Nabarup Ganguly- Unconventional Seismic Events on a Subduction Zone

Detecting unconventional seismic signals related to subduction zone processes at depth in continuous ocean bottom seismometer (OBS) records requires the analysis and identification of noise due to instrumental problems, deployment sites or sea state conditions. The temporary OBS deployment at the Lesser Antilles subduction zone provides new insights into the feasibility of detecting unconventional signals such as non volcanic tremor (NVT), long-period (LP) or ultra-long period (ULP) events. Analysis of noise at an array comprising several sites and types of instruments and comparison with recordings on land shows transients in the noise. Episodes can be identified considering the diversity of sites and instrument types and comparing the seismic signals with meteorological and oceanographic data. In order to reliably detect NVT (1–10 Hz) originating from inside the solid Earth, one must first characterize noise induced by the activity of the atmosphere and hydrosphere at the sea-bottom as well as on land. The semidiurnal modulation of noise amplitude can be shown here not to be due to that of the NVT from a seismic source at depth which is related to the subduction interplate and whose activity is modulated by the tidal stresses as inferred for other megathrusts on emerged forearcs. Here, the semidiurnal modulation is rather due to the effect of the tides themselves, such as tidal currents, since they do not affect all types and all components of the unique multi-station array of OBS that could be deployed on this submerged forearc. The short period cut-off of the strong noise due to ocean surface infragravity waves increases to longer periods with OBS depth, thereby increasing the observational window with low noise to lower frequencies, and deep OBS sites may be advantageous for detecting LP events.

Full-size image (161 K)
Fig. 1. Location map of the OBS network deployed at the Lesser Antilles subduction zone beginning of 2007 offshore Martinique, Dominica, Guadeloupe and Antigua Islands. Orange squares are the 3-component BB-OBS equipped with Trillium 240 s seismometers and pressure sensors. Blue squares are the 3-component WB-OBS equipped with CMG40T seismometers and broadband hydrophones. Green squares are 3-component PMD Scientific wideband and short period hydrophones. Orange, green, magenta and blue circles represent the 3-component short-period seismometers and hydrophones with color indicating the pool (see color caption). For the GeoAzur seismometers, a color distinction is made between the first deployment (D1) from January to April 2007 and the second deployment (D2) from May to August 2007. Triangles are the OBS with 1-component short-period seismometer and hydrophone. A red square marks the site of FDF broadband land station of the GEOSCOPE global network on the Martinique Island. The black frame indicates the area of the present study.
Full-size image (59 K)
Fig. 2. From Julian day 151 to 175, 2007, a variation of vertical velocity RMS amplitude after band pass filtering between 1 and 8 Hz in 5-min windows of WB-OBS at 5 sites over (maximum inter-station distance 80 km), as labeled in Fig. 1 An episode of high noise lasting several days is suggested, which would have a semi-diurnal modulation of its amplitude and a shift in time among sites.
Full-size image (151 K)
Fig. 3. One-day waveforms with a different filtering applied (Julian day 63). The two upper traces of each frame represent the vertical and horizontal components of WB-OBS at site G4 and the two lower traces represent the vertical and horizontal components of BB-OBS at site J6. (a) 2–8 Hz band-pass-filtered, (b) unfiltered and (c) 0.005–0.05 Hz (200–20 s) band-pass filtered. Note a high amplitude noise signal with semi-diurnal repetition, except for BB-OBS vertical component and for its horizontal component with the 2–8 Hz filter applied.
Full-size image (172 K)
Fig. 4. Ground velocity spectrogram for BB-OBS at site J6. a) vertical and b) horizontal components from days 53 to 76 computed between 0.0035 and 0.45 Hz. Superimposed on b), theoretical oceanic tides, which have been calculated for a point near J6 (point 16.5°N 60.5°W) with the SPOTL software (Agnew, 1997). PM, DF and IGW in panel a stand for primary, double frequency microseisms and infragravity waves respectively.
Full-size image (86 K)
Fig. 5. One-day spectra (PSD, power spectral density) with shading identifying the domains where NVT, LP, and ULP seismic events have been reported from seismometers on emerged subduction forearcs. a) PSD computed for the BB-OBS at site H3 in red, for the WB-OBS at site G4 (CMG40T) in gray, and for the broadband observatory GEOSCOPE station FDF (STS2) in blue. PSD have been estimated for the between 1000 s period and 8 Hz for the vertical and horizontal components. Julian day 53. Twenty-four hour data were used for the calculation, and there was no significant earthquake during this day. Dotted black lines are the low-and high-noise models of Peterson (1993) for permanent land stations. Before estimating the PSD, we removed the mean and the instrument response and transformed to acceleration. PSD have then been smoothed using a moving average. IGW stands for infragravity waves b) pressure spectra of the broadband hydrophone in red (G4) and DPG in gray (H3). c) Coherence computed for site H3 between the vertical seismic component and the pressure from DPG.
Full-size image (68 K)
Fig. 6. 1–8 Hz band-pass-filtered vertical component RMS signal amplitude over a fortnight, 5-min windows with no overlap have been used for the RMS amplitude calculation (a) WB-OBS at site G3, (b) BB-OBS at site H3 and (c) FDF on-land seismograph. The 1–8 Hz filtered traces are compared to (d) wind speed variation computed at point 16.5°N 60.5°W. 0.1–20 Hz band-pass-filtered vertical component RMS signal amplitude for the same data, e) WB-OBS at site G3 and f) BB-OBS at site H3g) FDF. Filtered traces are compared to h) SWH, significant wave height at the same point as d). In d) and h), data are from the Wave Watch III model provided by the National Oceanic and Atmospheric Administration (NOAA) Environmental Modeling Center.
Full-size image (310 K)

Fig. 7. Vertical velocity component spectrogram computed for a 1.5-month period in the 1 to 10 Hz frequency band. (a) Short-period SP-OBS 105 at site G6 equipped with a 4.5 Hz sensor, (b) wide-band WB-OBS a site G3 equipped with a Güralp CMG40T sensor, (c) broadband BB-OBS at site H3 (Trillium 240 s) and (d) vertical channel of the permanent STS2 broadband station Geoscope station FDF. (e) Wind speed in m/s (same source as for Fig. 5, at a point in the OBS array, 16.5°N and 60.5°W).

This entry was posted in Seismicity by Nabarup Ganguly. Bookmark the permalink.

About Nabarup Ganguly

Associate Disciple-Educator, M.Sc. in (Geography & Disaster Management), M.A. in Education, B.Ed, First Class First, Rank holder, Gold Medallist, Author & Life Member; Guide and Counsellor, Inventor (Tripura State Council For Science & Technology), Geographer, Department of Geography, RIO+20, Brazil, South America

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s