Deep Towed Seismic Detailed Structure at the BSR

High resolution multichannel seismic data were acquired in 1997 using the Deep-Tow Acoustics/Geophysics System (DTAGS) in the vicinity of ODP Site 889/890 (Gettrust et al., 1999). The system was towed at a depth of about 1000 m in water depths averaging 1400 m. Because of its proximity to the seafloor, wide-angle reflection data may be recorded on the DTAGS streamer which has a relatively short length of 622 m. The wide angles facilitate velocity estimation and amplitude-versus-offset analysis. However, to allow velocity analysis and coherent stacking of the data, careful geometry corrections for the varying depth of the streamer were required. Consequently, a method of estimating the depths of the source and hydrophones using sea-surface reflection times was developed (Walia and Hannay, 1999). The frequency band of DTAGS was 250-650 Hz, providing a vertical resolution of ~ 2 m. As well, the near seafloor source and receiver configuration reduces the Fresnel zone size to about 25 m, significantly less than the Fresnel zone width of-400 m for 30 Hz sea surface data.

4 km

Figure 7. (a) A stack of the near-offset channels of the DTAGS array over same area showing fine structure within the uppermost 200 m of sediments, (b) Portion of an unmigrated multichannel seismic lineover same area as the DTAGS line in (a). Vent 1 is associated with a seafloor pockmark (resulting in a diffraction hyperbola). Note that the sediment horizons change dip at each of the other marked blank zones, indicating the presence of faults.

IDTAGSI

IDTAGSI

4 km

Figure 7. (a) A stack of the near-offset channels of the DTAGS array over same area showing fine structure within the uppermost 200 m of sediments, (b) Portion of an unmigrated multichannel seismic lineover same area as the DTAGS line in (a). Vent 1 is associated with a seafloor pockmark (resulting in a diffraction hyperbola). Note that the sediment horizons change dip at each of the other marked blank zones, indicating the presence of faults.

In a stack of the near-offset DTAGS data, a high amplitude reflector about 10 m below the seafloor can be resolved (Fig. 7a). This near-surface reflector may indicate a carbonate pavement precipitated from upward-advecting fluid (Fink and Spence, 1999).

Combined with 5 previous surveys in the area, the DTAGS data enable us to determine the frequency dependence of hydrate-related reflectors over the broad band from 20-650 Hz (Chapman et al., 2000). The BSR is very strong for low frequency data, but its amplitudes become smaller for the higher frequencies and it is barely discernable at the peak frequency of the DTAGS source (Fig. 8). This behavior suggests that the BSR is produced not by a sharp impedance contrast but rather by a negative velocity gradient. The vertical scale of the gradient is much smaller than the wavelength of the low frequency data, but greater than the wavelength of the DTAGS data. We have modeled this frequency dependence of amplitude using synthetic seismograms, and we infer that the thickness of the velocity gradient layer at the BSR is 6-8 m, with velocity decreasing by 250 m/s (Fig. 8)

Frequency (Hz) Velocity (m/s)

Figure 8. Observed variation of reflection strength with frequency (solid dots). The shaded area gives constraints from DTAGS data. The dashed lines are the modeled variations of reflection coefficient with frequency, for different thicknesses ) Ad of the velocity gradient layer which represents the BSR.

Frequency (Hz) Velocity (m/s)

Figure 8. Observed variation of reflection strength with frequency (solid dots). The shaded area gives constraints from DTAGS data. The dashed lines are the modeled variations of reflection coefficient with frequency, for different thicknesses ) Ad of the velocity gradient layer which represents the BSR.

4.5 3D Seismic: Seafloor Venting and Possible Pathways for Fluid Flow

A 3D seismic survey was carried out in 1999 using a single 40 cu. in. airgun over a multichannel grid consisting of 40 lines spaced at 100 m plus 8 crosslines. A number of narrow vertical blank zones were observed on many of the seismic sections (Fig. 7b). The region of numerous blank zones was imaged at greater detail over a single channel grid with 25 m line spacings. Similar blank zones had been observed on some of the 1997 DTAGS seismic sections, as well as on high frequency data collected by the R.V. Sonne in 1996 (Zuellsdorf et al., 2000. It was suggested that these blank zones represent fluid expulsion channels.

The surface expression of Vent 1 (Fig. 7b) was determined in detail by 3.5 kHz sub-bottom profiling, which indicated that the structure was circular with a diameter of about 500 m. Migrated multichannel seismic data clearly show that this blank zone is associated with a seafloor pockmark, and that the polarity of the seafloor reflection changes phase within the pockmark. which indicates the presence of free gas just at or below the seafloor. Vent 1 also bounds a region of anomalously high reflectivity near and below the BSR, which is observed to some extent in Fig. 7b but is particularly prominent in the crossline direction. This high reflectivity can also be explained by the presence of free gas, trapped in sediments beneath the BSR. As observed in Fig. 7b (and in 3D time slices), most of the blank zones can be associated with changes in dip of the sediment horizons. These changes can be traced in 3D across the grid. The blank zones are thus interpreted as faults, which act as conduits for the focussed upward expulsion of fluids and methane gas.

Figure 9. Downhole electrical resistivities from Site 889/890 and the reference site 888. Resistivities of the recovered core samples are also shown.

5. ELECTRICAL RESISTIVITY OF HYDRATES

Electrical resistivity is an alternative parameter that is very sensitive to gas hydrate, because electrically resistive hydrate replaces conductive seawater in the pore spaces. Thus, resistivity measurements have potential as an alternative to seismic analyses for determining hydrate concentrations in deep sea boreholes and for seafloor field mapping of hydrate occurrences.

5.1 Resistivity from ODP Downhole Logs and Core Sample Measurements

At ODP sites off Vancouver Island where a BSR is present, log resistivities have average values of about 2.0 ohm m in the 100 m interval above the BSR (Hyndman et al. 1999; Fig. 9). The abrupt downward increase in log resistivity at 130 mbsf is similar in character to the increase in log velocity (Fig. 5b), and both occur at the base of slope basin sediments overlying accretionary basin sediments. Over the same 100 m interval, the log resistivities average 1.0 ohm m at a nearby deep sea reference site where no gas hydrate is present. As discussed in section 7.2, the difference in resistivity can be used to estimate hydrate concentrations.

Seafloor data 493 m . minus no-Hydrate model f-. /

Furthest array electrode

Model using ODP II \ Site 889 log resistivity 1 I L-*-" ' (30% hydrate saturation / —7 above BSR) \ J^l /

M

Jr\j y _____ — — — ~~

j/\l ^ 1saturation

//% / No hydrate reference

Figure 10. Results of towed seafloor EM system near Site 889. The observed phase differences (dots) best match those for a model in which resistivity above the BSR is 2.0 ohm m, just slightly less than that determined from the Site 889 resistivity logs.

Was this article helpful?

0 0

Post a comment