Proceedings of the 7th International Conference on Gas Hydrates (ICGH 2011), Edinburgh, Scotland, United Kingdom, July 17-21, 2011. DO DISSOCIATING GAS HYDRATES PLAY A ROLE IN TRIGGERING SUBMARINE SLOPE FAILURES? A CASE STUDY FROM THE NORTHERN CASCADIA MARGIN Nastasja A. Scholz* Michael Riedel George D. Spence School of Earth and Ocean Sciences Uni versity of Victoria, BC CANADA Roy D. Hyndman Thomas James Kathrin Naegeli Geological Survey of Canada - Pacific Sidney, BC CANADA Brandon Dugan Rice University Houston, TX UNITED STATES OF AMERICA John Pohlman US Geological Survey Woods Hole Coastal and Marine Science Center Woods Hole MA 02543 USA Tark Hamilton Department of Chemistry & Geoscience Camosun College Victoria, BC CANADA ABSTRACT Swath bathymetry from the northern Cascadia margin offshore Vancouver Island, Canada reveals several submarine landslides on the seaward slopes of the deformation frontal ridges. The slides occur in the thick accretionary prism of the subducting Juan de Fuca and Explorer plates. Possible trigger mechanisms for the slope failures include strong earthquake shaking, pore pressure changes induced by sea-level changes, and the dissociation of gas hydrates. This study focuses on two slide features of different morphology, representing the two end-members of the range of morphologies found on the margin. Orca Slide resembles a debris flow, while Slipstream Slide consists of large internally coherent sediment blocks. The presence of gas hydrates is interpreted *Corresponding author: Phone: +12507216188 E-mail: nscholz@uvic.ca
from of a bottom simulating reflector (BSR) in regional seismic data around both slides. Near Orca Slide additional evidence for gas hydrate is available from core and logging data of Integrated Ocean Drilling Program (IODP) Expedition 311, Site U1326. We compared the timing of rapid local sealevel change with the results from radiocarbon age dating and sulfate-reduction rate modeling of previous studies on collected sediment cores to constrain the ages of the slides and potential trigger mechanisms. A link between slump-failures and gas hydrate dissociation is excess pore pressure. We determine the amount of excess pressure needed to trigger slope failure by examining sea-level changes, tectonic uplift, sedimentation rate, and sediment physical properties such as pore pressure, sediment shear strength, porosity, sediment permeability, and gas hydrate concentration. The estimation of the present state of slope stability, the reconstruction of the slide dynamics and of the subsequent forces on the overlying water column have the potential to contribute to the assessment of local tsunami hazard. Keywords: gas hydrate dissociation, hazard analysis, slope failure, tsunami, numerical modeling INTRODUCTION Possible trigger mechanisms for submarine landslides are abundant and include earthquake shaking, rapid sedimentation, fluid seepage, slope over-steepening, gas charging, and gas hydrate dissociation. Links between submarine landslides and gas hydrate dissociation have been examined extensively in the last three decades (e.g. [1], [2], [3], [4], [5], [6]). An understanding of submarine landslides is important because of their tsunamigenic potential and because of their impact on offshore infrastructure such as underwater cables. This study aims to assess the trigger mechanism and the role of gas hydrates in slope failure at the deformation front of the Cascadia subduction zone offshore Vancouver Island. Numerical modeling of the submarine landslide dynamics in this region will concentrate on the prevalent failure mechanisms and the factors controlling slide morphology. Estimations of the temporal evolution of velocity, acceleration, mass distribution, and run-out can then help to assess the induced forces on ambient water to assess the potential generation of tsunamis. NORTHERN CASCADIA MARGIN This study focuses on two slope failures on frontal ridges situated at the deformation front of the northern Cascadia subduction zone (Fig. 1). The Cascadia convergent margin offshore of the Canadian southwest and US-American northwest coast stretches from the Queen Charlotte Islands approximately north of Vancouver Island down to northern California. The entire length of the Juan de Fuca oceanic plate is currently subducting beneath the North America continental plate at a rate of about 46 mm/a. The young, buoyant subducting lithosphere has accreted a 3-4 km thick sediment section to the North America plate, forming a large accretionary wedge in relatively short geologic time. The process of accretion led to vertical thickening and bulk shortening of the we dge material as well as thrust faulting. The present day structure of the northern Cascadia margin has generated numerous mega-thrust earthquakes, with a recurrence interval of 300-700 years [7]; the last event was a magnitude Mw=9 earthquake in 1700 [8]. The continental slope of the northern Cascadia margin is also the location of gas hydrate occurrences, which extend to the frontal ridge near the deformation front (e.g. [9]; [10]; [11]). Several ODP/IOPD studies have focused on assessing the the nature of the gas hydrate system ([12]; [13]; [14]), and on gas hydrate concentration (e.g. [9]; [15]; [16]). Frontal ridges and their morphological appearance Processes such as offscraping, folding, and faulting that accompany subduction led to the development of anticlinal ridges along the deformation front. These ridges are discontinuous along the western coast of Canada, with average lengths of only 20-30 km, widths of a few kilometers, and heights of up to 600m above the surrounding seafloor. All of these ridges are approximately parallel to the margin but can deviate as much as 30 from this margin-parallel trend. Slope angles are consistently steeper
towards the incoming plate than landward as a consequence of the local tectonic regime (e.g. [17]). The sedimentary section just seaward of the frontal ridges and the deformation front comprises two main sedimentological sub-units. One unit consists of a rapidly deposited turbidite section composed of clayey silts, fine sands, and diagenetic carbonates of Pleistocene age. This unit reveals horizontal bedding in the upper 130 m and highly fractured bedding that dips with angles of 40-70 in the lower 170m. The deeper unit is a fine-grained Pliocene section of pervasively fractured hemipelagic material containing abundant authigenic glauconite ([12]). Se diment composition and the existence of fractures both might have influence on slide initiation, propagation, and final deposit characteristics. The slide deposits reveal different morphologies despite being situated only ~10 km apart. Two examples for the end members of a range of different morphological appearances are the informally called Orca Slide and Slipstream Slide (Fig. 1). Orca Slide has the overall appearance of a debris flow with its material potentially been fluidized in the process of sliding. Slipstream Slide is a blocky slide, with slump material having moved as a coherent and intact mass. Eight slides were analyzed in detail by [18] using bathymetry data providing constraints on failure area, mean slope angle, and estimates of mean slide volume. For Orca Slide and Slipstream Slide, results indicate failure areas of 2.7 km 2 and 5.0 km 2, mean slope angles of 17 and 16 (with local slope angles potentially being much steeper), and mean volumes of 0.41 km 3 and 0.25 km 3, respectively. For Orca Slide, IODP Site U1326 (Fig. 1) provides control on the sediment properties such as porosity, shear strength, p-wave velocity, and grain-size distribution. Figure 1: The Cascadia subduction zone (modified after [14]); yellow star indicates location of IODP Site U1326; scale holds for both frontal ridges A CONNECTION BETWEEN GAS HYDRATE OCCURRENCE AND SLOPE FAILURE? At the northern Cascadia margin, the area of inferred and mapped BSR includes the location of all the frontal ridges. Water depth in this area is in excess of 2000 m. Failure planes reveal steep angles and an earlier study of the Orca Slide showed the presence of nearly vertical normal faults striking perpendicular to the margin, delineating the failure area, and extending downwards through the GHSZ to the BSR (at ~250 mbsf) and deeper [19]. Results of IODP Expedition 311 and other studies conducted at the margin suggest that gas hydrates offshore Vancouver Island are mainly formed from biogenic methane ([11]). While average gas hydrate concentrations are as low as 5%, it was found that there are localized high concentrations of 50% and above, mainly in sandy turbidites. Higher concentrations are also found in fractures, veins, and nodules and thus hydrates are very heterogeneous in their distribution. Especially abundant gas hydrate occurrence is found in cold vents of this region.
Since gas hydrates are located on the continental slope and if present in sufficient amounts may increase sediment strength, their dissociation possibly influences slope stability. The boundary of the gas hydrate stability zone (GHSZ) can be affected by hydrostatic pore pressure changes due to a sea-level decrease, tectonic uplift, rapid sedimentation or erosion, or changes in sea bottom temperatures that occurred during the Pleistocene warming period. Decomposition of gas hydrate releases water and free gas, thus generating excess pore pressure. Excess pore pressure generated is a function of the amount of gas hydrate present initially, the rate of gas hydrate dissociation, sediment permeability, background fluid flux, capillary pressure, depth of the BGHSZ to the seafloor, and water depth (e.g. [20]). The most important factor controlling the magnitude of pore pressure generation is the hydraulic permeability (e.g. [21]; [22]). It influences the rate of pressure dissipation and has a complex relationship with gas hydrate presence and habit and the host sediment. Coarse grained sediment which is favourable for gas hydrate growth seems to have a decrease in permeability when gas hydrate is formed thus preventing or at least decelerating the dissipation of excess pore pressure. Fine-grained sediments in turn are more susceptible to hydraulic fracturing due to the high gas pressure needed to overcome capillary entry pressures. This would open or re-open fluid pathways and thus locally increase bulk sediment permeability. Spatial differences in permeability (high permeability fractures, variable permeability sands, and low permeability fine-grained matrix) potentially influences the distribution of over- and under-pressured regions (e.g. [2]). The BSR which separates gas hydrate-bearing sediment from sediments with some free gas also represents a potentially weakened zone along which pore pressure-induced strength contrast. It has been observed that failure plane and BSR approximately match in their respective depths ([19]). SLOPE STABILITY MODELING The aim here is to assess the trigger mechanism for both slides and to estimate the influence of gas hydrate dissociation on failure initiation and morphology. The question to ask is which conditions would have to be met to bring a slope to failure. Several scenarios are tested including the influence of steep slope angles, earthquake shaking, and localized increase in pore pressure. Since trigger mechanisms are abundant and the probability of one mechanism to be the trigger may differ during different geologic periods, estimating the age of the slides is important to narrow down possible causes for slope failure. Slope stability analysis is used to assess the state of slopes and to quantify the forces needed to bring a slope to failure (e.g. [23]). These are important aspects for the estimation of the potential role that gas hydrate may play in slopefailure initiation. We employ two different techniques to assess the influence of several triggering factors on slope stability. One method simulates the evolution of basin fluid flow and the possible generation of localized overpressure based on basin geometry and sedimentation history. In the second method we estimate the forces needed to induce slope failure using a limit equilibrium approach which includes the effect of earthquake accelerations. Pore pressure modeling The basin flow software BASIN2 [24] calculates the evolution of overpressure and fluid flow regimes in basins based on the assumption that pressure from the overburden drives sediment consolidation. If the rate of sediment loading exceeds the rate of fluid expulsion, pore fluid pressure increases above hydrostatic pressure. To predict fluid pressure evolution, BASIN2 couples Darcy s law, conservation of fluid mass, sediment loading, and pore fluid thermal expansion and uses information about local topography, stratigraphic architecture, sedimentation rate, and sediment compressibility and permeability [24]. Figure 2a shows a preliminary result for the calculation of overpressure for the Orca Slide, using a simplified geometry of the frontal ridge and a constant sedimentation rate of 200m/Ma. The factor of safety (FS) can be used to estimate criteria for failure initiation. The FS is defined as the ratio between driving and resisting stresses. This involves calculating force and moment equilibria and searching for the limiting force at which slopes are still stable (FS equals unity). The stability analysis (Fig. 2b) indicates that the FS for Orca Slide is much too high for slide initiation due to sedimentation history and local topography. To further improve the results in Fig. 2, detailed bathymetric profiles across the two frontal ridges will help to constrain the present day geometry of the failure plane.
a) b) Figure 2: Preliminary results using a simplified geometry of the Orca Slide ; a) overpressure; b) factor of safety Influence of earthquake shaking Earthquake shaking is incorporated by adding the forces due to critical horizontal acceleration and the vertical effective stress both representing driving stresses [24]. The infinite FS calculation with earthquake acceleration is: FS c 'vh cos2 P * tan f ' vh cos sin Feq (1) Here c is cohesion, σ vh is hydrostatic vertical effective stress, P* is overpressure, Θ is seafloor slope angle, Φf is internal friction angle, and F eq is the earthquake acceleration shear stress parallel to the slope (e.g. [25]). Critical horizontal shaking kcrit with the FS equal to unity can then be calculated by using Feq k crit v cos2 (2) with σv a s the total vertical stress. Replacing F eq in (1) with (2) and solving for kcrit the intensity of an earthquake that is needed (FS=1) can be calculated. The resulting critical horizontal acceleration then has to be related to the peak spectral acceleration kpsa. [24], using the relationship kpsa = kcrit /(0.15*3.5). The chosen values in the above relationship are appropriate for California and the Gulf of Mexico ([24]). Both factors are functions of earthquake recurrence intervals and local soil conditions and have to be adjusted according to the local tectonic setting, which has yet not been achieved for the area of this study on the Cascadia margin. Future research is necessary to better define the relation between k PSA and k crit for Cascadia. As a next step ground-motion attenuation relationships can be used to calculate maximum distances of earthquakes of a certain magnitude to cause failure. As megathrust earthquakes have occurred in the past with assumed recurrence rates of ~ 500 years, it is likely that one or more such high magnitude events have triggered slope failure at the deformation front. A preliminary result estimating pairs of peak ground acceleration and distance for a hypothetical magnitude M w=8.5 earthquake are shown in Fig. 3. We calculated the potential for slope failure during critical shaking using the method of slices, which divides the ridge into a series of slices, each 30 m in width and extending to a depth corresponding to the thickness of the slide material above the gliding plane. This provides a more rigorous assessment of failure in 2D as compared to the 1D, infinite slope approximation. The ground motion attenuation relation of [26] was used to predict the peak ground acceleration at a given hypocentral distance for an earthquake of magnitude 8.5. This attenuation relationship, which is based on a stochastic finite-fault model, is appropriate for the Cascadia margin and for megathrust earthquakes at distances as small as 50 km or less. The estimated peak ground acceleration is 0.3g for an earthquake source at a distance of 40 km from the frontal ridge. This is the approximate distance of the frontal ridge from the top of the continental slope, which roughly marks the most seaward extent of the seismogenically locked zone where a megathrust earthquake can be generated [27].
Figure 3: Preliminary result for the peak ground acceleration caused by a hypothetical magnitude 8.5 earthquake with distance (ground motion attenuation relationship according to [26]); black line showing k PSA with hypocentral distance; the blue line indicates location of the updip limit of the locked zone; the green dashed line indicates estimated peak ground acceleration for an earthquake of magnitude 8.5 at this location Gas hydrate stability conditions Sediment cores taken in 2008 have been used for radiocarbon age dating and results suggest an age of the slides of at least 9,000 years but at most 14,000 years ([28]). This time interval is in good agreement with the ages of the turbidite sequences used to date the occurrence of past megathrust earthquakes on the Cascadia subduction margin ([7]). Additionally, sulfate-reduction-rate modeling on pore-water sample chemistry have led to similar age estimations ([29]; [30]). Another source for age constraints can be derived from the history of local, relative sea-level change because it represents important information on the temporal development of the overall gas hydrate stability conditions in the sediments. Values for relative sea-level take into account the effects of glacial rebound after ice masses have retreated following the end of an ice age. During the Last Glacial Maximum, Vancouver Island was covered by the Laurentide ice sheet which had its peak extent around 20,000 years BP. Deglaciation was rapid around 14,000 years BP ([31]) and sea-level changed rapidly between 15,000 and 8,000 years BP, and especially between 14,000 and 12,500, as we ll as bet ween 11,000 and 10,000 years BP. Glacial isostatic rebound (GIA) can be modeled numerically ([32]) and results on the relative sealevel curve for the specific locations of the slumps were used as input to estimate the role of changing sea-level on gas hydrate stability. The effects of these changes on the top and bottom of the GHSZ were estimated by calculating the BSR depth using the local P-T-st ability curve (modified after [33]) using values for the salinity and geothermal gradient as established at Site U1326. To calculate the local upper limit of the GHSZ the method of [34] was used. The assumption herein is that the upper limit of hydrate occurrence lies at a depth at which methane concentration exceeds methane solubility. This involves the calculation of mass balance of methane as it is generated and transported in the sediment column. Microbial methane production and methane diffusion, as well as the sedimentation rate are taken into account as boundary conditions. Results show that changes in overburden pressure equivalent to a decrease in sea-level by 120m lead to a BSR depth ~10m shallower than today if instantaneous reactions to the change in pressure are assumed (Fig. 4). In contrast, the calculated upper limit of gas hydrate occurrence did not change significantly with changing sea-level. Fig.4: Comparison bet ween calculated BSR depths for present day conditions (black) and 15,000 BP (red) for Orca Slide; depth are in meter below sea-level (mbsl) The thermal gradient may still be in the process of thermal re-equilibration since temperatures at the failure plane have suddenly been exposed to sea water temperatures at the time of slumping. The depth-difference between a theoretical pre-slide BSR and the seismically inferred BSR today can be used to calculate the time that the phase boundary would need to move the observed distance. Pore pressure increase due to ensuing gas hydrate dissociation can be quantified. A geomechanical model which relates volume change with the generation of excess pore pressure (e.g. P* in Equation 1) taking into account equilibrium temperature and pressure of gas hydrate at standard conditions, porosity, a confined compression modulus for soil, and the degree of
saturation was used to evaluate slides in the Beaufort Sea and the Gulf of Mexico [35]. This approach can be employed to define the pore pressure and stability response of the frontal ridges off Cascadia due to gas hydrate dissociation caused by environmental change. CONCLUSIONS Two submarine landslide events located at the deformation front of the northern Cascadia margin are investigated. Possible trigger mechanisms for an active, convergent margin such as earthquake shaking and over-steepening are taken into account. The possible role of the dissociation of gas hydrate in the initiation of the slides and in slide propagation will be quantified. Gas hydrate concentration estimates are constrained by seismic, logging, and coring data and available high resolution bathymetry helps to define slope geometries. As an important input for slide age estimation the local sea-level history is used to calculate the boundary of the GHSZ over time. With reduced pressure due to a sea-level low-stand near the end of the Last Glacial Maximum, BSR sub-seafloor depths were shallower by ~10 m compared to the current depths. Effects on the top of gas hydrate occurrence zone appear to be negligible. Rapid sea-level changes occurred REFERENCES [1] McIver R.D., 1982, Role of naturally occurring gas hydrates in sediment transport, AAPG Bulletin, 66 [2] Xu W. & Germanovich L.N., 2006, Excess pore pressure resulting from methane hydrate dissociation in marine sediments: A theoretical approach, Journal of geophysical research, Vol.111 [3] Paull C.K., Ussler W. III., Dillon W.P., 1991, Is the extent of glaciation limited by marine gas hydrates?, Geophys. Res. Lett., 18, 432-434 [4] Iverson R.M., 1997, The physics of debris flows, Review of geophysics, 35(3), 245-296 [5] McDougall S., Hungr O., 2005, Dynamic modelling of entrainment in rapid landslides, Can. Geotech. J., 42, 1437-1448 [6] Locat J., Demers D., 1988, Viscosity, yield stress, remolded strength, and liquidity index relationships for sensitive clays, Can. Geotech. J., 25, 799-806 [7] Goldfinger C., Nelson C.H., Johnson J.E., 2003, Holocene earthquake records from the Cascadia subduction zone and northern San Andreas Fault based on precise dating of offshore turbidites, Annu. Rev. Earth Planet. Sci., 31, 555-577 around 9,000-10,000 BP and 13,000-14,500 BP, which compares we ll with estimated slide ages from 14 C radiocarbon dating, sulphate-reductionrate modeling, and from the comparison of turbidite sequences used to date the occurrence of past megathrust earthquakes. As a first step a simplified geometry and sedimentation history has been used to calculate overpressure generation. Preliminary results indicate insufficient amounts sedimentation-induced overpressure to cause slope failure. Preliminary results on slope stability are derived from a factor of safety analysis. Critical horizontal earthquake shaking was calculated and minimum distances for a hypothetical magnitude 8.5 event were estimated. Next st eps will use improved geometry and sedimentation constraints to recalculate overpressure and factor of safety values. Numerical modeling of the slide process will then focus on the development of slide morphology with time. The aim is to investigate prevalent failure mechanisms on the Cascadia margin and to assess possible interaction of slide mass with gas and water released by gas hydrate dissociation. Further research will quantify effects of the slide propagation, on the overlying water column, and the subsequent potential for tsunami generation. [8] Hyndman R.D., 1995, Giant earthquakes of the Pacific Northwest, Scientific American, 273(6), 68-75 [9]Hyndman R.D., Davis E.E., 1992, A mechanism for the formation of methane hydrate and seafloor bottom-simulating reflectors by vertical fluid expulsion, J. Geophys. Res., 97(B5) 7025-7041 [10] Hyndman R.D. & Spence G.D., 1992, A Seismic Study of Methane Hydrate Marine Bottom Simulating Reflectors, Journal of Geophysical Research, 97(B5), 6683-6698 [11] Riedel M. et al., 2006, Gas hydrate on the northern Cascadia margin: regional geophysics and structural framework, Proceedings of the Integrated Ocean Drilling Program, Volume 311 [12] Westbrook G.K., Carson B., Musgrave R.J. et al., 1994, Proc. ODP, Init. Repts., 146 (Pt. 1): College St ation, TX (Ocean Drilling Program). doi:10.2973/ odp.proc.ir.146-1.1994 [13] Tréhu A.M., Bohrmann G., Torres M.F., Colwell F.S., 2006, Leg 204 Synthesis: gas hydrate distribution and dynamics in the central Cascadia accretionary complex, In: Tréhu A.M., Bohrmann G., Torres M.F., Colwell F.S. (Eds.), Proceedings of the Ocean Drilling Program, Sc ientific Results Volume 204, 1-40
[14] Riedel M., Collett T.S., Malone M., 2010, Expedition 311 synthesis: scientific findings, Proc. IODP, 311, doi:10.2204/iodp.proc.311.213.2010 [15] Yuan T., Hyndman R.D., Spence G.D., Desmons B., 1996, Velocity structure of a bottomsimulating reflector and deep sea gas hydrate concentrations on the Cascadia continental slope, Journal of Geophysical Research, 101, 13655-13671 [16] Yuan T., Spence G.D., Hyndman R.D., Minshull T.A., Singh S.C., 1999, Seismic velocity studies of a gas hydrate bottom-simulating reflector on the northern Cascadia continental margin, Amplitude modeling and full waveform inversion, Journal of Geophysical Research, 104, 1179-1191 [17] Davis E.E. & Hyndman R.D., 1989, Accretion and recent deformation of sediments along the northern Cascadia subduction zone, Geol. Soc. of Am. Bull., 101, 1465-1480 [18] Naegeli K., 2010, Internship Report, Geological Survey of Canada, Sidney, BC [19] Lopez C., Spence G.D., Hyndman R.D., Kelley D., 2010, Frontal ridge slope failure at the northern Cascadia Margin-normal fault and gas hydrate control, Geology, 38(11), 967-970 [20] Crutchley G. et al., 2009, The potential influence of shallow gas and gas hydrates on sea floor erosion of Rock Garden, an uplifted ridge offshore of New Zealand, Geo-Mar Lett., DOI 10.1007/s00367-010-0186-y [21] Nimblett J., Ruppel C., 2003, Permeability evolution during the formation of gas hydrates in marine sediments, J. Geophys. Res., 108(B9): 2420 [22] Kleinberg R.L., Flaum C., Griffin D.D., Brewer P.G., Malby G.E., Peltzer E.T., Yesinowski J.P., 2003, Deep Sea NMR: methane hydrate growth habit in porous media and its relationship to hydraulic permeability, deposit accumulation, and submarine slope stability, J. Geophys. Res. 108(B10): 2508 [23] Spencer E., 1967, A method of analysis of embankments assuming parallel inter-slice forces, Geotechnique, 17, 11-26 [24] St igall J & Dugan B., 2010, Overpressure and earthquake initiated slope failure in the Ursa region, northern Gulf of Mexico, Journal of Geophysical Research, Vol. 115, B04101, doi:10.1029/2009jb006848 [25] ten Brink U.S., Lee H.J., Geist E.L., Twichell D., 2009, Assessment of tsunami hazard to the U.S. East Coast using relationships between submarine landslides and earthquakes, Mar. Geol., 264(1-2), 65-73 [26] Gregor N.J., Silva W.J., Wong I.G., Youngs R.R., 2002, Ground motion attenuation relationships for the Cascadia subduction zone megathrust earthquakes based on a stochastic finite-fault model, Bull. Seismol. Soc. Am., 92(5), 1923-1932 [27] Wang, K., R. Wells, S. Mazzotti, R. D. Hyndman, and T. Sagiya, A revised dislocation model of interseismic deformation of the Cascadia subduction zone, J. Geophys. Res., 108(B1), 2026, doi:10.1029/2001jb001227, 2003. [28] Enkin R.J., Hamilton T.S., Riedel M., Pohlman J., 2010,Gravity driven slumps and turbidites from the frontal ridge of Cascadia s accretionary wedge Canada: controls from deglacial sedimentation, tectonics, and gas hydrates, EOS Trans. AGU, 91(26), Jt. Assem. Suppl., Abstract OS3A-06 [29] Pohlman J.W., 2006, Sediment biochemistry of northern Cascadia margin shallow gas hydrate systems, PhD Thesis, The College of William and Mary in Virginia [30] Pohlman J.W. et al., 2009, Methane sources and production in the northern Cascadia margin gas hydrate system, Earth and Planetary Science Letters, 287, 504-512 [31] Dyke A.S., 2004, An outline of North American deglaciation with emphasis on northern Canada, in: Quaternary glaciations extent and chronology, part II. North America, Ehlers J. and Gibbard P.L. (eds.), Developments in Quaternary Sc ience 2, 373-424 [32] James T. et al., 2009, Sea-level change and paleogeographic reconstructions, southern Vancouver Island, British Columbia, Canada, Quaternary Science News, 28, 1200-1216 [33] Sloan E.D., 1998, Clathrate hydrates of natural gases, Marcel Dekker, New York Spence G.D., Hyndman R.D., Chapman N.R., Wright F., 1992, [34]Malinverno A., 2010, Marine gas hydrates in thin sand layers that soak up microbial methane, Earth and Planetary Science Letters, 292(3-4), 399-408 [35] Nixon M.F., Grozic J.L.H., 2007, Submarine slope failure due to gas hydrate dissociation: a preliminary quantification, Can. Geotech. J., 44, 314-325