Bottom Simulating Reflections As Regional Gas Hydrate Proxies 41 Nature of BSRs

BSRs are the most widely used indicator for gas hydrates in marine sediments. They are located at the base of the hydrate stability zone (Chapter 4), where pressure and temperature conditions are at the phase boundary between hydrates and free gas. Because gas hydrate stability is much more sensitive to temperature than to pressure, BSRs approximately follow isotherms. In undisturbed sediments, isotherms are parallel to the seafloor, which is why BSRs mimic the seafloor (Fig. 4).

The negative polarity of BSRs indicates a Vp decrease across the stability boundary. In principle, this contrast may be caused by elevated Vp in gas-hydrate-bearing sediments above the base of the gas hydrate stability zone (BGHS) (Hyndman and Spence, 1992) and/or by low velocities in gas-charged layers beneath it (Korenaga et al., 1997). Results from various studies indicate that, with some possible exceptions, strong BSRs are principally caused by a drastic decrease of Vp due to free gas, with a relatively small contribution from an overlying "hydrate wedge" (Korenaga et al., 1997). At some locations, however, weaker BSRs may be caused by gas hydrates without underlying free gas (Minshull et al., 1994; Pecher et al., 1996b) (Fig. 4b). Important evidence

W LOWER SLOPE LIMA BASIN E

W LOWER SLOPE LIMA BASIN E

Fig. 4: (a) BSR off Peru. Structural interpretation according to von Huene et al. (1996). TWT: two-way traveltime. The shallowing of the BSR toward the continent is caused primarily by a decrease of pressure and the resulting move of the phase boundary toward lower temperatures, i.e., closer to the seafloor (an additional effect comes from an observed regional increase of heatflow). Note the weakening of the BSR close to its landward termination, which coincides with normal faults. The normal faults may act as conduits that allow gas from the BSR to escape, (b) Velocity-depth profiles from waveform inversion across the BSR. The right profile at the location where the BSR is weak may indicate that the BSR is mainly caused by elevated Vp above the BGHS, perhaps related to gas hydrates. The other two profiles suggest a thin free gas layer beneath the BSR

Fig. 4: (a) BSR off Peru. Structural interpretation according to von Huene et al. (1996). TWT: two-way traveltime. The shallowing of the BSR toward the continent is caused primarily by a decrease of pressure and the resulting move of the phase boundary toward lower temperatures, i.e., closer to the seafloor (an additional effect comes from an observed regional increase of heatflow). Note the weakening of the BSR close to its landward termination, which coincides with normal faults. The normal faults may act as conduits that allow gas from the BSR to escape, (b) Velocity-depth profiles from waveform inversion across the BSR. The right profile at the location where the BSR is weak may indicate that the BSR is mainly caused by elevated Vp above the BGHS, perhaps related to gas hydrates. The other two profiles suggest a thin free gas layer beneath the BSR

for the nature of the BSR comes from waveform inversion, which optimizes a subsurface velocity model by comparing synthetic seismograms to observed data. Waveform inversion is necessary to elucidate the details of the velocity structure around BSRs, because those structures appear to be thinner than a seismic wavelength. The BSR waveform is therefore a complex superposition of reflected signals from the top and base of the layer, which is impossible to model with oversimplified AVO models (e.g., compare the results of Katzman et al. (1994) to those of Korenaga et al. (1997) in the same area).

A major question regarding the operation of the hydrate/gas system is the thickness, origin, and dynamics of the free gas zone. VSPs conducted during Ocean Drilling Program Leg 164 on the Blake Ridge revealed that the hydrate zone there is underlain by a surprisingly thick gas zone (at least 250-m thick) (Holbrook et al., 1996). The free gas zone is characterized by a region of unusually low Vp (Fig. 5) that coincides with a band of high reflectivity that follows stratigraphie layers. This can be explained by slight variations in gas concentrations across layer boundaries, since Vp is very sensitive to gas at low

line 1

Fig. 5: Reflection seismic section (left) and Vp from VSPs on the Blake Ridge. The BSR at 995 and 997 is at -440 mbsf. The only reasonable explanation for the low Vp beneath the BSR is a thick free gas zone. Note the zone of high reflectivity beneath the BSR. This is probably caused by slight variations of free gas concentration across layer boundaries. After Holbrook et al. (1996).

Sites 994/995/997

1400 1600 1800 Velocity (m/s)

line 1

Sites 994/995/997

1400 1600 1800 Velocity (m/s)

Fig. 5: Reflection seismic section (left) and Vp from VSPs on the Blake Ridge. The BSR at 995 and 997 is at -440 mbsf. The only reasonable explanation for the low Vp beneath the BSR is a thick free gas zone. Note the zone of high reflectivity beneath the BSR. This is probably caused by slight variations of free gas concentration across layer boundaries. After Holbrook et al. (1996).

concentration (Fig. 6). A similar band of high reflectivity has been reported recently offshore Pakistan (Grevemeyer et al., 2000). Drilling on the Cascadia margin (MacKay et al., 1994) and off Chile (Bangs et al., 1993), on the other hand, showed that the gas zones there are seismically thin (off Cascadia, earlier results from waveform inversion (Singh et al., 1993) were confirmed). The cause of these large differences in the thickness of gas zone are not understood; however, sediment permeability may play a key role.

The character of BSR reflections depends on the frequency content of the data. Recently, BSR reflection strength on the Cascadia margin has been shown to be frequency-dependent, with weaker reflection amplitudes for higher signal frequencies (Yuan et al., 1996). This may be caused by a steep velocity gradient at the BGHS, rather than a first-order step: low-frequency signals "see" a step function, thus generating a strong reflection, while higher-frequency signals see the gradient, leading to weaker reflections. Importantly, DTAGS data, which have unusually high lateral and vertical resolution, show that the BSR on the Blake Ridge is not the continuous reflection observed in conventional surface-towed data, but is rather a set of "shingled" reflections along dipping layers that terminate at the BGHS (Fig. 3). The lower resolution of surface-towed seismic data smears out these shingles, generating a smooth, continuous reflection (Wood and Ruppel, 2000). Thus the very characteristic that distinguishes the BSR as a phase-boundary reflection ~ the cross-cutting of strata reflections - is itself an artifact of the relatively low frequencies of surface seismic data.

1600

& 1200

Vp o 800

Was this article helpful?

0 0

Post a comment