TLE Workshop: Ionospheric

1
TLE Workshop Ionospheric Magnetospheric
Effects
Lightning-induced Effects in the Ionosphere and
the Radiation Belts U. S. Inan1 1Space,
Telecommunications and Radioscience Laboratory
Stanford University, Stanford, California
94305 http//www-star.stanford.edu/vlf/ DEMETER
ICE and IDP data provided by courtesy of M
Parrot2 JA Sauvaud3 2LPCE/CNRS, 3A Avenue de
la Recherche Scientfique, Orleans,
France 3CESR/CNRS, 9 Avenue du Colonel Roche,
31028 Toulouse cedex 4, France With
contributions from Denys Piddyachiy, Bob Marshall
and Erin Selser Gemelos
2
Lightning-Induced Electron Precipitation (LEP)
3
From Sauvaud et al. 2008
200 keV electron distribution over 14 months
4
LEP Pitch Angle Distribution Inan et al.,1989
Even in the best LEP cases, electrons just
enter the very edge of the loss cone
5
LEP Events on DEMETERInan et al. 2007
6
Enhanced Precipitation Region Maintained by a
Thunderstorm
7
VLF Streaks on DEMETERParrot et al., 2008
8
Satellite-Based Detection of NPM-Induced
Precipitation

• Background energetic electron flux measured by
  DEMETER at longitude of NPM is relatively high.
• For the short-duration DEMETER passes, NPM may
  not induce precipitation which is significant
  compared to the measured background flux.
• Subionospheric detection offers longer-duration
  analysis with fewer complications due to
  background flux.

Bounce- Drift- Loss Cone Differential
Detected by DEMETER
At 700 km altitude and L 2, DEMETER will
detect particles of aeq 14-20.
9
LEP on 06-Oct-2007
10
LEP on 06-Oct-2007 (zoom in)
11
LEP Events Over the Eastern United States
12
LEP on 06-Oct-2007 (zoom out)
13
LEP over Eastern U.S.
14
LEP over the Eastern United States (at longitudes
of SAA)
15
LEPs at Longitudes of the SAA
16
Light-ion Depletion at LEO and Plasmapause
Position
17
Typical Equatorial Density Profile with
Plasmapause
18
Cyclotron Resonant Energy as a function of
L-shell
19
LEP at Longitude of the SAA
20
LEPs Off the East Coast of the United States
(Just East of SAA)
21
Next pass to the west (LEPs masked by intense SAA
precipitation)
22
LEPs as DEMETER moves into the SAA from the east
23
LEPs as DEMETER moves into the SAA from the east
24
LEPs observed in the presence of intense SAA
precipitation
25
DEMETER Passes over an Intense Thunderstorm
26
DEMETER Passes over an Intense Thunderstorm
27
DEMETER Passes over an Intense Thunderstorm
28
DEMETER Passes over an Intense Thunderstorm
29
DEMETER Passes over an Intense Thunderstorm
30
DEMETER Passes over an Intense Thunderstorm
31
DEMETER Passes over an Intense Thunderstorm
32
V-shaped event and LEP
33
Seasonal Variation Driven by Lightning?
34
(No Transcript)
35
(No Transcript)
36
Lightning-induced Electron Precipitation
37
Subionospheric VLF Remote Sensing
Many VLF transmitters operate worldwide,
providing a range of coherent laser-like signals
with which to probe the ionospheric regions
through which they propagate
38
Holographic Imaging of Lower Ionosphere
VLF receivers at 13 high schools Provides
excellent opportunities for outreach
39
(No Transcript)
40
Non-ducted LEP Events
(a)
(c)
In general whistler waves propagate in non-ducted
mode, illuminating large regions of the
radiation In each event, onset delay (Dt)an
onset duration (td) are measurable, corresponding
to wave/particle travel times and duration of LEP
pulse
(d)
(b)
41
Spatial Extent of LEP Events
VLF Amplitude Data for 24 March 2001
Dashed line VLF paths perturbed solid line ones
are not theoretical precipitation region
superposed

• Full extent of the ionospheric disturbance
  produced by an LEP burst (due to a single flash)
  is captured
• Corresponding region of the inner radiation belt
  is affected by whistler waves from a single
  lightning flash

42
LEP Events in Australia
Dt 1.28 sec DA 0.9 dB td 1.95 sec tr 162
sec
43
Theoretical Modeling
Peter and Inan 2006
44
Multi-mode VLF Propagation
45
Electron Precipitation

• Model calculates electron precipitation given
  lightning flash location and spectra, trapped
  radiation belt flux levels, plasmaspheric
  density, etc.
• Gives electron precipitation flux as function of
  L-shell, longitude, time, and energy
• Used to determine expected VLF signal
  perturbations

Model based on Bortnik 2004, results from Peter
and Inan 2007
46
VLF Signal Perturbations

• Reasonable agreement between modeled and observed
  VLF signal perturbations
• - Consistent in terms of location and scale
• - Each asterisk denotes individual VLF signal
  path

Peter and Inan 2007
47
Precipitation Metric (G)

• Define metric G to quantify electron
  precipitation along VLF signal path

Where E electron energy, t time of
precipitation, and F differential number flux
48
Quantifying Precipitation

• For the cases examined, precipitation along the
  VLF signal path is directly proportional to the
  observed VLF signal amplitude perturbation
• Conversion Ratio Y relates precipitation ( G ) to
  VLF signal perturbation ( DA ) via
• G / DA
• With the application of methodology to other VLF
  signal paths and other event types Y will need to
  be recalculated

49
Inverting to Precipitation
- With Y, we can infer the total precipitation
associated with a lightning flash using only
observations of VLF signal perturbations -
Integrate DA across paths and multiply by
conversion ratio Y to estimate total number of
energetic electrons (100-300 keV) precipitated (?)

• Example Calculation
• Perturbation area 1.46 x 106 dB-m
• Total Loss (?) 1.60.3x1016 electrons

50
Early/fast VLF Events
51
Sprites and Early/fast VLF Events Haldoupis et
al., 2005
52
18 August 1999 Marshall et al., 2007

• About 120 sprites over 6 hours
• 62 of sprites have associated early/fast events
  (52/84)
• Accounting for multiple sprites
• No causative CGs gt 32 km from any path

53
15 July 1995

• 34 sprites observed in 2 hours
• Early/fast events seen for all but 2 sprites
• 2 early/fast events seen without sprites

54
Sferic Bursts VLF Events Johnson and Inan,
1999
55
(No Transcript)
56
In-cloud lightning activity sprites van der
Velde et al., 2005
Nancay, France
57
What causes sferic bursts?AE51A-0263 R A
Marshall, U S Inan, W Lyons

• Rapid attenuation with long distance
• Indistinguishable mess of VLF energy individual
  sferics cannot be deciphered with VLF bandwidth
  (100 kHz sampling)
• Intra-cloud lightning
• horizontal dipole radiation has a null at low
  elevation angles, and high attenuation rates at
  high elevation angles (higher-order modes in the
  E-I waveguide)

58
LMA data shows intense IC activity associated
with Sprite
Many sprites are delayed from CG - why?
Sprites with larger delay typically smaller CG
59
Sprite sferic burst comparisonsAE51A-0263 R A
Marshall, U S Inan, W Lyons

• Analyzed thousands of samples of VLF data
  correlated with sprite observations
• Measured burst activity using quantitative energy
  measurements
• 90 of sprites have associated burst activity
• Only 50 of non-sprite related large CGs (gt 50
  kA) have associated burst activity
• Considerations
• Sprite lists used may not be entirely accurate
  some sprites missed, some phantoms. Clean-up in
  progress
• Some non-sprite CGs may have in fact produced
  sprites, since all CGs within 1000 km of YR were
  considered

60
Sprite-related sferic burstsAE51A-0263 R A
Marshall, U S Inan, W Lyons
61
Taranenko et al. 1993 (1D)?
62
Cross Sections of Inelastic Processes in Air
63
Other Model Concerns
Ionospheric Electron Density Profiles
64
Anisotropic 3D Simulation of Elves
65
Variations in the Ionospheric Profile
66
Density changes and optical emissions from 20 V/m
CG

• Note asymmetry due to magnetic field

67
Optical Emissions from 20 V/m vertical discharge

• Note asymmetry due to magnetic field

68
Multiple horizontal pulses 2 V/m
At 2 V/m, ionization threshold is not reached,
but optical is present Multiple pulses
cumulatively deplete ionospheric region
Asymmetry due to Earths magnetic field Here
are 20 pulses realistically, a sferic burst
may contain 100s of pulses
69
Multiple horizontal pulses 3 V/m
At 3 V/m, ionization threshold is reached, but
over the VOLUME perturbed, ionization is still
small Net change in electrons ? 5 1013 per
pulse Again, realistically, a sferic burst may
contain 100's of pulses of different amplitudes
70
Effect of Dipole Orientation (3 V/m, 5 km
altitude)?
71
Differences in Recovery Time
72
An unusual Early/fast event with 20 minute
recovery
73
Model of Ionospheric Chemistry

• Extension of GPI, new items highlighted
• 5 constituents
• Ne - electrons
• N - light positive ions O2, NO2, N2,
• Nx - positive ion clusters H(H2O)n
• N- - light negative ions O-, O2-,
• Nx- - heavy negative ions NO3-, NO3-(H2O)n

• Coefficients
• ai - mutual neutralization
• ad, adc - dissociative recombination
• b - attachment
• 3-body and 2-body (in E)
• g - detachment from light negative ions (value
  uncertain)
• Electron affinity0.4 eV ? highly dependent on T
• During daytime photodetachment0.4 s-1
• Also due to active species, Nac O, N, O2(a1Dg)
• gx - detachment from heavy negative ions,
  approximately0 (electron affinity 3.91 eV),
  photodetachment 0.002 s-1 (during daytime)
• B - rate of conversion N ? Nx
• A - rate of conversion N- ? Nx-
• Cosmic ray only source at low altitudes Qcr
  (peaks at 15 km)

74
Motivation for the new model Background electron
density

• Electrons
• Negative ions N-tot N- Nx-
• Solid lines new 5-species model dashed
  4-species GPI model
• GPI model overestimates electron density at low
  altitudes

75
Coefficients as a function of altitude (nighttime)

• Altitude dependence of detachment coefficient g
  is more complicated than suggested by GPI
• Increase of g at hgt70 km is due to active species
• The effective (average) geff is decreased
  compared to g due to presence of Nx- which has no
  detachment
• Two-body b is maximized around breakdown electric
  field

76
Electrical conductivity

• Initial ionization is caused by a giant blue jet
• Relaxation of ionization takes 2 stages
• Free electrons are quickly attached, t1 b-1
  (when geff0). In the presence of the electric
  field, initial relaxation is faster due to higher
  attachment coefficient b
• Positive and negative ions recombine, t1
  (aiNi)-1 104 s

77
Propagating VLF Modes

• Modes are calculated using Wait 1970 and taking
  into account the curvature of the Earth

78
VLF Losses

• Electric field present gt very rapid initial
  recovery
• Long enduring recovery is well exhibited
• Initial rapid recovery only last for the first
  few ms, which would not be detectable in typical
  VLF data

79
The End