Sites 994, 995, and 997, on a northeast-bearing transect across the Blake Ridge, were selected because geophysical information pertinent to gas-hydrate occurrence are particularly distinct. This information includes (1) a well-developed bottom-simulating reflector (BSR), which marks on marine seismic records, the base of gas-hydrate stability, and present at Sites 995 and 996, and (2) the occurrence at all three sites of seismic blanking, a region of low impedance contrast on marine seismic records above the BSR, and a possible indicator of gas hydrate (Lee et al., 1994). Holes were drilled to depths exceeding that of the BSR at ~450 mbsf (Table 1). Results obtained from samples collected at these three sites are listed in Table 3. As in the case of samples recovered from the diapirs, all samples from this transect of three sites contained hydrocarbon gases, with methane the most abundant. Distributions of methane with depth are shown in Figure 3 and are discussed below.
Site 994 is considered the reference site on the three-hole transect; no BSR is observed at this site, but the BSR is well-developed at the other two sites. Methane concentrations at Site 994 increase rapidly from 80 to 34,000 無/L between 8.9 and 38.9 mbsf, whereas sulfate is depleted to 0 mM in the same interval and remains near this value to the bottom of the hole. Higher molecular weight hydrocarbon gases are also present, but the amounts relative to methane are always small as reflected in the C1/(C2 + C3) ratios, which decrease with depth and are always larger than 200.
Although no BSR is present at Site 994, gas hydrate was recovered at 259 and 261 mbsf. The residual methane concentrations do not show anomalously high values in this stratigraphic interval. Thus, residual methane measurements do not appear to be useful in identifying intervals of gas-hydrate occurrence. Inorganic geochemical and logging data from this site suggest, however, that disseminated gas hydrate occurs intermittently throughout the stratigraphic interval extending between ~195 and ~450 mbsf (Paull, Matsumoto, Wallace, et al., 1996). Even in the transition interval at the projected depth of the BSR at ~450 mbsf, there is no marked change in the trend in concentrations of methane (Fig. 3).
Site 995 was the first site on Leg 164 at which drilling penetrated through a strong BSR. The site is located 3.0 km northeast of Site 994, and the drill cored the same stratigraphic interval as at Site 994. As at the previous site, residual methane concentrations increase rapidly and immediately with depth reaching maximum concentrations at 32 and 121 mbsf of 24,000 and 26,000 無/L, respectively. In general, methane concentrations are lower at this site than at Site 994; sulfate concentrations below ~20 m are depleted (~0 nM). Higher molecular weight hydrocarbon gases are present in amounts comparable to those at Site 994, and C1/(C2 + C3) ratios show the same trends.
In spite of the fact that a strong BSR was penetrated at Site 995, no samples of gas hydrate were recovered in any core from this site; however, chemical and thermal data indicate that gas hydrate is likely present in a disseminated form in the section of the cored sequence between about 190 and 460 mbsf (Paull, Matsumoto, Wallace, et al., 1996). In this stratigraphic interval, methane concentrations are generally in the range from 10,000 to 20,000 無/L. Below the BSR, the concentrations are less than 10,000 無/L in the zone where free gas exists beneath the gas hydrate (Fig. 3). Thus the presence of a free-gas zone is not obvious from the residual gas results.
Site 997 is located on the topographic crest of the Blake Ridge, and there is a very well-developed BSR that was penetrated during drilling. The site is about 6.6 km northeast of Site 995, and the stratigraphic interval of the two sites is similar. As at Site 995, concentrations of residual methane increase rapidly at the top of the hole, reaching a maximum of 62,000 無/L at 70 mbsf. With depth, the concentrations are variable and usually larger then 10,000 無/L and generally greater than the amounts observed at Site 995; sulfate concentrations at Site 997 are depleted (~0 mM) below 23 mbsf (Paull, Matsumoto, Wallace, et al., 1996). Higher molecular weight hydrocarbon gases are present in amounts usually exceeding those observed at Site 995, but the C1/(C2 + C3) ratios are similar and show a decreasing trend with depth.
In contrast to Site 995, at Site 997 one large gas-hydrate sample was recovered from 331 mbsf; however, the presence of dispersed gas hydrate was inferred, based on inorganic chemical, thermal, and logging data, to be present over a zone extending from 180 to 450 mbsf (Paull, Matsumoto, Wallace, et al., 1996). At Site 997, beneath the BSR, there was a slight increase in methane concentration in the free-gas region (Fig. 3), but this increase may not be significant as the free-gas region was not reflected in the residual gas results from Sites 994 and 995 (Fig. 3).
The concentrations and distributions of residual hydrocarbon gases as determined by the procedures described here are similar to results obtained shipboard using a headspace procedure in which ~5 cm3 of sediment was heated in a sealed glass vial (Paull, Matsumoto, Wallace, et al., 1996). In contrast, our procedure used ~170 cm3 of sediment. The main advantage of the larger sample size is that more gas can be extracted, and this gas can be analyzed for other components and for isotopic compositions.