NO EVIDENCE FOR ENHANCED GAS FLUX FROM THE BLAKE RIDGE DEPRESSION

C. K. Paull and W. Ussler III
MBARI, Moss Landing, California

Piston cores were collected from the floor, flanks, and background sediments associated with the Blake Ridge Depression to determine if this area is a gas-venting site. The hypothesis was that if the depression is associated with focused methane flux, either authigenic deposits indicative of methane-related diagenesis or steeper geochemical gradients would occur for the methane and methane oxidation sensitive pore water constituents in cores from within the depression.

Introduction

Many authors have postulated that the release of large volumes of methane into the Earth’s oceans from gas hydrate stored within the shallow geologic reservoirs along the world’s continental margins would alter the isotopic composition of the dissolved inorganic carbon pool in the ocean and/or the methane concentration of the atmosphere. These conjectures are based on both the size of the marine gas hydrate reservoir and the distinct isotopic composition of methane.

The viability of these concepts depends on establishing mechanisms to transfer methane from gas hydrate deposits upwards into the ocean and atmosphere and on defining sites where this may have happened. Some mechanisms may apply wherever gas hydrates occur within seafloor sediments (e.g., thermal decomposition or pressure reduction), while other mechanisms require special local conditions. One site that has been identified as an area that might be associated with volumetrically significant methane venting is the Blake Ridge Depression.

Blake Ridge Depression

The Blake Ridge is a large sediment drift off the southeastern United States known to contain extensive gas hydrate deposits. Most of the ridge crest is associated with generally smooth seafloor and laterally continuous strata. However, one area on the ridge crest near the 2,650 m contour is associated with complex bathymetry and contains a >20 km long and >100 m deep bathymetric depression (Fig. 1), referred to as the Blake Ridge Depression (BRD). This feature is nearly enclosed with relatively steep-sided bounding escarpments. The topography has been interpreted as a giant collapse depression (Dillon et al., 2001), and more recently as a sediment wave field (Holbrook et al., 2002), based on interpretation of seismic reflection profiles from the region.

Bottom simulating reflectors (BSR) are noticeably absent in sediments beneath the depression, while the seafloor surrounding the depression has well-developed BSRs. This suggests that free gas is common below the BSR in sediments surrounding the depression, but is absent underneath the depression. The absence of free gas under the BRD has led to speculation that large quantities of methane from the sediments under the BRD may have vented to the ocean or atmosphere system (Dillon et al., 2000; Holbrook et al., 2002). In the collapse model, the methane would have escaped along the bounding faults while the sediment drift model infers that there may have been channels of enhanced permeability in the sedimentary structures and that some gas escaped as a consequence of erosion associated with lateral migration of the sediment waves. Both models predict enhanced methane flux to the seafloor within the BRD.

A piston coring cruise was conducted to investigate the origin of this structure, its association with the surrounding gas hydrate deposits and the potential role of this structure in stimulating methane gas venting. Our approach was to sample sites where gas venting is postulated to have occurred and to measure the methane-sensitive pore water constituents, particularly sulfate. The concepts behind using sulfate as a proxy for methane flux are outlined in Borowski et al. (1999).

Methods

Thirty-three piston cores were collected from within and around the BRD in July 2000 using the R/V Cape Hatteras (Table 1). Pore water samples extracted from whole-round sub-samples of the cores were analyzed shipboard to determine their methane, sulfate, chloride, ammonium, and dissolved inorganic carbon concentrations using methods described in Borowski et al. (1999). Carbon isotope measurements of dissolved methane and carbon dioxide preserved in serum vials were made on selected samples at the University of North Carolina at Chapel Hill. Percent carbonate, and ¹³C and ¹⁸O measurements of disseminated carbonate in sub-samples of the squeezed sediments were conducted at the Stanford Stable Isotope Lab at Stanford University in Palo Alto, California. Additional sediment sub-samples were selected for analysis of the ¹⁴C content of the carbon contained in the disseminated organic matter. These analyses were preformed at the Center for Accelerator Mass Spectrometry facility at the Lawrence Livermore National Laboratories in Livermore, California. Nannofossil assemblages in a sample from the base of each core were determined by Tim Bralower. After the cruise, cores were split horizontally, described, and placed in the Woods Hole Oceanographic core repository.

Results

The lithology of cores from the BRD consists of nannofossil-rich hemipelagic sediment of variable consistency. Soft sediments occurred throughout some cores, while in other cores soft sediments were underlain by distinctly firmer sediment of a similar lithology. The firmer sediments were encountered primarily on the steep scarps and floor of the depression. Carbonate content of the sediments sampled varied from 0.1% to 49.8% with a mean value of 21.0% + 9.5% (n=69), with no significant difference between the softer and firmer sediments. Although the cores were collected to establish whether gas venting has happened, no indications of methane-associated diagenesis, more permeable horizons, or vein-like features were noted.

Thirteen ¹⁴C measurements made on sediment samples from between 60 to 400 cmbsf (cm below seafloor) in PC-2, PC-11, and PC-30 indicate that the upper soft sediment unit has undergone more or less continuous deposition in the Holocene with sediment accumulation rates of 110, 61, and 50 cm per thousand years. Four measurements on PC-19 show that only the uppermost sample (61.5 cmbsf) is Holocene, whereas ¹⁴C-depleted sediments occur below 116 cmbsf. The nannofossil assemblages in samples from the bottoms of the cores confirm that the firmer sediments are Pleistocene in age (Table 1). The older sediments came from cores collected on the exposed steeper sides and floor of the depression or along the northern eroded flank of the Blake Ridge.

Methane concentrations were measured on 237 samples. Only twelve samples had methane concentrations of over 100 µM. These methane-rich samples all came from PC-26 and PC-28 between 1,236 and 1,440 cmbsf. The remaining 225 samples had a range of methane concentrations between 0 and 85 µM with a mean of 1.5 µM. These methane profiles are typical of others measured along the Blake Ridge, and none indicated there is a significant advective flux of methane to the seafloor at these core sites.

Sulfate concentrations were measured on 235 pore water samples and ranged from seawater-like values at the top of the cores (~28 mM) to sulfate-free waters at the bottoms. Over the range of measurements, sulfate concentration profiles decrease approximately linearly with depth. Pore water sulfate concentration measured in 11 of the 12 methane-rich samples has a mean value of 0.3 mM. Thus, the most methane-rich sediment samples came from below the sulfate methane interface (SMI).

While only PC-26 and PC-28 penetrated the SMI, the depth to the SMI in the other cores can be estimated assuming a linear gradient was maintained below the deepest sample (Table 1). The shallowest SMI were in cores that came from near the crests of the sediment ridges within the BRD, ranging between 10 and 14 mbsf (meters below seafloor). Cores outside the BRD had SMI at depths between 13 and 16 mbsf, which is consistent with the previous measurements that have been made in the region (Borowski et al., 1999). Cores from the floor of the BRD, which were specifically targeted to sample potential gas migration routes, had deeper extrapolated SMI depths with values ranging from 21 to 38 mbsf. The deepest extrapolated SMI depths were found in cores collected from the flanks of the escarpments, ranging from 12 to 71 mbsf.

Twenty-four methane ¹³C values ranged from –102.7‰ to –62.2‰ (PDB). The ¹³C values of the CO₂ on 9 of the same samples ranged from –36.4‰ to –30.4‰ (PDB). The range of ¹³CCH₄ values demonstrates that this is microbially-generated methane and the extremely negative ¹³CCH₄ values near the SMI are similar to those previously reported. The low values of the ¹³CCO₂ are also suggestive that methane oxidation has occurred within samples obtained near the SMI. The ¹³C composition of CO₂ in these samples, which is distinct from seawater-like values, indicates that if substantial amounts of authigenic carbonate have precipitated within the sediments, it should leave a distinct carbon signature. Stable carbon isotopic values of disseminated carbonate in 68 squeeze cake samples range from –5.13‰ to 1.3‰ (PDB). However, only two sample have values more negative than –1.9‰ and both were from samples with less than 1‰ CaCO₃. Thus, these ¹³C measurements do not indicate that methane-derived authigenic carbonates have precipitated in these sediments.

Discussion and Conclusions

The projected depths to the SMI in cores from the crests of the interior ridges and outside the BRD are distinctly shallower than those from the floor or steep surfaces of the bounding escarpments. This implies that less methane is diffusing out of the depression’s floor and through its bounding scarps than from the surrounding area with a thicker sediment drape. This data is inconsistent with the BRD being an area where methane is leaking onto the seafloor today. One interpretation of these data is that any methane formerly contained in the strata exposed on the scarp faces has already been vented into the ocean. However, neither the chemical gradients nor absence of authigenic minerals provide evidence that methane venting has occurred in the past.

References

Borowski, W.S., Paull, C.K., and Ussler, W. III, 1999, Global and local variations of interstitial sulfate gradients in deep-water, continental margin sediments: Sensitivity to underlying gas hydrates, Marine Geology, v. 159, p. 131-154.

Dillon, W.P., Nealon, J.W., Taylor, M.H., Lee, M.W., Drury, R.M., 2001, Seafloor collapse and methane venting associated with gas hydrate on the Blake Ridge – Causes and implications to seafloor stability and methane release, in Paull, C.K., and Dillon, W.P., eds., Natural Gas Hydrates, Occurrence, Distribution and Detection, AGU Geophysical Monograph 124, p. 211-234.

Holbrook, W.S., Lizarralde, D., Pecher, I.A., Gorman, A.R., Hackwith, K.L., Hornback, M., Saffer, D., 2002, Escape of methane gas through sediment waves in a large methane hydrate province, Geology, v. 30, p. 467-470.

Table 1. Core locations, length, water depth, projected depth to the SMI, age of the core catcher sample based on calcareous nannofossil assemblage, and environment are listed for piston cores collected within and around the Blake Ridge Depression.