Previous   |   Next

doi:10.2204/iodp.proc.311.213.2010

A modified model of fluid expulsion and gas hydrate formation on the northern Cascadia margin

The combined scientific results from Expedition 311 allow us to modify the earlier fluid expulsion model proposed by Hyndman and Davis (1992). A conceptual diagram incorporating the main features of this revised model is shown in Figure F12. As in the earlier model, pervasive fluid expulsion as a result of tectonic thickening and shortening of the sediment package within the accreted wedge is required to explain the observations. The margin is characterized by nonuniform fluid expulsion rates, which result in the evolution of different fluid sources in the uppermost few hundred meters below seafloor along the margin. At Sites U1326 and U1325 (within the first 10 km from the deformation front), salt generation from the ash-to-zeolite transformation dominates. In contrast, fresher pore fluids are observed farther from the deformation front (Sites U1327 and U1328) as a result of deeper rooted smectite-to-illite transformation and freshwater expulsion that becomes more prevalent landward (especially 15–20 km away from the deformation front). However, farther landward, the fluid expulsion rate decreases and the amount of freshening in the pore fluid is reduced (as seen at Site U1329). Data from Site U1329 may also indicate communication with a different (and likely more landward) fluid source of basaltic origin (other than the downgoing crust). However, the data are nonconclusive and the issue is further complicated by the presence of a pronounced unconformity, where a 5 m.y. long record of sediments is missing.

With the exception of the cold seep Site U1328, methane is produced predominantly in situ within the GHSZ. Microbes utilize organic matter (deposited either by pelagic or turbidite sedimentation) to produce methane that is consequently incorporated into gas hydrate. Continuous sedimentation and associated burial, as well as lateral transport during accretionary prism formation, result in the enrichment of the δ¹³C composition of CH₄ with distance from the deformation front and with depth at each site along the Expedition 311 transect as the organic matter is progressively degraded landward as well as with depth.

The preference for gas hydrate occurrence in coarser grained sediments has been observed at many marine (Collett et al., 2008; Weinberger et al., 2005; Kimura, Silver, Blum, et al., 1997; Ginsberg et al., 2000) and terrestrial (Dallimore et al., 1999; Boswell et al., 2008) locations and in laboratory experiments (e.g., Tohidi et al., 2001; Uchida et al., 1999). The separation of high gas hydrate concentrations within the sandier turbidite sequences and low to absent gas hydrate concentrations within the finer grained sediments without any evidence for an underlying gas migration component can be explained by the model of Malinverno (2010), in which gas is microbially generated in the finer grained sediments and then transported by diffusion into the sandier sediments, where it accumulates to saturations in excess of local solubility.

Although these models and associated geochemical and isotopic data show that in situ methane production from carbonate reduction is the main source of methane and is sufficient to explain all observations of gas hydrate distribution, concentration, and mode of occurrence along the Expedition 311 transect, the advection of methane is still a requirement in the overall formation of gas hydrate within the accretionary prism. If in situ methane production (peaking near the sulfate–methane transition zone and then exponentially decreasing with depth) is the sole source for gas hydrate formation, gas hydrate should also be abundant in the sediments of the abyssal plain as the total organic carbon content at Site 888 (average = 0.54 wt%; Cragg et al., 1996) is relatively similar to that at all other sites along the Expedition 311 transect (U1325 = 0.52 wt%, U1326 = 0.42 wt%, U1327 = 0.7 wt%, U1329 = 0.61 wt%, and U1328 = 0.5 wt%; data from Kim and Lee). However, logging and coring at Site 888 did not show any evidence of gas hydrate.

Fluid advection is observed and constrained by concentration gradients of multiple pore water constituents (e.g., Cl, Br, Li, and Sr), and fluid sources can be relatively deep (~1 km). Dissolved methane can also be transported by these advecting fluids; however, the depth of the methane does not necessarily have to be deep, and migration pathways can be relatively short (i.e., entirely within the gas hydrate occurrence zone). Continuous recycling of gas hydrate and free gas is occurring at the BGHSZ because of the active tectonic deformation of the accretionary prism, combined with rapid sedimentation and mass wasting processes. The characteristic downhole isotopic signature of the dissolved methane can result from the mixing of "freshly" produced methane by in situ bacterial processes and some "older" methane advected from below.

The vertical and lateral extent of the local gas hydrate occurrence is governed through temperature and pressure, local rates in fluid advection and sedimentation, and the abundance of organic matter and microbes, as well as by appropriate host sediment (sand). Thus, a rather heterogeneous picture of gas hydrate occurrence is derived with a high degree of variability on the kilometer scale from site to site along the Expedition 311 transect, on the 10–100 meter scale between adjacent holes at one site, and on the meter to submeter scale vertically within each hole. Focused fluid flow along faults can generate cold vents with massive gas hydrate formation near the seafloor, as seen at Site U1328. Methane in the shallow (<40 mbsf) gas hydrate accumulation at the cold vent is transported from greater depth. A model of cold vent–related fluid flow and gas hydrate formation was previously published by Riedel et al. (2006). In this model, methane-rich pore fluid and/or free methane gas is passed through the sediment column along a series of filamentous fractures. As methane solubility drastically decreases near the seafloor, massive gas hydrate is formed and excess methane is vented into the overlying ocean. The result is the formation of a massive gas hydrate cap and widespread seafloor carbonate formations and chemosynthetic communities.

The BGHSZ (seismically defined as a BSR) is a temporal trap for some amounts of free gas below, which gives rise to the seismic velocity decrease across the interface. Because there appears to be only little migration of free gas from greater depths on a regional scale, the free gas trapped below the BGHSZ could best be explained as being derived from gas hydrate recycling. The combination of continuous sedimentation processes and tectonic uplift with relatively rapid fluid migration (subsaturated in methane) results in the recycling of gas hydrate at the BGHSZ and the accumulation of a thin free gas zone as described by Haacke et al. (2007).

Top of page   |   Previous   |   Next