BACKGROUND

Geological Setting

The area of this investigation is on the accretionary prism of the Cascadia subduction zone (Fig. F1). The Juan de Fuca plate converges nearly orthogonally to the North American plate at a present rate of ~45 mm/y (e.g., Riddihough, 1984). Seaward of the deformation front, the Cascadia Basin consists of pre-Pleistocene hemipelagic sediments overlain by a rapidly deposited Pleistocene turbidite for a total sediment thickness of ~2500 m. Most of the incoming sediment is scraped off the oceanic crust and folded and thrust upward to form elongated anticlinal ridges with elevations as high as 700 m above the adjacent basin. The thrust faults near the deformation front penetrate nearly the entire sediment section (Davis and Hyndman, 1989).

During Leg 146 three sites (Fig. F1) were drilled that are of significance to Expedition 311. Site 888 was drilled in the deep Cascadia Basin and is near the proposed location of Site CAS-04B and served as a reference site for a "non-gas-hydrate" environment. Sites 889 and 890 were drilled midslope over a strong BSR. These two sites are most relevant to the proposed Expedition 311 Sites CAS-01B and CAS-01C. The following descriptions were taken from the Leg 146 Initial Results volume (Westbrook, Carson, Musgrave, et al., 1994).

Site 888 lies in an outer part of the Nitinat Fan, 7 km seaward of the toe of the accretionary wedge. This site provides a reference for the type, age, and physical properties of sediment in the sedimentary section that is stripped from the ocean crust to form the accretionary wedge. The three holes at Site 888 penetrated into the top 600 m of the sedimentary section, which is 2.5 km thick at this location. Three lithostratigraphic units were recognized.

Magnetic polarity analyses and biostratigraphy indicate that the entire cored section at Site 888 is younger than 600 ka.

The geothermal gradient has been established as 68°C/km. Thermal conductivity increases in value downward through the uppermost 200 mbsf to a mean value of 1.23 W/m·K for the section below that depth.

Measurements of porosity and shear strength indicate that sediments in the cored section are underconsolidated. Wireline density and neutron porosity logs show that the minimum porosity of the section lies at 300 mbsf. The downward increase in porosity shown in the logs beneath 300 mbsf may be an artifact of poor hole condition. However, porosity measurements conducted on the cored material show a downward decrease beneath 300 mbsf.

The general state of undercompaction indicated by the logs and physical property measurements may be attributed to rapid deposition, especially of the sandy section in Unit II. The rate of sedimentation in the upper 100 mbsf has been close to 1 m/k.y., and sedimentation rates in the remainder of the cored section have at least matched that rate and were probably greater.

The pore water geochemistry in the section varies downward in response to bacterial sulfate reduction, carbonate diagenesis, and fluid flow within some intervals. Pore water chlorinity varies only between 543 and 571 mM. However, small-scale variation in the section may be indicative of fluid flow to maintain observed pronounced minima.

Overall, the organic carbon is at a low concentration (0.2–0.4 wt%). Concentration of methane in the upper 200 mbsf is below 5 ppmv. Ethane, propane, and butane are present only in trace amounts (C₁/C₂ > 1000), indicating that methane is of bacterial origin.

Sites 889/890 are located on the mid-continental slope off Vancouver Island at water depths of 1315 and 1326 m, respectively (Fig. F5). Coring began in bedded slope-basin sediments and at Site 889 extended into underlying deformed sediments of the accretionary wedge. Site 890 was cored to only 50 mbsf to sample the near-seafloor sediments. A major objective of this site was to investigate the BSR at a depth of 225 mbsf. Holes 889A, 889B, and 889C penetrated the BSR.

Sediments at Sites 889/890 range in age from late Quaternary to late Pliocene. A hiatus in the record is present at 87 mbsf, separating upper Quaternary from lower Pleistocene deposits. Three lithostratigraphic units were determined.

Unit I includes clayey silts, fine sand, and diagenetic carbonates. It extends from the surface to a depth of 128 mbsf. Unit I comprises slope and slope basin sediments that are hemipelagites, turbidites, and mass flow deposits.

Unit II is similar to Unit I, but it is noticeably more consolidated than the overlying unit and is highly fractured. Diagenetic carbonates were observed throughout the entire section. Unit II is thought to consist of abyssal plain silts and clays that were post-depositionally fractured during accretion. Unit II extends to 301.5 mbsf, beneath which the glauconite content increases sharply at the top of Unit III. The sediments of Unit III appear to be abyssal plain deposits like those in Unit II above, but the abundant authigenic glauconite suggests deposition under suboxic conditions.

The downhole variation in consolidation is clearly shown in the distribution of bulk density and porosity. Unit I is characterized by normally consolidated deposits in which porosity declines regularly with increasing depth. Between 128 and 260 mbsf (upper Unit II), the porosity decreases rapidly and the sediments become overconsolidated to the base of Unit II. There are distinct excursions from the general porosity decrease with depth that correlate with variations in the diagenetic carbonate cementation and organic geochemistry (organic carbon content, methane, N, and S concentrations). The position of the BSR falls within one of these excursions, but it is not unique. Unit III exhibits an apparently anomalous increase in porosity with increasing depth.

The inorganic chemistry of the pore waters defines two zones: within lithostratigraphic Unit I, sulfate declines to 0 at 10 mbsf, accompanied by an increase in alkalinity, and as a result Ca²⁺ and Mg²⁺ concentrations decrease rapidly from 0 to 60 mbsf. Chloride concentration declines from 550 mM at the sediment/water interface to 363 mM at the base of chemical Zone 1 (130 mbsf).

Chemical Zone 2 (130–386.5 mbsf) shows nearly constant Cl^– concentration of 370 mM. The concentration of Na⁺, Mg²⁺, and phosphate are also nearly constant, indicating that the low Cl^– concentration is a dilution effect.

Lithostratigraphic Unit I yielded elevated methane concentrations (60,000–80,000 ppmv) below the sulfate reduction zone. Methane declines to 30,000 ppmv at the base of Unit I. Over this same interval, ethane and propane were essentially absent, and the C₁/C₂ ratio was > 1000, indicating a microbial methane source. A small spike in the ethane concentration (33 ppmv) at 129 mbsf (associated with the Cl^– minimum) suggests a deeper source for fluids at that level.

Gas composition changes markedly in the interval 130–247 mbsf, with headspace (HS) samples containing 30,000–77,000 ppmv methane, 5–35 ppmv ethane, and 0.5–3.4 ppmv propane. Below 250 mbsf the methane concentrations become highly variable. Ethane increases markedly below 300 mbsf, and propane increases rapidly below 360 mbsf. Within this interval, the C₁/C₂ ratio declines from values of ~2000 to <100, indicating a mixture of thermogenic and bacterially derived gas. There is no evidence in either the logs or the cores for the accumulation of gas hydrate in massive form, but the temperature of –1.4°C measured in a core at 220 mbsf could have been produced by the dissociation of gas hydrate.

Six in situ temperature determinations were made with the Water Sampling Temperature Probe and Adara tools. Temperature increases linearly with depth, with a gradient of 54°C/km.

The BSR at Sites 889/890 is situated 276 ms two-way traveltime below the seafloor. Time-depth curves derived from the vertical seismic profile (VSP) and sonic log indicate that the BSR is located at 225 mbsf (Hole 889B). Although the acoustic log does not exhibit a substantial decrease in velocity across the BSR, the VSP data define a rise in velocity just above the BSR with a distinct low-velocity zone beneath it. Velocities are lower than those of seawater and are typical of the presence of small amounts of free gas. The discrepancy between VSP and acoustic log results may be attributed to drilling disturbance, which could deplete the gas phase in the immediate vicinity of the drill hole.

Gas hydrate concentrations were previously estimated from the downhole acoustic and resistivity logs, multichannel seismic (MCS) core analysis, and VSP velocities and from pore water freshening (e.g., Yuan et al., 1996, 1999; Hyndman et al., 1999). Gas hydrate concentrations from the different methods vary slightly but were estimated to be on average 20%–35% of the pore space over a 100 m thick interval above the BSR.

These are high gas hydrate concentrations for a marine gas hydrate occurrence, and are not observed on other margins. ODP drilling and logging at the Blake Ridge offshore South Carolina (passive margin environment) showed gas hydrate concentrations that were on average lower than 10% (Paull, Matsumoto, Wallace, et al., 1996). Results from Leg 204 at southern Hydrate Ridge located on the southern Cascadia margin offshore Oregon, with a very similar tectonic environment to offshore Vancouver Island, showed very low gas hydrate concentrations of <5% except for the unusual summit of Hydrate Ridge (Trehu et al., 2004). An earlier approach by Ussler and Paull (2001) for interpreting chlorinity data from Leg 146 by using a smooth chlorinity baseline suggested lower gas hydrate concentrations at Sites 889/890, but were in an apparent contradiction to other geophysical results using electrical resistivity and seismic velocity.

In preparation for Expedition 311, the gas hydrate concentrations along the northern Cascadia margin were recalculated using Leg 146 acoustic/electrical resistivity logs and pore water chlorinity/salinity data (Riedel et al., in press a). New estimates show that the concentrations could alternatively be between 5% and 10% on average of the pore volume from ~130 mbsf to the BSR (at ~230 mbsf) if different baselines are used in the individual calculations.

No conclusive decision can be made on the chlorinity baseline without additional pore water chemistry data (lithium and strontium data) to confirm that the model by Torres et al. (2004) derived from Leg 204 results is also valid for the Sites 889/890. Therefore, the gas hydrate concentrations along the northern Cascadia margin based on chlorinity data may either be as low as a few percent, as suggested earlier by Ussler and Paull (2001), or as high as 40%, previously suggested by Hyndman et al. (1999).

Gas hydrate concentrations were calculated from the resistivity data using Archie's law. Archie's law consists of several empirical parameters (referred to as a, m, and n), which were determined to be a = 1.4, m = 1.76, and n = 1.0 by Hyndman et al. (1999) using the Leg 146 core data at Sites 888 and 889/890. However, Collett (2000) redefined these parameters to a = 1.0, m = 2.8, and n = 1.9 by using the Site 889 resistivity and neutron porosity logs. The differences in these empirical parameters result in highly different estimated gas hydrate concentrations. The Leg 146 logging data are of relatively poor quality and neither the new parameters nor the previous analyses fit all data well, and a large uncertainty remains in the results. However, results from Leg 204, obtained from a tectonically similar environment, suggest that the Archie parameters by Collett (2000) may be preferable.

Finally, new velocity references were calculated from the Site 889 porosities using various published empirical relations between porosity and velocity (Jarrard et al., 1995; Hyndman et al., 1993) and the Lee et al. (1993) weighted equation. All the newly proposed baselines are significantly shifted toward higher seismic velocities relative to the former baseline defined by Yuan et al. (1996) and Hyndman et al. (2001). Thus, they result in gas hydrate concentrations that are only between 5% and 10% of the pore volume. However, significant uncertainty remains in the applicability of the empirical parameters for each of the individual equations.

If the model by Torres et al. (2004) proposed for Hydrate Ridge (Leg 204) is applicable to Site 889 and if the Archie parameters by Collett (2000) are correct, these two methods (although individually very uncertain) confirm results from the acoustic velocity analyses of Lee et al. (1993). Thus, the total gas hydrate concentration at Sites 889/890 may be much lower than previously assumed.

Extensive site survey data exist on the Cascadia margin to support the operations of Expedition 311: