1. Sprite Research in Brazil: an overview Fernanda São Sabbas National Institute for Space Research - INPE, São José dos Campos, SP, Brazil Acknowledgements: Thanks to COST P-18 and to FAPESP

2. Outline Observation of TLEs: 4field campaigns: 3 international, 1 Brazilian Building infra-structure for sprite research: acquisition of equipment, education of students, collaboration network, need more sprite scientists Data analysis of sprites, lightning and thunderstorms from all campaigns Modeling sprite generation: quasi-electrostatic model of CG electric fields

3. Campaigns Summary Total TLEs: 18 sprites on the first campaign 11 TLEs (9 sprites and 2 halos) on the second >500 TLEs (>400 sprites and >100 halos) on the third 27 TLEs (sprites) on the last campaign Parent +CGs: from 26 to 150 kA when covered by BLDN. Type of thunderstorms: small MCSs associated with cold fronts small isolated storms large MCSs (3 rd and last campaigns) Charge moments calculated for 1 st campaign (Tohoku Univ. ) 240- 1500 C.km, with a mean <600 C.km, and 3 rd campaign (Duke Univ.) 100-800 C.km, with a mean ~255 C.km.

4. FAPESP Project: DEELUMINOS 1 – To Build infrastructure by buying the necessary equipment to observe Sprites, and TLEs in general, and the necessary computational resources for data analysis and modeling 2 – To perform campaigns and routine observations in Brazil, and participate of campaigns around the globe, to observe TLEs and their relationship with other phenomena in the MTI, collecting data with the purchased equipment 3 – To analyze the data obtained in the campaigns and routine observations 4 – To model the generation mechanisms, the relationship of TLEs with other MTI phenomena, their effects in the atmosphere using computer simulation Several of the planned activities involve collaboration with Brazilian and International experts in a number of research fields

5. Distribution of the distance between the sprite and parent +CG flashes shows that most sprites are laterally displaced from +CG [Wescott et al., 1998, 2001, São Sabbas et al., 2000, 2003a]. Not explained by current models using laminar atmosphere. Motivation for model

6. Possible sources of inhomogeneities are: Density variations caused by upward propagating gravity waves generated by convective activity in thunderstorms [Rowland et al., 1996; Pasko et al., 1997b]. Patches of electron temperature and density perturbations caused by interference patterns in the lower ionosphere generated by intracloud lightning [Valdivia et al., 1997]. Ionization trails from meteorites [Wescott et al., 2001]. Motivation for model

7. Sprites and neutral density perturbations

8. Poisson’s equation is solved using a Matlab Sylvester Eq. Solver [L z  +  L r = -(1/  0 )  tot ] Continuity equation is discretized using standard first finite differences method Boundary condition:  /  r = 0 everywhere  Quasi-electrostatic model

9. Conductivity

10. Self-consistent formulation: depends on E

11. Charge and Electric field

12. Laminar cases

15. perturbed cases

17. Negative perturbed case

18. Sprite Research in Brazil: an overview Fernanda São Sabbas National Institute for Space Research - INPE, São José dos Campos, SP, Brazil Acknowledgements: Thanks to COST P-18 and to FAPESP

19. Second Campaign, GW and sprite convective sources Grayscale GOES 12 IR image showing the locations of the cloud-top regions with T< -66 o C color enhanced in 2o C intervals. Dry to wet transition: extra tropical cyclone genesis over Atlantic, advected moisture from the Amazon, deep convection over Central Brazil, cold fronts. Tropopause around -76º C, equivalent to ~15200 m.

20. Second Campaign, GW and sprite convective sources

22. Second Campaign Summary Convective plumes: –high incidence of low peak current –CGs (bulk). –Vigorous convective plumes capable of generating strong gravity waves in areas with T < -76 o C, with dT ~ -2 o to -8 o C from average. Typical tropopause overshoots of ~200–3100 m. –Plumes with ~30 min lifetime and ~5-20 km diameter. Diameter correlated with dT (57%). Gravity waves: a total of 120 display observed, 94 short-period 26 medium-scale waves. Spread F and plasma bubbles observed on many nights. Coincident observations of GW and bubbles.

23. Overview of Sprite Campaigns in Brazil Fernanda São Sabbas (1), Michael J. Taylor (2), Dominique Pautet (2), Matthew Bailey (2), Natalia N. Solorzano (3), Jeremy N. Thomas (4,5), Robert H. Holzworth (4), Steven A. Cummer (6), Osmar Pinto Jr. (1), Sharon Vadas (7), Peter Stamus (7), David C. Fritts (7), Nelson J. Schuch (8), Vinicius T. Rampinelli (1), and Caitano L. da Silva (8,9) 1 – National Institute for Space Research - INPE, São José dos Campos, SP, Brazil 2 – Utah State University - USU, Logan, UT, United States 3 – Digipen Institute of Technology, Redmond, WA, United States 4 – University of Washington - UW, Seattle, WA, United States 5 – Geomagnetism Program, USGS, Denver, CO 6 – Duke University – DU, Durham, NC, United States 7 – Colorado Research Associates – CoRA/NWRA, Boulder, CO, United States 8 – Southern Regional Space Research Center-CRSPE/INPE, Santa Maria, RS, Brazil 9 – Universidade Federal de Santa Maria - UFSM, Brazil

24. Third Campaign Goal: continuation of the first campaign Ground site at Southern Space Observatory – SSO Observation period: 02-03/2006 International collaboration: INPE, UW, USU, Duke INPE facility at University of Santa Maria, near Southern Space Observatory (29.5° S, 53.8° W)

25. Observed TLE rate grows with storm size until data gap in IR images WWLLN peaks at ~04 UT and TLEs at ~05 UT (Solorzano et al. and Sao Sabbas et al., AGU Fall Meeting, 2007) Third Campaign, TLE, WWLN and cloud top T

26. Argentina all CGs: ~ -40 o C in the beginning of the night Maximum occurrence associated with ~ -50 o C US -CGs: ~ -62 o C in the beginning of the night Maximum ocurrence associated with ~ -68 o C Third Campaign, cloud top T and CG rates

27. Argentina all CGs: ~ -40 o C in the beginning of the night Maximum occurrence associated with ~ -50 o C US +CGs: ~ -69 o C most of the night Third Campaign, cloud top T and CG rates

28. Argentina sprites + halos + elves: ~ -45 o C in the beginning of the night ~ -55 o C in the end of the night US sprites: ~ -63 o C < T c < ~ -68 o C throughout the night Third Campaign, cloud top T and sprite rate

29. 3 rd most active sprite storm reported Most sprites in stratiform region More sprites above cloud regions warmer than -60 C compared with US High Plains Halo altitude (83 ± 2 km) and diameter ( 58 ± 2 km) similar to US High Plains Impulsive charge moment changes appear lower than U.S. High Plains (mean of 255 C.km)– more analysis needed Rare -CG sprite-halo observed, only 4 th confirmed, first time over land-based mesoscale storm Third Campaign Summary

30. Goal: to establish local capability of performing sprite observations Ground site at Southern Space Observatory – SSO Observation period: 10/2007 Southern Space Observatory (29.5° S, 53.8° W) Fourth Campaign, First Brazilian Measurements

31. Fourth Campaign, Instrumentation Sprite imager: CoolView EM 1000/TV, Photonic Science Sony digital tape recorder KIWI-OSD GPS time inserter

32. Fourth Campaign, First Brazilian Measurements

34. TECHNICAL SPECIFICATIONS Resolution: 1004x1002 pixels EM-CCD Pixel size: 8 micron pixels Read out rate: 20MHz Digitisation: 12-bit or 16-bit digitisation Frame rate: 17 frames sec at 17MHz Analogue output: 25 f·p·s CCIR Cooling: –30°C, to minimise dark current Read out modes: Full CCD, windowing (any size/shape) and binning, with 100 f·p·s or better. Analogue provides for a sub- area of pixels to be read at a faster rate. Gain: Electron multiplier gain from 1 up to 1000 times Spectral response: 400nm-1080nm Typical CCD %QE > 65% @ 550 - 700nm Readout noise: < 1 electron (at maximum CCD gain) Dark current < 1 electron/pixel/sec (typical delta of 45 O C from ambient operating temperature) Read out: LVDS or CCIR Camera size: Head 76mm (Ø) x 190mm (L) Camera Weight: 1.25kg Trigger mode: Allows image capture on demand Low Light Level Cameras TECHNICAL SPECIFICATIONS Resolution: 512x512 pixels Pixel size: 15 micron pixels Read out rate: 10, 5, 3, 1 MHz Digitisation: 14-bit (16-bit @ 1 MHz) digitisation Frame rate: 34 to several 100s Cooling: thermoelectric cooling down to -90°C Read out modes: Full CCD and binning Gain: Electron multiplier gain from 1 up to 1000 times Hermetically sealed housing: Designed for deep cooling Spectral response: 400nm-1080nm Typical CCD %QE > 65% @ 420 - 820nm (90% @ 500-650 nm) Readout noise: < 1 electron (with electron multiplication) Dark current < 1 electron/pixel/sec (below -70°C ) Read out: digital Camera size: 17 cm x 14 cm x 15 cm Camera Weight: 3.1 kg Andor's iXon DV887 back illuminated EMCCD Photonic Science CoolView EM 1000/TV

35. Third Campaign, atmospheric T profile

36. Second Campaign, gravity waves GOES 12 convection observed in the IR channel at 20:54 UT on 24 October 2005. Cloud top temperatures are color coded, with solid (dashed) contours showing GW momentum flux magnitudes at 21:55 (22:15) UT at 200 km altitude.

37. First Campaign Goal: simultaneous in-situ electric and magnetic field measurements onboard stratospheric balloons and sprite ground/airborne observations Ground site at Cachoeira Paulista, and INPE airplane Observation period:11-12/2002 and 02-03/2003 International collaboration: INPE, UW, USU, UAF

38. First Campaign Summary A total of 18 sprites were recorded during the first sprite campaign in Brazil. Two possible elves were also imaged in March 2003. Parent +CGs detected by BLDN for 16 of these events, peak currents ranged from 26 to 150 kA. The possible elves were associated with very large +CGs (101, 146 kA). Satellite image data and lightning data show they were occurred over small isolated storms within strong cold fronts. ELF data from Syowa Station, Antarctica, and Onagawa, Japan, indicate charge moments in the range of 240-1500 C.km, with a mean <600 C.km.

39. Second Campaign Goals: to investigate the role of GWs in seeding sprites and strong ESF and plasma bubbles. Campaign period: 09-11/2005 Instrumentation: digisondes, ionosondes, VHF radars, GPS receivers, airglow and sprite imagers, etc, located in several ground sites in Brazil. International collaboration: INPE, CoRA/NWRA, USU, UnB, APL/JHU, UTD, CU, JRO, and others deep convection HF GWs, T ~ 10 – 30 min mesopause GWs (OH) ~ 90 km F-layer GWs (1356, 6300) ~ 200 km sprites