Integrated Ocean Drilling Program (IODP) Proposal GAS HYDRATE ON THE CASCADIA MARGIN

This proposal is for an IODP program to constrain models for the formation of deep sea gas hydrate in subduction zone accretionary prisms. The objectives include the deep origin of the methane, its upward transport, its incorporation in gas hydrate, and its subsequent loss to the seafloor. The main attention is on the widespread seafloor-parallel layer of dispersed hydrate located just above the base of the stability field (~250 m below seafloor). Such layers may make up the largest volume of hydrate globally. In the model, methane is carried upward through regional grain-scale or small-scale fracture permeability, driven by the tectonic consolidation of the accretionary sedimentary prism. Also important is the focussing of a portion of the upward methane flux into localized plumes or channels to form concentrations of nearseafloor hydrate. The amount of hydrate in local concentrations near the seafloor is especially important for understanding the response of marine hydrate to climate change. Long-term monitoring in the boreholes, especially over a strong earthquake concentration (Nootka fault), will assist in determining the role of strong shaking in the sediment consolidation, episodic upward fluid transport, and hydrate formation. The proposal is for drilling, downhole measurements, and long-term recording at a transect of sites across the Northern Cascadia large accretionary prism, from the first indication of hydrate just landward of the deformation front, to the upper continental slope where hydrate is no longer stable. The sites will track the history of methane in an accretionary sedimentary prism from: (1) its production by mainly microbiological processes (and to a lesser degree thermal processes) over a thick sediment vertical extent, (2) its upward transport through regional or locally focussed fluid and free gas, (3) its incorporation in the regional hydrate layer above the BSR or in local concentrations at or near the seafloor (and how the methane gets to near the seafloor without forming hydrate at greater depth), to (4) methane loss from the hydrate by upward diffusion, or eventually by the base of the stability field moving upward to the seafloor as the water depth shoals landward though sediment tectonic thickening, and (5) methane oxidation and incorporation in seafloor carbonate, or expulsion to the ocean. The proposal builds on the previous Cascadia hydrate drilling of Leg 146 in the area that included hydrate Sites 889-890 and deep sea reference Site 888, on more recent Leg 204 off Oregon, as well as previous hydrate drilling elsewhere. The recent Mallik II hydrate drilling program in the Canadian arctic provided important testing of a number of new measurement technologies. Previous drilling mainly addressed the nature and distribution of hydrate occurrence. A series of important new technical developments in measurement and borehole recording have made achieving the new objectives of tracking the formation to the dissociation of hydrate now very feasible. Further technical developments are in progress, so the discussion here may be modified to take advantages of new technologies. Borehole instrumentation is developing rapidly. Important facilities for this proposal include, (1) the now well-developed CORK monitoring of downhole pore pressure, permeability, and thus upward fluid flow, (2) Modular Formation Dynamics Tester (MDT), (3) Log-While-Drilling (LWD), (4) Distributed Temperature Sensors (DTS), (5) Infrared (IR) imaging of the recovered core, and (6) Pressure Core Barrel sampler for hydrate, free gas, and fluid recovery under insitu conditions. Long-term monitoring of the boreholes may be facilitated by the proposed NEPTUNE seafloor cable system. The area on the continental slope off Vancouver Island is one of the most completely studied of accretionary prism gas hydrate, together with Hydrate Ridge on the central Cascadia continental slope. There are detailed and comprehensive site surveys and other field studies, including: very extensive seismic reflection surveys (conventional MCS, 3D MCS, high resolution surveys, deep-towed DTAGS, and OBSs), heat flow, electrical sounding, seafloor compliance, piston coring, swath bathymetry, and acoustic imaging. The Lithoprobe program has allowed close integration of marine and land geophysical and geological data, and continuity of geological structure from marine to land. The previous Cascadia and other hydrate drilling has been analysed and interpreted, and the hydrate formation theories have now been developed into quantitative models that may be tested by an IODP program. INTRODUCTION This proposal is for an IODP program on the northern Cascadia margin to constrain models for the formation of deep sea gas hydrate in subduction zone accretionary prisms. Natural gas hydrate occurs beneath some continental slopes and in arctic permafrost areas. The arctic occurrences can have very high concentrations, but appear to contain less total hydrate than marine occurrences. Recent studies have indicated that the largest occurrences of gas hydrate may lie in nearly horizontal layers several hundred metres beneath the seafloor of continental slopes, especially in the large subduction zone accretionary sedimentary prisms. Such hydrate and underlying free gas produce the ubiquitous BSR. Hydrates do occur on passive margins, but they are less common and appear usually to contain lower concentrations. The importance of deep sea gas hydrates is highlighted by the Japanese second $70 Million, 5-year government-industry program, and large research and survey programs of U.S.A., India, Korea, Canada, Germany, and several other countries. The two marine hydrate areas that have received the most detailed scientific study, including previous drilling by ODP, are the Blake Ridge region off eastern U.S.A. (a passive margin setting), the Cascadia margin off Oregon and Vancouver Island (subduction zone accretionary prism). Other hydrate drilling on an accretionary prism has been off SW Japan (by JNOC/JAPEX). The Blake Ridge has a widespread bottom-simulating reflector (BSR) but the hydrate concentrations are relatively low. Very important new information for arctic hydrate has been obtained from the Mallik I and Mallik II drilling in northern Canada (Dallimore et al., 1999, 2002; also see Mallik and ICDP web sites). The very high concentrations encountered at the Mallik site (up to 80% pore volume) suggest that such arctic deposits may be the first to be exploited for energy. However, the total amounts in such occurrences appear to be less than the total in marine hydrate. If our conclusions for the distribution and concentration of most hydrate are correct, accretionary prism hydrate is the most important both for the long-term energy potential of hydrate, and for the role that natural gas hydrate plays in climate change. The larger clastic sedimentary prisms (including Cascadia, S.W. Japan, Alaska, S. Chile, and Makran) make up approximately 20 % of the total of subduction zone margins. Within accretionary prisms, the largest amount of hydrate appears to occur in a very widespread layer located just above the BSR (e.g., Trehu et al., 2002). However, also important is the focussing of a portion of the upward methane flux into localized plumes or channels to form concentrations of near-seafloor hydrate. The amount of hydrate in the widespread layer above the BSR (~200 m below seafloor), compared to that in local concentrations near the seafloor is especially important for understanding the response of marine hydrate to climate change. Near-seafloor hydrate will respond much more quickly to ocean temperature changes Figure 1. General location of proposed drilling transect near previous ODP Site 889/890. A BSR is present on about 50% of the mid-continental slope (shaded area). compared to hydrate several hundred metres below the seafloor. For the region of ODP Site 889/890, Taylor et al. (2002) calculated that a 30 m thick hydrate deposit lying at the base of the stability field would dissociate due to seabed warming in approximately 8000 years. Nearseafloor hydrate could dissociate much more quickly, and be much more responsive to humaninduced global warming (e.g., Wood et al., 2003). It remains a puzzle how methane can get to near the seafloor without forming hydrate at deeper levels. Off Vancouver Island a hydrate BSR occurs in a 30 km wide band parallel to the coast beneath much of the continental slope (Figure 1). The hydrate is concentrated in a layer 50-100 m thick, just above the base of the hydrate stability field, which is located 200-300 m below the seafloor. The hydrate concentrations reach about 30% of pore space base upon seismic velocity and electrical resistivity data. The surveys and studies that have been carried out, and the evidence for the presence and content of gas hydrate have been summarized in two recent comprehensive review articles (Hyndman et al., 2001; Spence et al., 2000). Models for formation of wide-spread gas hydrate layer A general model for deep sea gas hydrate formation by removal of methane from upwardly expelled fluids was developed earlier for this area (Hyndman and Davis, 1992). Mainly biogenic methane, inferred to be produced over a thick sediment section, is carried up to form hydrate when it enters the stability field. The hydrate concentration is predicted to be greatest just above the BSR as is usually observed. Subsequently a model has been proposed for how free gas and resulting BSR will be formed as the base of the hydrate stability moves upward due to, post-Pleistocene seafloor warming, uplift, and sediment deposition (Westbrook et al., 1994; Paul and Ussler, 1997; von Huene and Pecher, 1998). Physical and mathematical models have been developed for the formation of hydrate involving upward methane advection and diffusion (e.g., Xu and Ruppel, 1999; Rempel and Buffett, 1997, and others). Testing these models and determining the appropriate model parameters requires: (1) accurate definition of the vertical distribution of hydrate and gas, (2) accurate formation temperatures to define the base of the stability field, (3) physical and fluid chemical data, and downhole measurements and recording, that define the vertical advection rates of fluids and of methane, (4) better calibration of the effect of hydrate and gas concentrations on velocity, resistivity, and other physical parameters for interpretation of both downhole data and seafloor measurements and surveys, (5) determination of the sediment pore pressure and permeability that drive the upward advection. Recently evidence for focused fluid/gas flow and gas hydrate formation has been identified at at least two sites on the Vancouver Margin. The first site is an active cold vent field associated with near-surface hydrate close to ODP Site 889/890 that has been focus of intense research (e.g. Riedel et al., 2002). Studies include high-resolution bottom profiling, 3D seismic surveys, piston coring and ocean-bottom video surveying and sampling with the remotely operated vehicle ROPOS. These vents represent fault-related conduits for focused fluid and/or gas migration associated with massive hydrate formation within the fault zone and represent therefore the opposite mechanism to the widespread fluid-flow. It is so far unknown how important these cold vents are in the total budget of fluid flow in an accretionary prism. Drilling at the vent field would help to constrain the significance of fault-related fluid and/or gas flow. The second site is located in Barkley Canyon at shallower water depth around 800 m and hosts massive gas hydrate outcrops. Some of these outcrops have been accidentally dredged by a fish boat in summer 2000, and an estimated 1.5 t of gas hydrate was brought to the surface in the fishing net (Spence et al., 2001). The role of earthquake shaking on fluid expulsion and hydrate formation There may be important time variations in fluid expulsion and hydrate formation, especially the influence on fluid expulsion and hydrate formation of very strong shaking. Most of the fluid expulsion could occur as a consequence of the megathrust earthquake rupture located immediately beneath the continental slope hydrate deposits. Magnitude ~9 thrust earthquakes occur approximately every 600 years on this margin, with the last event 300 years ago. In order to look at the effects of strong shaking we need to find a location with more frequent large (but not megathrust) earthquakes. The Nootka transform fault that extends beneath the accretionary prism approximately orthogonal to the margin of central Vancouver Island provides an ideal opportunity (Figure 2 ). It generates M>6 events approximately at intervals of 20 years and M>4 every 4 month. Accelerations are estimated to occur at 1% g every 3 – 12 month and 5 % g every 2 – 10 years. This part of the proposal involves only one site, with the key installation of an ACORK to monitor time variations in pore pressure and fluid flux associated with earthquake shaking. The NEPTUNE seafloor cable observatory system proposed for the Juan de Fuca plate may provide an ideal method for long-term monitoring (see NEPTUNE web site http://www.neptune.washington.edu/). SCIENTIFIC GOALS The proposal follows the goals for gas hydrate drilling of the ODP Gas Hydrates Program Planning Group, i.e., (1) Study the formation of natural gas hydrate in marine sediments; (2) Determine the mechanism of development, nature, magnitude and global distribution of gas hydrate reservoirs; (3) Investigate the gas transport mechanism, and migration pathways through sedimentary structures, from site of origin to reservoir; (4) Examine the effect of gas hydrate on the physical properties of the enclosing sediments, particularly as it relates to the potential relationship between gas hydrates and slope stability; (5) Investigate the microbiology and geochemistry associated with hydrate formation and dissociation. These scientific goals are an expansion of the latest achievements of ODP Leg 204, dedicated to study gas hydrates at Southern Hydrate Ridge (Trehu et al., 2002). Leg 204 was entirely focused on the specific structure of Hydrate Ridge and has only limited potential for applications at different continental margins. Figure2. Nootka fault seismicity.