The Woomera Infrasound

1
The Woomera Infrasound and Seismic Experiment
David Brown1 Clive Collins1 Brian Kennett2 1.
Geoscience Australia 2. Australian National
University
Geological Survey of Canada Australian Dept. of
Defence
1
2
Introduction

• The Woomera Experiment
• An international explosives trial held at
  Woomera, South Australia, in September, October
  2002
• Conducted for defence research by
• UK Ministry of Defence
• Australian Department of Defence
• Purpose
• ammunition stewardship program
• Participation by
• Canada
• Netherlands
• Norway
• USA
• Singapore
• Two explosions
• 27000 kg ammunition

IS07
2
3
Introduction
Flinders Ranges
1165m
167m
900 m
315m
3
4
The Source
Source Information
4
5
The Source
Explosives Container
Instruments, tamping
Earthen walls 3 sides Concrete roof
6
The Source
27000 kg shot
Shock wave

• Some statistics
• for a block of rubble
• T 33.57 sec
• 508 m
• V 165.2 m/s
• H 1380.5 m
• q 84.7 deg

Shock wave
7
The Source
Norwegian house
Australian house
7
8
Receiver Information
Seismic Receiver Information
Infrasound Receiver Information
8
9
Receiver Information
Spectrogram 20,000 lb Chemical Explosion _at_ 250
km 180 deg. backazimuth

• Broad-band signal
• signal duration 0.6 min
• dominant period 1.6 sec
• drop-off frequency
• 6db down gt 2.0 Hz
• 12 db down gt 2.0 Hz
• Fstat
• _at_ max SNR 4.3
• _at_Max Power 21.5
• Frequency
• _at_max SNR 2.0 Hz
• _at_max Power 0.47 Hz
• Need to design a 3-sensor array with maximum
  capability around 0.5 Hz. ie, average intersensor
  spacing of around 330 m.

10
Receiver Information
Broken Hill Station
Finite-frequency array response
10
11
Receiver Information
Broken Hill Station
11
12
Receiver Information
Finite-frequency array response
Oodnadatta Station
12
13
Receiver Information
Oodnadatta Station
13
14
Receiver Information
Finite-frequency array response
IS07 Station
14
15
Receiver Information
15
16
The Signals
Broken Hill Station 27000 kg
16
17
The Signals
Broken Hill Station 27000 kg
0.8 3.2 Hz
17
18
The Signals
Broken Hill Station 27000 kg
342 m/s 289 deg
354 m/s 290 deg
348 m/s 290 deg
454 m/s 256 deg
477 m/s 257 deg
490 m/s 241 deg
18
19
The Signals
IS07 27000 kg
352 m/s 167 deg
19
20
The Signals
Oodnadatta Station 27000 kg
0.8 3.2 Hz
20
21
The Signals
Broken Hill Station 5000 kg
0.4 1.6 Hz
21
22
GA Infrasonic Processing for CTBT monitoring

• Infrasonic Processing at Geoscience Australia
• Will process 5 IMS stations IS03, IS04, IS05,
  IS06, IS07
• Will observe the following processing philosophy

Single station processing Seeking significant
signals on individual stations
Automatic internal alert notification
duration gt 2 minutes Fstat gt 10.0 SNR gt 1.5
57 days 38185 detects 98 arrivals
INFER Automatic detector
DISCIN Simple discrimination criteria
detection
arrival
Interactive Analysis review
Interactive Analysis scanning
DISCEX High-level discrimination procedures
arrival
1 arrival
Manual external alert notification
22
23
GA Infrasonic Processing for CTBT monitoring

• Signal parameter estimation at sparse arrays
• Global minimisation of the misfit between
  theoretical and stacked beam powers
• theoretical side-lobe pattern will be
  imprinted on the stacked beam power

Define broad-band theoretical array response to be
Define broad-band stacked beam-power to be
Define the misfit function using an norm
as
Use the Sambridge Neighbourhood Algorithm to
converge to the region of best fit.
24
GA Infrasonic Processing for CTBT monitoring
Signal Reception and Detection Sparse Arrays

• Array response
• Beam power as a function of slowness for I07AU
  array
• Zero slowness
• finite frequency signal centred at 0.875 Hz
• 9 x 9 81 beams were used
• maximum beam power centred on zero slowness

Sy (s/km)
Sx (s/km)
25
GA Infrasonic Processing for CTBT monitoring
Signal Reception and Detection Sparse Arrays
Time (sec)
N
2880
1440
2160
0
720
3600
Synthetic implant, azimuth, 128 deg.
W
E

• Array response
• Beam power as a function of slowness for I07AU
  array
• with implanted signal
• 9 x 9 81 beams were used
• 0.875 Hz
• maximum beam power centred on slowness
• corresponding to the implant azimuth (128 deg)

Slowness plane
S
26
GA Infrasonic Processing for CTBT monitoring
Signal Reception and Detection Sparse Arrays
misfit surface
27
GA Infrasonic Processing for CTBT monitoring
Signal Reception and Detection Sparse Arrays
N
N
W
W
E
minus
E
N
S
S
81 beams 1.5 deg accuracy in azimuth
BSSA, 2003, Vol 93 p.1765-1772
equals
E
W
S
28
GA Infrasonic Processing for CTBT monitoring
Network processing Infrasonic source location

• Determine the Great-circle intersection point
• Determine the travel-times to receivers using a
  two-value constant velocity model
• 290 m/s (stratospheric propagation)
• 240 m/s (thermospheric propagation)
• Determine if predicted travel-times match
  observed arrival times for both stations (up to
  some variance)
• Use uncertainty in measured azimuth to determine
  uncertainty in predicted source location

INFRA_LOC_0 automatic 2 station source location
detection
origin
INFRA_LOC_1 automatic source location (refinement
level 1)
origin
gt 2 stations
Updated travel-time Information. GT information
for testing Data Fusion
INFRA_LOC_2 automatic source location (refinement
level 2)
origin
28
29
Summary

• The Woomera Infrasound and Seismic experiment
• Characterised by unusually fast travel-times
• May be timing errors (unlikely)
• Characterised by significant off-great circle
  azimuths
• may be side-lobe detections (unlikely)
• May be acoustic signal generated by the seismic
  interaction with the Flinders Ranges (?)
• GA Infrasonic Processing
• Seeks significant signals on individual stations
  first
• Basic set of discrimination criteria for
  automatic alert notification
• Duration, Coherence, Energy
• Performs 2-station source location using constant
  velocity model
• Will experiment with the Kennett procedure for
  sparse array processing
• May help the spatial aliasing problem at sparse
  arrays

29