Lightning-driven electric and magnetic fields measured in the stratosphere during the Brazil 2002-03 Balloon Campaign

1
Lightning-driven electric and magnetic fields
measured in the stratosphere during the Brazil
2002-03 Balloon Campaign
Jeremy N. Thomas (U. of Washington,
jnt_at_u.washington.edu), Robert H. Holzworth (U. of
Washington, bobholz_at_ess.washington.edu), Michael
P. McCarthy (U. of Washington, mccarthy_at_ess.washin
gton.edu), Osmar Pinto Jr. (INPE, Brazil,
osmar_at_dge.inpe.br)
I. Abstract
IV. Thunderstorms during Flight 1
V. Quasi-electrostatic fields due to nearby
lightning compared to a numerical model
During the Brazil Sprite Balloon Campaign
2002-2003, thousands of the lightning-driven
electric and magnetic field changes were measured
in the stratosphere above 30 km in altitude.
Two-hundred of these lightning events occurred
within 75 km horizontal distance of the balloon
payload and drove quasi-electrostatic field (QSF)
changes of up to 140 V/m measured at the payload.
Although the sprite imaging cameras were blocked
by clouds during these balloon flights, these QSF
can be predicted at sprite altitudes using the in
situ data along with a numerical QSF model. It is
found that the electric fields at sprite
altitudes (60-90 km) never surpass breakdown when
the fields for each of these 200 nearby lightning
events are propagated to the mesosphere. In
addition to measuring nearby electric fields, the
balloon payloads measured thousands of ELF to VLF
electric and magnetic fields from distant
lightning (gt75 km). These ELF to VLF fields
generally agree with ground-based measurements
Cummer et al., GRL, 25, 1281, 1998 and models
Pasko et al., GRL, 25, 3493, 1998 but are in
disagreement with the previous balloon-borne
measurements during the Sprites99 Campaign
Bering, Adv. Space Res., 34, 1782, 2004.
Comparison between model (red line) at 34 km
horizontal distance from payload and 34 km in
altitude and data (blue points) for two CG
strokes at 050 s and 064 s which occur at about
000009 UT. top panel vertical electric field.
bottom panel radial electric field.
Eighty minutes of dc vertical electric field data
during Flight 1. The two largest field changes
measured occurred at about 000009 and 001603
UT, seen as large negative transients, and were
correlated with CG flashes measured by the
Brazilian Lightning Network (BIN).
GOES8 Satellite IR image, 2345 Dec. 6, 2002
(courtesy of CPTEC, Brazil)

• II. Motivations
• Address the following questions regarding how
  thunderstorms affect the middle and upper
  atmosphere
• Are the magnitudes and relaxation times of nearby
  lightning-driven quasi-electrostatic fields (QSF)
  above thunderstorms sufficient for sprite
  production and growth?
• Do lightning-driven ELF to VLF electromagnetic
  field changes measured in the stratosphere
  indicate that current is flowing in the
  mesosphere after CG strokes?
• Do QSF and ELF/VLF models and other experiments
  agree with these measurements?

Model prediction of the vertical electric field
magnitude vs. altitude at R 0 km using the best
fit parameters to the measured electric field
change at Z 34 km, R 34 km. The vertical
electric field magnitude at three instants in
time (1 ms before the 1st CG stroke, 1 ms after
the second CG stroke, and 350 ms after the
second CG stroke) is compared to the various
breakdown thresholds.
Flight 1 Trajectory
Flight 1 Trajectory and CG Lightning
Quasi-electrostatic fields due to nearby -CG and
IC lightning
III. The Brazil Balloon Campaign 2002-2003
Overview
VI. ELF to VLF field changes due to distant
(gt75km) lightning

• Primary Measurement Objective Measure nearby
  (lt75 km) lightning-driven quasi-electrostatic
  fields.
• Logistics Balloon payloads launched from
  Cachoeira Paulista, Brazil
• (about 200 km northeast of Sao Paulo) on Dec. 6,
  2002 and March 6, 2003.
• On-board Measurements dc to VLF vector electric
  fields, VLF magnetic fields, x-rays, and optical
  power at a float altitude of 32-35 km.
• Results The electric and magnetic fields driven
  by thousands of lightning events were measured by
  the payloads, including 200 events with 75 km.

IC Flash
-CG Flash

• VII. Conclusions
• The amplitudes and relaxation times of the nearby
  lightning events generally agree with the
  numerical QSF model developed by the authors
  verifying that the QSF approach is valid.
• Positive cloud-to-ground (CG) lightning and
  in-cloud (IC) lightning generally produce larger
  nearby electric field changes compared to
  negative cloud-to-ground lightning (-CG), which
  agrees with the correlation between large CG
  strokes and sprites.
• It is found that the electric fields at sprite
  altitudes (60-90 km) never surpass conventional
  breakdown when the fields for each of these 200
  nearby lightning events are propagated to the
  mesosphere. The relativistic runaway threshold is
  surpasses from some of these events but at too
  high of an altitude to allow for enough avalanche
  lengths (e-foldings of the electron population)
  to cause optical emissions.
• The typical relaxation times of these predicted
  electric fields (70 ms at 70 km) are long enough
  to allow sprites to initiate and grow.
• The Brazil payloads measured ELF to VLF field
  changes for every CG stroke at 75-600 km distance
  from the payload and rarely measured delayed ELF
  pulses after these CG sferics. This disagrees
  with the Sprites99 payloads which rarely measured
  the CG sferics but for 90 of CG strokes measured
  a delayed ELF pulse, which they have attributed
  to mesospheric current flow Bering, Adv. Space
  Res., 34, 1782, 2004 Bhusal, Adv. Space Res.,
  34, 1811, 2004.

ELF to VLF fields were measured for every CG
stroke detected by BIN. For 2 (15/750) of -CGs
and 17 (31/184) of CGs a CG delayed pulse was
measured which may indicate mesospheric current.
This bias to CGs suggests that mesospheric
currents are driven by large charge moment
strokes and that the QSF from these strokes
initiate the mesospheric breakdown. The example
shown here is a 111 kA CG 328 km horizontal
distance from the payload.
Payload Configuration, Nov. 2002
Acknowledgments Work supported by NSF grants
ATM-0091825 and ATM-0355190.