R/V/ Cape Hatteras cruise CH 18-95 utilized seismic methods to study of the gas hydrate deposits of the Blake Ridge region, offshore South Carolina, in conjunction with drilling of Site 994 and 995 of the Ocean Drilling Program (ODP) Leg 164. Gas hydrates are a solid structure in which ice forms a cage around a guest gas molecule (usually methane). Natural hydrates are stable in two environments: permafrost regions because of low temperatures (continental hydrates) and deep sea sediments because of high pressure (oceanic hydrates). The amount of oceanic hydrates is believed to exceed that of continental hydrates by about two orders of magnitude (e.g. Sloan, 1990).
Scientific interest in natural gas hydrates has increased during the last decade, and this cruise, together with ODP drilling, is part of this growing interest. The significance of natural gas hydrates is largely due to the huge volume of methane which is assumed to be stored in hydrates. The amount of carbon present in hydrates worldwide exceeds that from fossil fuels by a factor of two (Kvenvolden, 1988). Consequently, hydrates represent a huge potential energy reservoir. Development of the Siberian Messoyakha continental gas hydrate proves the technical feasibility of a long-term methane production from hydrates (e.g. Makogon, 1988).
A second reason hydrate is significant is their effect on seafloor stability. The base of the gas hydrate zone may represent a zone of weakness within the sediment column because hydrates, which act as bonding agents within the hydrate bearing layer, may inhibit normal sediment consolidation and cementation. Also, free gas may be accumulated at the base of the hydrate stability zone leading to excess pore pressure. Along the southeastern U.S. coast, locations of slope failure apparently concentrate slightly seaward of the line at which the hydrate stability zone intercepts the seafloor, although the gentle dip of the seafloor of <6 at these depths would indicate a relatively stable slope (Booth et al., 1994), an observation which clearly supports the hypothesis of hydrates being potentially involved in marine slope failure.
A final reason for studying gas hydrates is their potential link to climate changes. The amount of ethane stored in hydrates is believed to be ~ 3000 times the amount currently present in the atmosphere Kvenvolden, 1993). Since methane is a greenhouse gas, release of hydrates would have climatological implications. This process appears to be very complex, involving several factors which are difficult to constrain, in particular how much methane released at the seafloor may propagate through the water column without being dissolved and finally reach the atmosphere (e.g. Kvenvolden, 1994).
Prominent seismic reflectors, referred to as bottom simulating reflectors (BSRs) or gas hydrate reflectors, are often associated with the base of the gas hydrate stability zone. They are caused by a negative seismic impedance contrast which is mostly assumed to be due to the presence of free gas trapped beneath low-permeable hydrated sediments (e.g. White, 1977, Singh et al., 1993, MacKay et al., 1994). BSR occurrence in reflection seismic data has so far been the most important marker for the presence of hydrates in marine sediments. However, not all hydrates which have been recovered during drilling were located above BSRs.
During the last few years, a number of innovative techniques have been developed and applied to identify and possibly quantify the occurrence of oceanic gas hydrates: e.g. amplitude reduction (Dillon et al., 1994), large-aperture reflection and refraction (e.g. Katzman et al., 1994), and vertical seismic profiling (e.g. MacKay et al., 1994). Common to these approaches is the sensitivity of both compressional (P-) and shear (S-) wave velocity to the presence of gas hydrate: rigid (higher velocity) hydrate is assumed to replace part of the (lower-velocity) pore waters during hydrate formation.
This report describes cruise CH 18-95, in which gas hydrates were studied on the Blake Ridge using single-channel seismic (SCS) reflection profiling, large-aperture ocean bottom seismometer (OBS) lines and walkaway vertical seismic profiles (VSPs). The OBS data (in which seismometers are located on the seafloor) are well suited to resolve small-scale velocity variations in hydrate layers (e.g. Katzman et al., 1994, Korenaga et al., 1995, Spence et al., 1995). The walkaway VSP data (in which seismometers are clamped within a borehole at various depths) enhance the seismic ray coverage. Both of these large-aperture data sets take advantage of possible P- to S-wave conversion at the seafloor and the BSR. S-wave velocity may be more sensitive than P-wave velocity to the presence of small amounts of hydrate in sediments if hydrate acts to bond sedimentary grains (e.g. Lee et al., 1993).
Cruise CH 18-95 occurred in conjunction with drilling of ODP Leg 164, Holes 994 and 995 on the Blake Ridge in order to collect zero-offset and walkaway VSPs during drilling and to conduct isotopic experiments that could not be conducted on the drill ship because of contamination concerns. A key objective of this two-ship work is to relate physical properties of the sediments (e.g. seismic velocity) with hydrate concentration. The results of these "calibration" analyses, when combined with drilling properties and ongoing laboratory measurements in the US Geological Survey Gas Hydrate Laboratory (Winters et al., 1995), will enable researchers to understand how small-scale fluctuations observed in seismic data constrain hydrate distribution and concentration. This may ultimately lead to new techniques and approaches for mapping hydrate occurrence in marine sediments world wide.
|R/V Cape Hatteras|
Seismic Source - The GI gun worked without failing during the entire operation, which is remarkable considering the overall length of the profiles (ref. also figure 2, figure 3, figure 4, and figure 5 ). Harold Williams performed some maintenance work on the gun between both legs. Problems with the shooting system were limited to a few irregularities with the triggering system and to the failures of the shot logging system during the second walkaway experiment. The latter problem should be addressed before using this system again, because its failure makes evaluation of any data from two-ship seismic experiments as well as OBS data considerably more difficult and very time consuming.
The signature of the GI gun is excellent, which is e.g. clear from fig. 6 and Fig. 7. No deconvolution was applied to those data. Tuning however, was made difficult due to the lack of recording a clean far-field output signal. Future operations would probably benefit from the deployment of a single hydrophone at some depth below the ship to monitor the tuning of the gun.
Single Channel Seismic Data - Fig. 6 shows the dip line over the ODP Leg 164 transect at Blake Ridge. It is obvious from these data (as it was from the data used to locate the drill sites), that the BSR weakens considerable between Sites 994 and 995. However, the BSR does not disappear completely. Strong dipping reflections beneath the BSR, which display considerably lower amplitudes once they cross the BSR, may indicate that gas beneath the base of the hydrate stability zone is aligned parallel to sediment bedding, an effect previously observed in high resolution deep-towed data (Wood et al., 1995). The sharp pulse (negative/positive phase) of the signal from the BSR indicates that gun tuning was optimum. Penetration of the seismic data down to more than 0.5 s beneath the BSR demonstrates that this GI gun is an excellent seismic source for this environment, i.e. a few thousand meters of water, a few hundred meters of sediments and the desire to obtain as high a resolution as possible.
This is again demonstrated in fig. 7, which displays a migrated seismic section above the Blake Ridge Diapir. Migration was performed assuming water velocity resulting in a slight undermigration of the data, which is reflected by some remaining diffraction hyperbola at the crest of the diapir. Otherwise, however, diffraction hyperbole collapsed quite well. A clear BSR can be seen northeast of the diapir (right side) at about 3.4 s TWT at shot no. 1500, shallowing as it approaches the diapir. On the southwestern side, the BSR weakens. About 100 shots (ca. 3 km) away from the diapir, the BSR disappears completely as a single reflection and the base of the gas hydrate stability zone can only be inferred from the amplitude decrease of horizons crossing this base. The pull-up of the BSR close to the diapir possibly reflects a rise in heat flux above the diapir. A change of the chemical composition of pore water, however, might also contribute to this effect (salt, e.g., shifts the phase boundary of methane hydrates towards lower temperatures). The Blake Ridge Diapir was drilled as part of ODP Leg 164. Results from drilling should provide an explanation for this pull-up.
OBS Data - The OBS operation went smoothly without any significant problems. This is also reflected in the high quality of the data. Fig. 10 displays the vertical component of the OBS which was located at Site 994D (BRH 4). Apparent noise in some traces around shot no. 400 can be easily eliminated by applying a spike deconvolution, something we already have tested. The strong amplitudes of the refracted wave at high offsets (even visible in the multiple signal) prove that the GI gun is perfectly suited for this type of experiment, during which seismic waves travel relatively long distances through sediments. This observation also demonstrated that the GI gun was a suitable seismic source for the walkaway VSP experiment. Figure 8 shows the deployment locations.
Walkaway VSP Experiment - The principle setup of the walkaway VSP experiment at Site 995B is shown in fig.11 . Assuming that we will be able to recover the data acquired when the shot logging system failed, a 100% data coverage was achieved at this site. Four of the eight receiver stations were placed above the BSR resulting in a high coverage of rays within the layer of hydrated sediments if reflections from the BSR can be identified in the data. This will probably allow a good tomographic reconstruction of P-wave velocity around the drill hole. Ray coverage beneath the BSR depends on whether sufficient energy is reflected at horizons beneath the receiver positions.
Fig. 12 displays a section of the data (approximately 40% of the whole spread) from the deeper receiver position at Site 994D at 649.9 mbsf (well beneath the base of the gas hydrate stability zone). A slight bubble pulse is seen about 0.15 s after the main signal. The otherwise clean record having a high signal to noise ratio demonstrates the suitability of the GI gun for this experiment. Some reflected energy may be present at the left (northeastern) part of this section characterized by a shallower slope than the direct arrivals. The data from Site 995B are noisier which is not surprising considering the storm we encountered during these measurements. Fig. 13 shows a section of the data gathered at the shallowest receiver position (175.8 mbsf). The BSR can be clearly identified in these data. The source signature is worse than for all the other seismic data, which might also be due to the severe weather conditions. Two strong bubble pulses will make it necessary to deconvolve the data before interpretation. Despite the high noise level, the direct arrival can be traced over almost the entire data set, indicating again the proper choice of the source. No converted shear wave could be identified in the horizontal components so far. This was expected, however, since shear wave energy is probably quite weak and some processing to reduce the high noise level such as coherency filtering will be required before possible shear wave arrivals might become visible.
Despite the noise due to bad weather, the walkaway VSP experimental setup turned out to be well suited for studies in conjunction with ODP drilling. Problems with the VSP tool in the borehole apparently had been resolved perfectly before drilling of Hole 995B. R/V JOIDES Resolution reports will certainly comment in detail on how to address such problems for future drilling. At the shooting ship, our only suggestion for improving the experiment, apart from securing adequate shot logging, would be to try to record the far field of the GI gun to enhance tuning as well as for post-cruise processing. Ideally, this could be done by both towing a hydrophone at some distance beneath the gun to record the near-vertical signature and by recording the signal at the drill ship to obtain the near-horizontal GI gun signal.