Transcript Diapositive 1

Sprite and lightning infrasound measurements
during the 2005 Eurosprite campaign
T. Farges1, E. Blanc1, P. Herry1, T. Neubert2
1: CEA / DASE
2: DNSC
CEA/DASE T. FARGES
June 23, 2008
1
2005 Eurosprite campaign
Several instruments have been set up to study TLEs in the frame of CAL
network (Arnone et al., IRF report, 2008):
• cameras at Pic du Midi (F), as in 2003, and Puy de Dôme (F)
• Meteo observations: Meteorage lightning location, radar imaging of cloud
• VLF: narrowband receiver (Crete) and wideband receiver (Nançay, F)
• ELF: Finland, Poland, Crete, Israel and Hungary (Schumann resonance)
• ULF: Finland, Poland, Crete
• infrasound: Flers (F), as in 2003, and St Just (F)
CEA/DASE T. FARGES
June 23, 2008
2
Objectives and lightning activity
Two storms have been thoroughly studied to characterize:
1. lightning-induced infrasound
2. sprite-induced infrasound, particularly those produced close to the
infrasound station (< 200 km)
2 sprites observed by
Pic du Midi cameras
CEA/DASE T. FARGES
June 23, 2008
3
Lightning-induced infrasound and distance limit for detection
08/31/2005
Lightning distance from St Just station
Signal measured by one of the four
sensors
in St Just

Lightning-induced
infrasound
amplitude > background signal
when lightning distance < 60 km
Signal filtered from 0.1 to 9 Hz
(sampling
frequency
= 20 Hz)
individual
spectrum
CEA/DASE T. FARGES
June 23, 2008
4
Azimuth determination and one-to-one identification
Calculation of azimuth of arrival and horizontal apparent speed with the
Progressive Multi-Channel Correlation method (Cansi, GRL, 1995)
speed of sound
(km)
array configuration
(km)
CEA/DASE T. FARGES
June 23, 2008
5
Comparison of lightning azimuth with infrasound azimuth
Lightning distance (km)
 Very good agreement between
lightning azimuth and infrasound
azimuth
Remarks:
1. Good correlation is found also for
lightning distance > ~200 km
2. Less (even lack of) detection for
distance from 75 to 200 km.
infrasound
azimuth
50
10
h (km)
shadow zone
100
CEA/DASE T. FARGES
June 23, 2008
d (km)
200
6
Comparison of lightning azimuth with infrasound azimuth
Lightning distance (km)
 Very good agreement between
lightning azimuth with infrasound
azimuth
Remarks:
1. Good correlation is found also for
lightning distance > ~200 km
2. Less (even lack of) detection for
distance from 75 to 200 km.
infrasound
azimuth
CEA/DASE T. FARGES
3. long duration (~1 minute) and
large azimuth range (more than 30°)
events while lightning-induced
infrasound lasts few seconds and is
very localized (azimuth range < 5°)
June 23, 2008
7
Infrasound from sprites
first identification : EuroSprite campaign in 2003
Lightning map with Météorage data
sprite
location
Study case for July 21st storm :
• 28 sprites observed
• from 02:00 to 03:15 UT
• at 350 – 500 km from Flers
camera
• detection of several infrasound with
specific signature : chirp signal
(Liszka, 2004)
• delay between the time of arrival of
optical and infrasound signals is in
agreement with the calculated
infrasound propagation time (ray
tracing model + assumption on the
source location)
(Farges et al., GRL, 2005)
CEA/DASE T. FARGES
June 23, 2008
8
Other examples
2 successive images (40 ms)
 single chirp signal
CEA/DASE T. FARGES
maybe more complex
June 23, 2008
9
Main characteristics of infrasound of sprites
• The duration is directly linked to the
size of the sprite in the direction of
observation (Farges et al., GRL, 2005)
slope = 0.31 km/s
 mean propagation
~1000
speed km
• Detection range : ~ 1000 km
(Neubert et al., JASTP, 2005)
• Several chirps were detected
during dawn or morning time (before
or after sunrise, respectively) : sprite
detection is available even if
cameras are blind due to sky
brightness or masking by cloud
(Farges et al., GRL, 2005 and
Neubert et al., submitted, 2008)
CEA/DASE T. FARGES
03/07/21
camera
observation
D region
starts to
be ionized
Infrasound
chirps
June 23, 2008
10
Main characteristics of infrasound of sprites
• The chirp signature is due to the horizontal extension of the source and the
low pass filter property of the thermosphere:
Thermosphere
low pass filter
125 km
Altitude
130 km
50 km
Source at
60 km
f (Hz)
10
1
140 s
Farges et al., GRL, 2005
• Recent study by Pasko and Snively (AGU, 2007) shows:
“The reported infrasound signatures are consistent with pressure
waves due weak air heating on the order of several degrees K in
a vertically extended cylindrical volume with radial dimension on
the order of several tens of meters, in agreement with known
morphology of streamer channels in sprites.
The presented modeling results support the ideas advanced in
[Farges et al., 2005] that the chirp-shape can be explained by the
horizontal size of the sprite.”
CEA/DASE T. FARGES
June 23, 2008
11
Camera observation of sprite on September 9th
parent lightning location
21:18:39
Camera
FOV
?
20:35:42
Cameras turned
clockwise
Both sprites appear on the verge of the camera
FOV at ~150 km from St Just
Small sprites produce weak and brief infrasound
which are undetectable.
CEA/DASE T. FARGES
Time
June 23, 2008
12
Infrasound measurements when storm produced sprites
Lightning distance (km)
From 21:00 to 22:30, several long
duration events appear when lightning
are mainly in the shadow zone.
Some of these long duration events
present chirp shape in spectrogram
CEA/DASE T. FARGES
June 23, 2008
13
Wave parameters of chirp signal
We use wave parameters (azimuth and apparent speed) and assume a
direct propagation from source to sensor to locate the source.
incidence angle vs. apparent speed:
cos  
Cs
Sapp
source
height
ds
incidence
angle
Source located in
altitude
azimuth
North
ds = <Cs> Dt
<Cs> = 0.31 km/s
Dt: propagation time
To calculate ds, we must assume that it is a sprite and select a compatible parent
lightning. Now, parent lightning and sprite occur simultaneously. We know then the
propagation time Dt.
CEA/DASE T. FARGES
June 23, 2008
14
Calculation of 3D location of chirp infrasound source
altitude range
of sprite
21:10:40.631
140 km
29.7 kA
+CG
2 sprites ?
sprite / CG distance ~ 100 km
This sprite (if it is one) could not be
observed by Pic du Midi cameras because
it appears completely outside of their FOV
CEA/DASE T. FARGES
June 23, 2008
15
Conclusion
Main results:
• Lightning-induced infrasound is detectable and identifiable (one-to-one)
when distance < 50-75 km. Detection beyond shadow zone is possible
(when wind strength is weak) but one-to-one identification is not.
• Several long duration and large azimuth range events appear during
thunderstorm. Some of them have a chirp shape and have high apparent
speed meaning that infrasound source is located in high altitude.
• For one case, we calculate the 3D location of source using wave
parameters and hypothesis on possible parent lightning location (assuming
that it is a sprite). Results show that source altitude range is compatible with
sprite altitude range.  first observation of sprite at short distance and new
infrasound signature for sprite
Future work:
• Refine 3D location using ray-tracing propagation program.
• Confirm this with a combined observation (camera + infrasound)
CEA/DASE T. FARGES
June 23, 2008
16