avalanche Sufficient field (E > 2 kV/cm) Sufficient potential drop (> 10s of MV) Sufficient density for collisions giving multiplication

9. Photograph by William Biscorner Positron & gamma-ray feedback Feedback limits the number of avalanche lengths possible; But feedback allows true discharge even with only a few avalanche lengths Feedback predicts correct TGF duration and approximate luminosity (Dwyer 2008) Plot courtesy of J. Dwyer

10. Photograph by William Biscorner Classes of high-energy observations Surges – second-to-minute enhancements seen from ground, balloons, aircraft – runaway without breakdown? Steps – microseconds, seen from ground in stepped leaders – 10s of keV energy – cold runaway in a small volume? Terrestrial Gamma-ray Flashes – millisecond duration, seen from space, ground – MeV energies – true runaway breakdown?

11. Photograph by William Biscorner Observations of surges Left: Eack et al. 1996, balloon; Right: McCarthy & Parks 1985, airplane Surge enhancements often terminated or re-started by lightning

12. Photograph by William Biscorner New result: surges have RR spectrum From Tsuchiya et al. 2007, PRL – ground-based observation, winter thunderstorms over Japan:

13. Photograph by William Biscorner Recent observations of steps Natural & rocket triggered lightning – Short bursts of x-rays during the dart-leader or step-leader phase – Energies up to ~250 keV – Cold runaway (E~E c ) in leader tips? J. R. Dywer et al, Science, 299, 694 (2003); J. R. Dwyer et al, Geophys. Res. Lett, 31, L01803 (2005)

14. Photograph by William Biscorner TGFs seen from the ground have RR spectrum From natural lightning: Moore et al. 2001, GRL From triggered lightning: Dwyer et al. 2004, GRL Time history (left) and spectrum (above) from Dwyer et al. 2004

15. Photograph by William Biscorner TGFs from space: BATSE and RHESSI: 1991-2000 Large area Trigger required 76 TGFs in 9 years 4 energy channels. Fishman et al. 1994, Science 2002-present Smaller area No trigger needed > 800 TGFs in 6 years Fine energy resol. Smith et al. 2005, Science

16. Photograph by William Biscorner From an accelerated electron to a count in your detector: Ionization losses compete with acceleration; scattering broadens the electron beam Bremsstrahlung softens photon spectrum relative to electron spectrum and gamma beam is broader than electron beam Compton scattering of gammas in atmosphere softens and broadens the beam further – softer photons are delayed Photoelectric absorption in atmosphere removes lowest-energy gammas Compton and Photoelectric effects in spacecraft mimic atmospheric effects (except for delay) Deadtime and pileup in detectors can further distort spectra

17. Photograph by William Biscorner Summed RHESSI TGFs have RR spectrum: Flattening of spectrum @ 1 MeV requires production altitude 15-21 km: near tropical tropopause (just above thunderclouds) Dwyer & Smith 2005 Carlson, Lehtinen and Inan 2007

18. Photograph by William Biscorner Summed RHESSI TGFs have RR spectrum: Depth also affects absorption at the lowest energies (harder to observe) Spectrum is softer away from beam axis (Ostgaard et al. 2008) Deep source implies much higher energies (can be hundreds of thousands vs. thousands of Joules) See spectral modeling talks by B. Hazelton T. Gjesteland N. Ostgaard P. Connell

19. Photograph by William Biscorner RHESSI TGFs: Seasonal & diurnal dependence follows lightning

20. Photograph by William Biscorner Deficit at high latitudes relative to lightning, consistent with production at tropopause (E. Williams et al. 2006, JGR) ?=?= Trop. vs. latitude, B.D. Santer et al. 2002

21. Photograph by William Biscorner (Cummer et al. 2005, GRL; Inan et al. 2006, GRL) What kind of lightning is associated with TGFs seen from space? Use observations of ELF/VLF radio “atmospherics/sferics” Statistical surveys: TGF sferics are present but weak Single cases close-up: these are +IC strokes: (Stanley et al. 2006, GRL) Highest altitude lightning – consistent with tropopause hypothesis – all lightning may have TGFs!

22. Photograph by William Biscorner (Cummer et al. 2005, GRL; Inan et al. 2006, GRL) What kind of lightning is associated with TGFs seen from space? Use observations of ELF/VLF radio “atmospherics/sferics” Statistical surveys: TGF sferics are present but weak Single cases close-up: these are +IC strokes: (Stanley et al. 2006, GRL) IS TGF CAUSE OR EFFECT OF LIGHTNING?

23. Photograph by William Biscorner (Cummer et al. 2005, GRL; Inan et al. 2006, GRL) What kind of lightning is associated with TGFs seen from space? Use observations of ELF/VLF radio “atmospherics/sferics” Statistical surveys: TGF sferics are present but weak Single cases close-up: these are +IC strokes: (Stanley et al. 2006, GRL) IS TGF CAUSE OR EFFECT OF LIGHTNING? NEED BETTER THAN RHESSI 2ms TIMING!

24. Photograph by William Biscorner Spectral evolution consistent with Comptonization B. Grefenstette et al. 2008 Greater soft lags in BATSE data (Nemiroff et al. 1997 JGR; Feng et al. 2002 GRL) are an instrumental deadtime effect

25. Photograph by William Biscorner Secondary electron beams MUST occur:

26. Photograph by William Biscorner Some BATSE events and one RHESSI event are clearly electron beams – but is the Compton mechanism adequate?

27. Photograph by William Biscorner Some BATSE events and one RHESSI event are clearly electron beams – but is the Compton mechanism adequate? See talks by B. Grefenstette B. Carlson

28. Photograph by William Biscorner Questions remaining: Causality: (are TGFs a cause or effect of IC lightning?): need timing Ubiquity: (do all types of lightning include TGFs?): need location Relation to surges: need wide dynamic range & high sensitivity Relation to optical phenomena (sprites, elves, etc): need common platform Size, nature of avalanche region: need larger and more sensitive dataset with localizations ? Are the RHESSI events the tip of the iceberg?

29. Photograph by William Biscorner Because BATSE was more difficult to trigger, BATSE TGFs are expected to be intrinsically brighter than RHESSI.

30. Photograph by William Biscorner Because BATSE was more difficult to trigger, BATSE TGFs are expected to be intrinsically brighter than RHESSI. The opposite is true. BATSE is ~4x saturated

31. Photograph by William Biscorner A new result (which shouldn't be): Most RHESSI TGFs count at maximum throughput in the brightest 50 microseconds. Stay tuned for statistical estimates of average saturation factor...

32. Photograph by William Biscorner Questions remaining: How bright can TGFs really get? Ask Sprite-SAT/ TARANIS/ ASIM... ? ?