COMPARISON BETWEEN THE STATED OBJECTIVES AND RESULTS OF THE LEG

Each drilling leg that is being considered by the ODP is outlined in a Drilling Prospectus that is circulated through the ODP community. The Prospectus documents the major objectives and a strategy to achieve the objective. Below, each of the 12 original objectives of Leg 164 are listed with a statement about what was accomplished toward meeting that objective.

1. Assessing the amounts of gas trapped in extensively hydrated sediments.

The traditional methods of core description do not work for estimating the amounts of gas hydrate because gas hydrates are unstable under surface conditions. Thus, estimates of the amounts of gas hydrate were made using the following proxy techniques. Shipboard estimates based on conservative calculations using the chloride values indicated up to 14% of the sediment volume of some samples and on average 1.3%, 1.8%, and 2.4% of the sediment volume above 450 mbsf was filled with gas hydrate at Sites 994, 995, and 997, respectively. The oxygen isotopic composition (based on a newly determined fractionation factor between water and gas hydrate) predict that 3%-6% of the sediment volume between 200 and 450 mbsf at Sites 994, 995, and 997 was occupied by gas hydrate (Matsumoto and Borowski, Chap. 6, this volume). Gas volumes from the PCS samples required that individual samples contained between 0% and 9% gas hydrate (Dickens et al., 1997). Calculations based on downhole resistivity log data indicate that 1%-11% of the bulk sediment volume was occupied by gas hydrate (Collett and Ladd, Chap. 19, this volume). Calculations based on downhole velocity log data indicate that either 12.1% (using porosity values from the bulk density) or 3.8% (using core-derived porosity) of the bulk sediment was gas hydrate filled (Lee, Chap. 20, this volume). Data from a shear-wave sonic tool indicates that 5%-10% of the pore space was occupied by gas hydrates (Guerin et al., in press). Seismic velocity data from the vertical seismic profiles suggest that at least 2% of the sediment volume is occupied by gas hydrate (Holbrook et al., 1996). The various techniques used to make these estimates operate on variable spatial scales and are based on simple assumptions. Nevertheless, the assumptions are fundamentally different between the techniques. However, the estimated values using the various techniques are broadly consistent and indicate that on average a few percent of the sediment volume was filled with gas hydrate, with some zones having higher concentrations.

Whereas the chloride data clearly indicated that the top of an extensive zone of gas hydrate extended between ~200 and 450 mbsf at Sites 994, 995, and 997, different interpretations of the chloride data produce different total amounts and distribution of the gas hydrate. The shipboard estimates were based on the departures in the chloride values from a baseline that was fitted to the measured values. However, this interpretation does not account for the occurrence of chloride values lower than seawater throughout most of the cored sedimentary section. Other ways of fitting and modeling the pore-water Cl^- values will produce higher estimates of the gas hydrate amounts (Paull and Ussler, 1997; Egeberg and Dickens, 1999). Egeberg and Dickens (1999) have modeled the data with respect to distribution of gas hydrate and predict that the top of the gas hydrate-bearing zone extends to ~40 mbsf. In their model, the gas hydrate formation starts within ~15 m of the base of sulfate reduction, where the dissolved methane concentrations in the pore waters are still only a few millimols/liter (Hoelher et al., Chap. 8, this volume).

Comparisons between the amounts of gas recovered from the PCS (Dickens et al.,1997 and Chap. 11 and Chap. 43, both this volume) and the traditional ODP headspace gas measurements (Kvenvolden and Lorenson, Chap. 3, this volume) indicate that headspace gas measurements can grossly underestimate in situ gas concentrations, especially where the concentration exceeds surface saturation values. The Leg 164 PCS data may be the first accurate measurements of gas concentrations in gas-rich marine sediments drilled by ODP (Dickens et al., 1997 and Chap. 11, this volume).

2. Contributing to an understanding of the lateral variability in the extent of gas hydrate development.

The measured variations in the lithology, physical properties, pore-water geochemistry, and downhole logs measurements among Sites 994, 995, and 997 were subtle. Estimates of the amount of gas hydrate that occurred in Site 994 (which was not associate with a BSR) and at two sites with BSRs (Sites 995 and 997), did not differ by more than a factor of two (Collett and Ladd, Chap. 19, this volume; Lee, Chap. 20, this volume). Lateral variations in the velocity fields, based on new amplitude vs. offset and acoustic tomographic analyses of multichannel profiles near the Blake Ridge drilling transect show limited lateral variations across an even longer transect (Tinivella and Lodolo, Chap. 28, this volume).

3. Refining the understanding of the relationship between bottom-simulating reflectors and gas hydrate development.

The bottom-simulating reflectors (BSRs) in the seismic reflection data from Sites 995 and 997 are associated with free gas in the underlying sediments (Holbrook et al., 1996; Collett and Ladd, Chap. 19, this volume; Lee, Chap. 20, this volume; Tininvella and Lodolo, Chap. 28, this volume). Most of the gas hydrate that was inferred to occur during Leg 164 was not near the depths associated with the BSR, but rather dispersed throughout at least the overlying ~250-m-thick interval.

4. Investigating the distribution and in situ fabric of gas hydrates within sediments.

Collectively, the inferred occurrences of gas hydrate indicated that it occurs either as (1) dispersed within the pores of fine-grained sediments without obviously disturbing the sediment structure or (2) massive hydrate within large cavities that are perhaps fractures and faults.

The distribution of the fine-grained gas hydrate within these lithologically uniform-appearing sediments (Balsam and Damuth, Chap. 31, this volume) was surprisingly heterogeneous. Data from diverse sources (e.g., visual observations of gas hydrate, core temperatures, chloride values, well logs) show that the amount of gas hydrate varies on scales of centimeters to meters. Visually obvious lithologic explanations were not seen during the shipboard core descriptions to explain the variations in gas hydrate amount; similarly, shore-based geochemical studies of the sediment chemistries have not revealed obvious explanations for these variations either (Watanabe et al., Chap. 15, this volume, Lu et al., Chap. 14, this volume. However, detailed grain-size analyses suggest that the samples with low chloride values, and thus increased amounts of gas hydrate, typically were slightly coarser grained (Ginsburg et al., Chap. 24, this volume).

On a larger scale, two zones lying between 185 and 260 mbsf and 380 and 450 mbsf contained distinctly more gas hydrate. Whereas methane recycling near the base of gas hydrate stability can easily explain the higher concentrations in the lower zone, the upper concentration zone is more difficult to explain if the lithologies were truly uniform. However, detailed studies of the lithology indicate increased abundance of siliceous microfossils in the upper zone (Kraemer et al., Chap. 23, this volume; Ikeda et al., Chap. 35, this volume). Kraemer et al. (Chap. 23, this volume) suggests that the relatively large and round pores may provide appropriate nucleation sites for gas hydrate formation.

5. Establishing the changes in the physical properties associated with gas hydrate formation and decomposition in continental margin sediments.

Shipboard physical properties measurements only show a slow progressive change in the sediment characteristics that are associated with compaction. However, the distinct variations in well-log responses suggest the in situ properties of the hydrate-bearing materials are distinct. Most of the models used to calculate amounts of gas hydrate assumed that the gas hydrate was passively filling the sediment pores or coating the grains (Guerin et al., in press) rather than preferentially cementing the grain contacts together.

6. Determining whether the gas captured in gas hydrates is produced locally or has migrated in from elsewhere.

The chemical and isotopic composition of the gases evolved from gas hydrate collected as the PCS degassed or sampled from the cores samples (Lorenson and Collett, Chap. 4, this volume; Matsumoto et al., Chap. 2, this volume; Paull et al., Chap. 7, this volume) indicate that the gases are primarily microbially produced via CO₂ reduction. Wehner et al. (Chap. 5, this volume) infer that traces of ethane and propane found in the absorbed gases indicate migration from below.

Although small in comparison with the near-surface population, substantial bacterial populations were measured throughout the sections (Wellsbury et al., 1997; Wellsbury et al., Chap. 36, this volume). Moreover, the concentrations of acetate increase at depths that are approximately coincident with the base of gas hydrate stability. Acetate is a substrate that is known to support substantial communities of methane-producing bacteria, but is usually depleted in marine sediments after just a few meters of burial. However, on the Blake Ridge, acetate reappears as a significant component in the pore water at ~350 mbsf, its concentration continues to increase to the bottom of the holes, and concentrations of up to 17 mM are reported at 728 mbsf (Wellsbury et al., 1997; Egeberg and Barth, 1998). Wellsbury et al. (1997) infers that these bacteria are actively turning over the existing acetate. However, models of the isotopic composition of the recovered gases and the known fractionation effects suggest that the majority of the methane and dissolved carbon dioxide were not locally produced (Paull et al., Chap. 7, this volume). The models indicate that the majority of the gas has migrated into the system from elsewhere.

Models of the observed changes in the pore-water chemistry (e.g., Egeberg and Barth, 1998; Egeberg and Dickens, 1999) suggest that there is a slow upward migration of pore water (~0.2 m/k.y). These models are further supported by the Cl^- isotope data (Hesse et al., Chap. 12, this volume).

Egeberg and Dickens (1999) also present pore-water bromide concentrations that exceed 3 mM, which is among the highest values ever recorded for a deep-sea sediment. Because organic matter decomposition is the only source of increased Br, and the Br concentrations are significantly higher than can be produced from in situ organic matter decomposition, they infer that there is a flux of Br from another area where there is considerable organic matter decomposition.

Over the Blake Ridge Diapir, the upward migration of gas is clearly focused along a high-angle fault system that forms a conduit to bring gas to the seafloor. Whereas regional seismic data indicate that faults that extend to the seafloor are unusual, high-angle faults that extend from the BSR to near the seafloor are very common. Woods et al. (unpubl. data) infer that these common high-angle structures may be acting as important conduits for the vertical transport of pore water and gas.

7. Investigating the role of gas hydrates in the formation of authigenic carbonate nodules.

Authigenic carbonates with relatively depleted ¹³C values occur from ~20 mbsf downward to ~180 mbsf at Sites 994, 995, and 997 (Pierre et al., Chap. 13, this volume; Rodriquez et al., Chap. 30, this volume). The onset of isotopically light carbonates is coincident with the base of sulfate reduction (Borowski et al., Chap. 9, this volume; Hoelher et al., Chap. 8, this volume; Rodriquez et al., Chap. 30, this volume). Distinct seepage-related diagenetic carbonates were also identified at Site 996, where methane is escaping to the seafloor along the fault conduits (Naehr et al., Chap. 29, this volume). Authigenic carbonate precipitation is also promoted by increased alkalinity associated with anaerobic methane oxidation (Borowski et al., Chap. 9, this volume).

Authigenic siderite is commonly associated with gas hydrate-bearing sediments (Matsumoto, 1989). Previously it was inferred that the formation of siderite occurred in association with gas hydrate decomposition at the base of the gas hydrate stability zone. However, the occurrence of siderite in these sites starts at about 180 mbsf, above the level of distinct gas hydrate (Pierre et al., Chap. 13, this volume; Rodriquez et al., Chap. 30, this volume) and well above the base of gas hydrate stability.

8. Refining our understanding of chemical and isotopic composition of gas hydrates.

The methane to ethane ratios (typically >1,000) and isotopic ratios (¹³C of methane <-60 Peedee belemnite [PDB]) of the gases recovered from the cores indicate that the gas is largely of microbial origin (Paull et al., Chap. 7, this volume). The isotopic compositions of the methane recovered from gas hydrate samples (-66 to -70 PDB and -201 to -206 standard mean ocean water [SMOW]) are essentially indistinguishable from the isotopic composition of the surrounding interstitial methane (Lorenson and Collett, Chap. 4, this volume, Paull et al., Chap. 7, this volume, Matsumoto et al., Chap. 2, this volume). However, the oxygen isotopic composition of the water ranged from 2.67 to 3.52, which is 3.5-4.0 heavier than the ambient interstitial waters (Matsumoto et al., Chap. 2, this volume).

The isotopic composition of the gas samples from the PCS showed significantly more scatter in values than the core gases from a similar depth (Paull et al., Chap. 7, this volume). However, the volume weighted averages for the total gas from a single PCS core is the same as that from adjacent free gas and headspace samples (Wallace et al., Chap. 10, this volume). This indicates that there is no isotopic fractionation of methane during core recovery despite considerable gas loss (upwards of 99.9%). The cause of the increased scatter for individual gas samples from a single PCS core remains unclear.

Measurements of the noble gas composition were attempted (Dickens and Kennedy, Chap. 16, this volume). The objective was to determine whether significant fractionation of noble gases and enrichment of xenon occurs during gas hydrate formation. All specimens of gas hydrate from Leg 164 that were analyzed are enriched in xenon relative to noble gas ratios in air and air-saturated seawater. However, the experiments were far from conclusive because of significant contamination.

9. Determining the gas composition, hydration number, and crystal structure of natural gas hydrates.

The gas that filled the hydrates was primarily methane with an isotopic composition that indicates that it was produced via CO₂ reduction (Matsumoto et al., Chap. 2, this volume, Paull et al., Chap. 7, this volume). Samples of gas hydrate that were recovered from the cored sediments yielded volumetric water to gas ratios that indicate that a minimum of 71% of the potential gas-bearing cages and, after corrections for potential contamination, suggested that probably most hydrate cages are gas filled (Lorenson and Collett, Chap. 4, this volume). Nuclear magnetic resonance measurements and X-ray diffraction analysis confirm the high cage occupancy and indicate that both the large and the small cages were filled with methane (Matsumoto et al., Chap. 2, this volume).

10. Determining the role of gas hydrates in stimulating or modifying fluid circulation.

Active fluid flow was demonstrated to be occurring on the top of the Blake Ridge Diapir at Site 996. Apparently, faults form conduits for advective degassing of methane associated with gas hydrate-related reservoirs on the Blake Ridge. The isotopic compositions of methane captured within sediments at <50 mbsf at this vent are consistent with the isotopic composition of methane sampled from the gas hydrate- and free-gas-bearing zones below ~150 mbsf at Deep Sea Drilling Project Site 533 (Brooks et al., 1983; Galimov and Kvenvolden, 1983) and on ODP Leg 164 (Paull et al., Chap. 7, this volume).

Models of the pore-water chemical gradients at Site 997 (Egeberg and Dickens, 1999) suggest that there is slow upward migration of pore water (~0.2 m/k.y.). However, the extent to which the gas hydrates themselves alter sediment properties and effect the patterns of fluid flow was not resolved. Similarly, the effect of the free gas in the sediments below the base of gas hydrate stability is unknown. The occurrence of pore waters that are significantly fresher than seawater throughout the sedimentary section will also increase the buoyancy of the pore waters.

11. Investigating the potential connection between major slumps and the breakdown of gas hydrate.

A direct connection between gas hydrate decomposition and slumping was not generated. However, physical properties of cores that came from depth on the Blake Ridge are significantly undercompacted. One explanation for this undercompaction is that the presence of free gas and gas hydrate in the sediments prevents normal compaction from occurring (Winters, Chap. 40, this volume).

12. Establishing the influence of the Carolina Rise Diapirs on the gas hydrates as well as the origin of the diapirs themselves.

The pore-water chloride profiles over the diapirs increase down core, suggesting that these are salt-cored diapirs. However, little was learned about the effects that the diapirs exert on the gas hydrate.