Infrasound from the 2004-2005 Earthquakes and Tsunamis near Sumatra
From poster presented at 2005 IRIS Workshop, Stevenson, WA and presentation at Acoustical Society of America 2005 meeting in Vancouver. The ASA page also includes a sound file with the seismic, hydroacoustic, and infrasonic arrivals observed at Diego Garcia.
Introduction
Multiple infrasound arrays in the Pacific and Indian Oceans that are part of the International Monitoring System (IMS) observed three distinct waveform signatures associated with the December 26, 2004 Aceh, Sumatra, earthquake and tsunami. The infrasound stations observed (1) seismic arrivals (P, S and surface) from the earthquake, (2) T-phases, propagated along the SOFAR channel in the ocean and coupled back to the ground, (3) infrasonic arrivals associated with either the tsunami generation mechanism or the motion of the ground above sea level, and (4) deep infrasound corresponding to the propagation of the tsunami into the Bay of Bengal. All signals were recorded by the pressure sensors in the arrays. The seismic and T-phase recordings are due to the sensitivity of the MB2000 microbarometers to ground vibration, whereas the infrasound arrivals correspond to dispersed acoustic waves propagated through atmospheric waveguides. It appears that the arrival of the tsunami, as well as oceanic infragravity waves following the tsunami, were not observed by the infrasound stations. A similar (but not identical) sequence of arrivals was observed at Diego Garcia during the March 28, 2005 Nias earthquake and the April 10, 2005 Mentawai earthquake, suggesting that ground motion efficiently generates infrasound in the Sumatra region. We show the prominent features of the arrivals, present infrasonic source location estimates, and consider whether infrasound may be used in conjunction with other technologies as a discriminant for tsunami genesis.
Telesonic spectra
We analyzed infrasound data from stations in Diego Garcia (I52GB), Palau (I39PW), Madagascar (I33MG), and Kenya (I32KY) for the aforementioned Sumatra earthquakes (Figures 1 and 2). Figure 3 presents spectrograms for the December 26, 2004, March 28, 2005, and April 10, 2005 earthquakes, showing the seismic, T-phase, and infrasonic arrivals recorded by the infrasound stations in Diego Garcia, Palau, Madagascar, and Kenya.
Figure 1. Regional map showing locations for the December 26 2004 and March 28 2005 earthquakes. Figure from Lamont-Doherty Cooperative Seismographic Network (LCSN) http://www.ldeo.columbia.edu/LCSN/
Figure 2. Detail map showing locations and magnitudes of the four Sumatra earthquakes discussed in this study (click to enlarge).
Figure 3. Click to enlarge image - it would be very hard to read otherwise. Spectrograms recorded at a) I52, Diego Garcia; b) I39, Palau; c) I33, Madagascar; d) I32, Kenya. Left-hand columns show spectrograms from 0.2-4 Hz, right-hand columns show spectrograms from 0.02-0.16 Hz. Spectrograms are aligned by seismic arrival times whenever possible.
Figure 3a closeup, click each panel to enlarge.
The April 10 seismic event consisted of two distinct earthquakes at 10:29:11 (Mb 6.7) and 11:14:19 (Mb 6.5), and although seismic signals were not observed by the microphones, they may have produced a faint T-phase arrival and two clear infrasound arrivals at the Diego Garcia station. At this range, most of the seismic (Aceh and Nias) and infrasound energy lies below 2 Hz, whereas the T-phase energy is above 1 Hz. In the higher frequency band, the infrasonic signal duration is on the order of 1 hour for the Aceh and Nias event, and much shorter for the Mentawai temblors. The Diego Garcia station was the only one to record any infrasound from the Mentawai event. In the low frequency band, and particularly below 0.1 Hz, the infrasound from the Aceh event lasts for over four hours, whereas infrasound from the Nias event is less energetic and after ~1 hour is masked by low-frequency noise attributed to weather. No seismic or infrasound arrivals were produced in the deep infrasound range for the Mentewai event. The Palau station (range of ~4300 km) recorded clear high-frequency infrasound for both Aceh and Nias. At the low frequencies, weather masked Aceh, but Nias produced a clear arrival. The Madagascar station has an arrival in the high-frequency band associated to the Nias event, and the Kenya station has a brief arrival in the low frequency band associated with the Aceh event. In order to optimize the azimuthal and temporal resolution of the array data and capitalize on the spectral diversity of the observed signals, we used the Progressive Multi-Channel Correlation method (Cansi, 1995) in separate shallow (0.5-4 Hz) and deep (0.02-0.16 Hz) infrasound bands to analyze the station data.
Infrasonic Source Distributions
Using the Ground to Space (G2S) atmospheric profiles (Drob et al., 2003) specific to the station location and time of the event, source locations (Garces et al., 2004) were estimated for the high-frequency (0.5-4 Hz) infrasonic arrivals associated with the Aceh and Nias earthquakes and tsunamis (although there was a tsunami associated with the March 28 event, it was small and did not pose a substantial threat). The nearest stations to the epicenters (Palau and Diego Garcia) provided the best data for both events, and Madagascar was also used for Nias. Figure 4 shows four candidate infrasonic locations for the Aceh event, two near the epicenter/tsunami source and two more near the highest mountains in Aceh. The Nias event yielded similar locations further south (Figure 5).
Figure 4. Map showing infrasound-based locations calculated for the Aceh earthquake/tsunami in the 0.5-4 Hz band. Different locations correspond to the use of different phases for the location.
Figure 5. Map showing infrasound-based locations calculated for the Nias earthquake in the 0.5-4 Hz band. Different locations correspond to the use of different sets of stations for the location.
Although the radiation of infrasound from large continental earthquakes is established (e.g. Le Pichon et al., 2003), our results strongly suggest that islands shaken during submarine earthquakes can produce infrasound. More intriguing is the possibility that the tsunami genesis process may be capable of radiating infrasound, as suggested by the second possible infrasonic source location near the epicenter. The arrivals in the deep infrasound suggest a unique interpretation for the Aceh tsunami. At Diego Garcia, we observed coherent infrasonic energy arriving for over 4 hours, with a clear backazimuth trend (Figure 6), first southwards towards the northern tip of Sumatra and then steadily swinging northwards towards and across the Bay of Bengal, ending at the southernmost tip of India and spanning an angular range of over 40 ° (Figure 7) .
Figure 6. Arrival azimuth and apparent horizontal phase velocity for the Aceh event as recorded in the deep infrasound at I52, Diego Garcia.
Figure 7. Map showing overall range of detection azimuths in the deep infrasound band for the Aceh event at I52, Diego Garcia.
Le Pichon et al. (2005) followed the fault rupture and the propagation of the tsunami into the Bay of Bengal, attributing the infrasound arrivals to the interaction of the tidal wave with the shoreline. Alternately, the source process may involve multiple reflections of the tsunami energy from steep bathymetry, leading to resonant triad interactions as encountered in open-sea swells (Willis et al., 2004). In contrast, for the Nias event the arrival backazimuths from Diego are steady, whereas the Palau arrivals last over three hours and swerve from south to north, spanning a range of ~30°. It is unfortunate that the deep infrasound range at Palau was dominated by weather noise during the Aceh event, so no comparison is possible to the Nias observations.
Concluding Remarks
From the prominent features of infrasonic arrivals and infrasonic source location estimates for the selected Sumatra earthquake and tsunami sequence, we deduce that:
1. Submarine earthquakes can produce infrasound. The sound may be radiated by the vibration of the ocean surface or the vibration of land masses near the epicenter.
2. Infrasound stations can also serve as seismic and T-phase detectors for large events.
3. Observed signals are substantially different for the three events, which may be due to source or propagation effects.
4. Small and large tsunamis may both produce infrasound. The source process is not understood.
5. There is a substantial difference between the information contained in the lower and upper frequency bands of the infrasound range.
The candidate source location near the epicenter, in conjunction to the unique signal observed at Diego Garcia for the Aceh event, suggests that infrasound may be used in conjunction with other technologies as a discriminant for tsunami genesis. Fundamental research is needed on how large earthquakes and tsunamis can produce low-frequency sound.
References
USGS Earthquake Hazards Program, http://earthquake.usgs.gov/ . The authors are aware of revised magnitude estimates for these events, which are outside the scope of this paper.
Cansi, Y, An automatic seismic event processing for detection and location: The P.M.C.C. method, Geophys. Res. Lett., 22 , 1021-1024, 1995.
Drob, D., M. Picone, and M. Garces, The Global Morphology of Infrasound Path Partitioning, J. Geophys. Res. , 108, 4680, 2003.
Garces, M., P. Caron, C. Hetzer, A. Le Pichon, H. Bass, D. Drob, J. Bhattacharyya, Can infrasound help with tsunami monitoring?. Submitted to EOS, 2005.
Garcés, M, P. Caron, and C. Hetzer, (2005a), Infrasound Associated with the
2004 Sumatra Megathrust Earthquake and Tsunami, ASA/CAA 05 Meeting, Vancouver , BC.
Garces, M, C. Hetzer, H. Bass, M. Hedlin, K. Lindquist, J. Olson, C. Wilson, D. Drob, and M. Picone, Forensic studies of infrasound from massive hypersonic sources, EOS , 85, no 43, 433,440, 2004.
Hedlin, M.A.H., Garces, M., Bass, H., Hayward, C., Herrin, G., Olson, J.V. and Wilson, C., Listening to the Secret Sounds of Earth's Atmosphere, EOS , 83, 557, 564-565, 2002.
Le Pichon, A., P. Herry, P. Mialle, J. Vergoz, N. Brachet, M. Garcés, D. Drob, L. Ceranna, Infrasound associated with large Sumatra earthquakes and tsunami. Submitted to GRL, 2005.
Le Pichon A., J. Guilbert, M. Vallée, J.X. Dessa and M. Ulziibat, Infrasonic imaging of the Kunlun Mountains during the great 2001 China earthquake, Geophys. Res. Lett ., 1029/2003GL017581, 2003.
Le Pichon, A., Y. Cansi, J. Guilbert, J. Vergoz, N. Brachet, Analysis of infrasonic waves generated by the rupture of the Sumatra Mw=9.0 in the Indian ocean, Geophysical Research Abstracts , Vol. 7, 08944, 2005.
Willis, M., M. Garcés, C. Hetzer, and S. Businger, Infrasonic observations of open ocean swells in the Pacific: Deciphering the song of the sea, Geophys. Res. Lett. , 31, L19303, 2004.