Previous   |   Next

doi:10.2204/iodp.proc.311.109.2006

Cold vents

Much of the recent work on gas hydrate off Vancouver Island has been on seismic blank zones that are interpreted to be cold fluid and gas vents and that contain shallow massive gas hydrate. One cold vent area has been well studied, but there is not widespread high-frequency seismic data to estimate how common such vents are on the margin, as they are not easily seen on conventional low-frequency MCS data. A number of cold vents characterized by seismic blank zones have been identified within a closely spaced area on an uplifted sediment block near Site 889 (Riedel et al., 2002), as well as ~2 km north and ~5 km southwest of the site (Riedel, 2001). This cold vent field, ~4 km x 2 km, consists of at least four vents associated with small-scale near-surface faults (Fig. F9). The most prominent vent (referred to as Bullseye), which is ~400 m across, has been the subject of intensive geophysical and geochemical studies since 1999 (Novosel et al., 2005; Lu et al., 2005; Riedel et al., 2002; Solem et al., 2002; Schwalenberg et al., 2005; Willoughby et al., 2005; Wood et al., 2002; Zühlsdorff and Spiess, 2004). A hydrate cap is evident in high-resolution seismic data, and gas hydrate within the upper 2–8 meters below seafloor (mbsf) was recovered by piston coring. Several different mechanisms have been proposed for the origin of the seismic blanking and the associated implications concerning the nature of the fluid venting. Riedel et al. (2002) suggested that much of the blanking is due to near-surface carbonate or the massive gas hydrate layer at the surface; additional blanking may be due to scattered accumulations of gas hydrate in the subsurface. Zühlsdorff and Spiess (2004) proposed that the oblong shape of the blank zone and its associated surface mound are created by natural hydraulic fracturing induced by a local elevation in the seafloor and BSR. The blanking is produced by free-gas bubbles within the fractures, particularly in the period following a fluid expulsion event. Wood et al. (2002) argued that the entire blank zone is a gas chimney produced by rising warm fluids that significantly perturb the base of gas hydrate stability upward. In a recent summary, Riedel et al. (2006) proposed a model that incorporates elements of the previous models, where the blank zone is produced by either gas hydrate or gas that has a channeled or filamentous distribution produced by fractures of varying scales in the sediments. Small fractures may be lined with high concentrations of gas hydrate, which can act as an impermeable barrier that traps free gas within the channels. However, the lack of a strong seismic velocity anomaly indicated that the total volume occupied by these fractures must be a small portion of the total volume of the vent, and so the overall amount of gas hydrate or gas must be volumetrically small. The model is also consistent with highly localized vent outlets that may be episodically active, perhaps at the time of great earthquakes. Understanding the composition, structure, gas hydrate and gas content, and origin of these blank zone structures is a very important topic for future study.

Summary of vent structure seismic observations

This 4 km x 2 km vent field is characterized by a dense population of seismic blank zones. These sub-vertical zones of reduced seismic reflection amplitude have been observed using a variety of seismic recording systems and source frequencies, but they are especially prominent in the higher frequency data. Blank zones were first identified on DTAGS data in 1997 using frequencies from 250 to 650 Hz (Wood et al., 2000) and were subsequently imaged using two-dimensional (2-D) and 3-D single- and multichannel seismic systems. The blank zones range from 80 to several 100 m in width, and blanking is restricted to the upper slope-sediment section above the BSR. The majority of the deeper accreted sediments, however, lack coherent reflectivity and blanking cannot be distinguished from the generally chaotic low reflectivity of the sediments at these depths that we associate with pervasive deformation. Blanking increases with increasing seismic frequency (Riedel et al., 2002). A typical characteristic of seismic images across the vent field are diffraction hyperbolas at the edges of the blank zones (Fig. F10), which artificially enhance stratigraphic reflector strength in the area and result in bright rims around the edges in 3-D seismic time slices (Fig. F11).

In lower frequency seismic data, several reflectors can be traced through the blank zones, but they show very little up-warping or down-warping indicative of changes in seismic velocity across the vents. Seismic velocity analyses carried out at several vents also did not detect any significant lateral changes. The lack of a seismic velocity anomaly may be explained by either the blanking produced by small amounts of gas hydrate or free gas or by a balancing combination of high amounts of gas hydrate with a few percent of free gas. Also, no significant velocity anomaly was detected using traveltime tomography from five OBSs deployed over the main cold vent (Zykov and Chapman, 2004), although massive gas hydrate was recovered from near the seafloor at this vent and the top of the gas hydrate was mapped seismically as a dome-shaped cap reflector (Fig. F11).

Summary of piston coring

The seismic surveys were complemented by two piston coring campaigns. In 2000, a total of 24 cores were taken in the area of the cold vent field, mainly over Bullseye vent and blank Zone 4 with additional cores located outside the blank zone area to verify background sediment properties. Further piston coring was carried out along an east–west transect across Bullseye vent in 2002 for more detailed geochemical analyses (Pohlman et al., 2003). The cored sediment mostly consisted of finely laminated glaciomarine clays and silts with occasional layers of sands and silts (Novosel, 2002). Detailed core physical property measurements were carried out to document sediment alterations as a result of upward methane advection and gas hydrate formation (Novosel, 2002; Novosel et al., 2005). Although thermal conductivity, density, and porosity showed only slight differences in sediments from within the seismically blank zones compared to the surrounding area, by far the strongest signal was detected in magnetic susceptibility (Novosel et al., 2005). The magnetic susceptibility of cores located within Bullseye vent is dramatically lower than the values for cores located outside this area (Fig. F12). The cores within the area of blanking have very low magnetic susceptibilities between 50 and 800 × 10^–6 SI, whereas cores from outside the blanking area have susceptibilities several orders of magnitude higher, from 2000 × 10^–6 to 4000 × 10^–6 SI. The reduction in magnetic susceptibility is associated with the formation of authigenic pyrite (Novosel et al., 2005).

Piston coring recovered massive hydrate at Bullseye vent at depths of 1–8 mbsf. The near-seafloor seismic data image the top of the hydrate as a shallow reflector around the center of the blank zone at similar depths to the cored gas hydrate (Fig. F11).

Geochemical profiles in the piston cores were also used to determined fluid flux rates (Solem et al., 2002). Sulfate reduction profiles as a function of depth represent a balance between upward methane advection and downward oxidation diffusion from the seafloor. In the background cores from outside the seismic blanking area to the southwest and northeast of Bullseye vent, the calculated methane flux varies from 10 to 19 mol/(m²·k.y.), and the sulfate/methane interface (SMI) is generally deeper than 6 m. Within Bullseye vent, the methane flux is higher, varying from 32 to 60 mol/(m²·k.y.), with the SMI being typically between 3 and 6 mbsf. In one core from the 2002 campaign, however, the SMI was very close to the seafloor. This site had the shallowest occurrence of near-seafloor massive gas hydrate (Pohlman et al., 2003).

Seafloor video observations

The 2000 coring cruise was followed by surveys with the Remotely Operated Platform for Ocean Sciences (ROPOS) in August 2000 and May 2001 to examine the seafloor environment around Bullseye vent. At several locations, large outcrops of carbonates were found (Riedel et al., 2002). Typically, the 10–15 cm thick carbonate sheets covered a seafloor area of several square meters and were located around relatively steep ridges ~3 m high. The carbonates often outcropped along multiple lineaments. Along one of these lineaments, samples of the carbonate outcrop were collected and analyzed for carbon isotopic composition. The low isotopic weights of about –45 indicate that the carbonate is formed mainly by oxidation of biogenic methane. Additional carbonate outcrops were found at blank Zone 3 as thin plates almost parallel to the seabed. There was no obvious surface elevation associated with the carbonate outcrops. A single live clam community of vesicomyids associated with worms and bacteria mats was observed at Bullseye vent during the ROPOS cruise in 2001 (Riedel et al., 2002). These chemosynthetic communities indicate active methane venting and related sulfide emissions. The clams formed small-scale colonies 1–2 m in diameter aligned around a small carbonate ridge. Several water samples were taken with ROPOS in a vertical profile above this clam colony showed a methane plume in the water column up to ~250 m above the seafloor.

New alternative surveying techniques

In the most recent studies, two new geophysical surveying techniques have been developed at the University of Toronto, CSEM sounding and seafloor compliance measurements, were applied at the vent field (especially Bullseye vent) to overcome some of the seismic imaging difficulties and related model uncertainties from the earlier studies (Schwalenberg et al., 2005; Willoughby et al., 2005) These methods are sensitive to physical properties of the sedimentary section, which are modified by gas hydrate, namely, electrical resistivity and bulk shear modulus.

Seafloor compliance is primarily sensitive to shear modulus, and therefore shear velocity, as a function of depth. Gas hydrate, which displaces pore fluids and cements the sediment grains, increases the shear modulus of the sediment column. Willoughby and Edwards (2000) found higher than normal shear velocities within the GHSZ from compliance studies near Site 889. This method was subsequently applied over the cold vent field. The new deployments over and adjacent to two of the vents show significant increases in shear modulus with depth (Fig. F13). The anomalous compliance results over the two vent sites indicate a large increase in shear modulus within the GHSZ, particularly near the base, suggesting a significant gas hydrate concentration.

The seafloor-towed CSEM system was successfully deployed in several areas on the northern Cascadia margin over the past several years (Edwards, 1997; Yuan and Edwards, 2001). Data from initial deployments around Site 889 showed an average resistivity of ~2 m for the 100 m thick layer overlying the BSR (Yuan and Edwards, 2001). This is in very good agreement with the downhole electric resistivity logs (e.g., Hyndman et al., 1999) and a factor of two higher than the log resistivities in deep-sea basin Site 888, where no gas hydrate is expected. As discussed above, it is not yet clear whether the high resistivity is mainly a result of hydrate or unusually low salinity pore fluid. More recent deployment of the CSEM array over the seismically mapped cold vents showed dramatic high-resistivity anomalies (Schwalenberg et al., 2005) as illustrated in Figure F14. The resistivity values within the anomalous zones rise locally to >5 m, about four times the regional background of 1.1–1.5 m. The resistivity values were converted to gas hydrate concentrations using both the Archie parameters of Hyndman et al. (1999) and those of Collett (2000). For both sets of coefficients, comparable amounts of gas hydrate are required to explain the observed resistivity anomalies. For Bullseye vent, the additional gas hydrate concentration reaches as much as 50% of the pore space (Schwalenberg et. al., 2005). It should be noted that the method has poor depth resolution and the concentration estimates are averages for a depth interval of several hundred meters.

Top of page   |   Previous   |   Next