Infrasound from lightning Jelle Assink and Läslo Evers Royal Netherlands Meteorological Institute Seismology Division ITW 2007, Tokyo, Japan
Jan 20, 2016
Infrasound from lightning
Jelle Assink and Läslo Evers
Royal Netherlands Meteorological Institute Seismology Division
ITW 2007, Tokyo, Japan
Low Frequency Array
• Astronomical initiative • Infrastructure ao. power, internet, computing and backup facilities• Dense (international) coverage
• Geophysical sensor network• Combined seismic/infrasound recording
LOFAR
LOFAR
Objectives
• Source identification through association• Atmospheric contribution to seismic noise• Seismo-acoustics by simultaneous observations• Local noise characterization
Practicalities• Adapt KNMI microbarometer for periods up to1000 s• Construct Very Large Aperture Infrasound Array 30 KNMI-mb’s at 1 to 10s of km
• Develop low cost infrasound sensor• Construct High Density Infrasound Array 80 sensor in 100x100 meter field
Cabauw Infrasound Array
• Combined meteo and infrasound project• Cabauw site: 215 m meteo tower• 3D sensing of the boundary layer
Objectives
• Detect gravity waves and other atmospheric phenomena• Applying infrasound technique to non-acoustic velocities• Relation between state of the boundary layer and infrasonic signal characteristics• 3D acoustical array for signal characterization as function of height
50 km
Source: NASA
Objectives• Detectability lightning discharges with infrasound
– To which extent– Distinction CC/CG– Source localization
• Content and behavior of related infrasound• Possible source-mechanisms• Wave propagation paths through atmosphere
• Comparison and verification KNMI lightning detection network based on EM (‘FLITS’)
Source mechanisms
• Few (1969): thermally driven expanding channel model, blast wave
• Bowman and Bedard (1971): convective system as a whole, vortices, mass displacement
• Dessler (1973): electrostatic mechanism, reordering of charges within clouds
• Liszka (2004): transient luminous events, such as sprites
Electromagnetic detectionKNMI FLITS network
LF antenna (around 4 MHz)
VHF array (around 110 MHz)
Electromagnetic detection
• FLITS: Flash Localisation by Interferometry and Time of Arrival System
• LF Antenna: Time-of-Arrival– Detection and localization– Discrimination CC/CG
• VHF array: interferometry– Detection and localization
• A minimum of 4 stations for unambiguous detections
Infrasound detection
KNMI IS network
Electromagnetic detections
at 01-10-2006
CC
CG
Cloud-to-Clouddischarge
Cloud-to-Grounddischarge
Infrasound & FLITS detections at DBN for 1-10-2006
CGCC
High F ISLow F IS
All-day observation summary• Correlation in time
between (nearby) discharges and coherent infrasound detections
• Nearby discharges:– High app. velocity– High amplitude– Coherent energy
over infrasound frequency band
Raw data
Time(s)
Pre
ssu
re(P
a)
Unfiltered data, strong front nose
Filtered data
Time(s)
Pre
ssu
re(P
a)
Bandpass 1-10 Hz, variety of impulsive events
Filtered data
Time(s)
Pre
ssu
re(P
a)
Bandpass 1-10 Hz, blast waves
Atmospheric attenuationInfrasound amplitude vs. distance from array
– Normalized for discharge size– Empirical attenuation relation: exponentially decaying?
Atmospheric attenuation
Log-log presentation
Atmospheric attenuationPower coefficient = 1 for cylindrical spreading
= 2 for spherical spreading
Conclusions
• CG discharges can be detected over ranges of 50 km, CC much harder to identify
• Thermally driven expanding channel model seems feasible, correlation with blast waves
• Small arrays needed for detection, 25-100 meters inter-station distance
• Attenuation: near-field infrasound indication for point source far-field cylindrical spreading
Detection and parameter estimation results
Either high apparent velocity and large azimuthal deviation or low apparent velocity and small azimuthal deviation
What propagation path allows 0.36 km/s?
Non-tropospheric velocity of 420 m/s between DBN and DIA
Head wave like propagation in high velocity acoustic channel
Strong winds cause high propagation velocity, large azimuthal deviations and steep incident angles
Raytracing with NRL-G2S models