Bichromatic infrasonic tremor from compact sources at Puu Oo

Document Sample
Bichromatic infrasonic tremor from compact sources at Puu Oo Powered By Docstoc
					Bichromatic infrasonic tremor from
compact sources at Puu Oo, Hawaii

 M. Garces, D. Fee, S. McNamara, and W. Bortz
 Infrasound Laboratory, University of Hawaii, Manoa

 R. Hoblitt and F. Trusdell
 USGS- HVO


 2006 Cities on Volcanoes, Quito, Ecuador
Overview




 •Observed diurnal variability of infrasonic tremor
 •Most signals originate from active vent
 •Tremor amplitude and detection affected by
 atmospheric conditions                               Photo courtesy of HVO
Previous infrasound work at Kilauea: Even apparently
steady processes can generate sound. Sustained
degassing? Lava falls inside conduits?




                        2002 Deployment - Multiple technologies
Array
location
• Volcano, Hawaii
• Array deployed on
Frank Trusdell’s
backyard
• ~12.5 km NW of crater
complex
• Deployed August 18th-
27th 2005
Instrumentation

   • 24-bit Geotech digitizer sampled at 20 Hz
   • Four Chaparral 2.2 microphones on high gain (2.0 V/Pa)
   • 60 m aperture triangular array with a central element
   • Dense tropical forest with porous hoses
   • Wind direction and speed within forest are not representative of the
   region’s wind patterns
       –Ground truth observations
       –Long wavelength turbulence penetrates forest
       –Data below ~0.3 Hz proxy for wind
Stacked and Filtered Waveforms
                                                                 •Aligned, stacked,
                                                                 filtered between 0.5-
                                                                 3 Hz
                                                                 •Very good
                                                                 correlation between
                                                                 waveforms
                                                                 •Consistent during
                                                                 low noise periods
                                                                 •Correlation
                                                                 deteriorates during
                                                                 high wind



  The pressure amplitudes 1 m from the vent would be ~260Pa, or 142 dB re 20 µPa. In
   the audible range, this would be commensurate to the sound level of a jet at takeoff.
Pu’u O’o Spectral Content   •Continuous harmonic
                            tremor from vent
                            •0.5-6 Hz
     Spectrogram    PSD
                             - Infrasonic tremor
                             - Increase in spectral
                            amplitude during night
                            • 0.02-.3 Hz
                              -Proxy for wind speed
                              -increase in spectral
                            amplitude during day

                            •Winds increase during
                            day raise noise and
                            lower signal

                            •Winds decrease during
                            night lower noise and
                            raise signal!
Array Detections
                   •Window length
                   60 sec, 90%
                   overlap
                   •0.5 – 6 Hz, 10
                   bands
                   •Detection: pixel
                   in time-freq
                   space

                   •~90% originate
                   from near crater
                   complex
PMCC Results 8/21/005 0000-2400
                                  •Diurnal

                                  variations

                                  •Very similar

                                  azimuths/speeds
Other Sources           Log10 Plot of PMCC Detections



 Plot of detections
 on Log10 scale
 •Skylights
    -lava falls?
    -lava-gas
    excitation within
    tube
 •Surf
 •Helicopters, planes
Conclusions

  • Diurnal changes in infrasonic tremor spectral amplitude and array
  detections
      –Source process change over time…so do source recordings
  • Atmospheric changes can significantly affect the propagation of
  infrasound
      –Tropospheric winds affect ambient noise levels, S/N, and
      detectability.
      –“Shadow zones” not stable
      –Boundary layer variability. Topography.
  • Proper site and instrument selection and recognition of the
  regional wind patterns is necessary to permit robust infrasonic
  remote sensing of volcanic processes
                                                             Photo courtesy of HVO
Summary of the Workshop on Acoustic
Remote Sensing of Volcanoes
Sponsored by:
Geological Survey of Canada
National Science Foundation
International Association of Volcanology and Chemistry of the
Earth's Interior
Instituto Geofisico, Escuela Politécnica Nacional, Quito, Ecuador
National Center for Physical Acoustics, University of Mississippi
Infrasound Laboratory, University of Hawaii
Thirty Participants
Agenda
Principles of Infrasound and Volcano Acoustics
Sensor Technology and Array Design
Data Acquisition Systems and Formats
Signal Processing
Signal, Noise, and Clutter
Canadian and US Prototype Operational System: ASHE
French and Australian Volcano Monitoring Studies
Science of Volcano Acoustics, and Infrasonic Signals from Ecuador, St.
    Helens, Kilauea, and Other Volcanoes. Selected Contributions from
    Participants.
                 ASHE - Acoustic Surveillance
                    for Hazardous Eruptions
                                  David McCormack, Geological Survey of Canada
                                       Henry Bass, NCPA University of Mississippi
                                               Milton Garces, University of Hawaii
                                                            Michael Hedlin, UCSD
                          Hugo Yepes, Escuela Politecnica Nacional, Quito, Ecuador


•Requests from ICAO, FAA
•Has the potential to continuously monitor relatively cheaply over
hundreds of kilometres
•Is now a mature, proven technology suitable for deployment in
remote areas
•Aviation community would like ‘5 minute warning’. There is a
direct tradeoff between spacing of arrays and latency of warning.
•‘Natural’ technology for direct measurement of atmospheric
perturbation
ASHE - Acoustic Surveillance
   for Hazardous Eruptions
               4-element infrasound array,
               2C broadband seismometer,
               Meteorological station
               Real-time satellite link

               3C broadband seismometer,
               Meteorological station
               Real-time satellite link



               Real-Time Data sent to
                 Geological Survey
                     of Canada
      Monitoring of Indonesian                                               3
     volcanoes with infrasound :
         preliminary results                              1
         SURvey of Volcanic Arc of Indonesia                         2                4
J. Guilbert1, P. Harjadi2, M. Purbawinata3, S. Jammes1,
               A. Le Pichon1, B. Feignier1
                       CEA/DASE



                                                                   Detections obtained
                                                    Pinatubo V.?   during the month of
                       Cendres V.?
                                                1                  March 2005
                                        4
                                                                   Clear and continuous
                                                                   infrasound detections
                                                                   are identified from
                                   3
                                            2                      Pinatubo, Semeru,
   Anak Krakatau V.                                                Krakatau, Talang and
                                            Semeru-Bromo V.        Barren Island volcanoes
                                                                   demonstrating the
                                                                   potential of infrasound
                                                                   technology for volcano
                                                                   monitoring
Multi-Technology observations of the Manam PNG
Volcano Eruption, January 2005: Potential for an
Enhanced Volcanic Ash Warning System
David Brown, Andrew Tupper, Doug Christie, Ima Itikarai



                                                 Dec 2005:
                                                      3 operational IMS stations,
     2643 km                                          sending to Vienna
                      2116 km
                                                      1 real time to Canberra (IS05)
                                    2600 km
                2600 km
2887 km                                          March 2006:
                                       3618 km       3 real time to Canberra
               IS07       3618 km
     IS07
                                                 Automatic processing in Canberra:
                                                     detection list
                                                     source location
  IS04
                  IS05
                                IS05
Detection of Infrasound from Manam Volcano, PNG
Paola Campus1, Douglas R. Christie and David J. Brown, Paola.Campus@ctbto.org




              Infrasonic signal detected    Infrasonic signal not detected

            Manam Subplinian/Plinian Eruption,
    January 27th, 2005 - Height of ash plume: 21 to 24 km
Manam, January 27th, 2005
Highlights:
The Important contribution IMS infrasound can make to
volcano monitoring:
   • Manam seismic observatory operated by RVO was destroyed,
     including some loss of life.
   • Substantial cloud cover on January 27 made satellite analysis
     difficult.
   • Consequently, an ash advisory was not issued for 14 hours


The Important issue of accessing and processing IMS
infrasound data in real time
   • IS07 processing could have issued alert 12 hours earlier
     (real-time data feed to Canberra was not completed at the time)