Bobb Carson

Methane-derived authigenic carbonates formed by subduction-induced pore-water expulsion along the Oregon/Washington margin

Authigenic magnesian calcite, dolomite, and aragonite are precipitated in the uppermost terrigenous sediments of the Washington/Oregon accretionary prism by subduction-induced dewatering. These distinctive carbonates are methane-derived and occur at sites of concentrated pore-water expulsion. Unique biologic communities that subsist, at least indirectly, on methane (Suess et al., 1985) are also found at some of these sites. The methane, which is dominantly biogenic, is carried to the uppermost sediments of the prism by fluids and is oxidized by sulfate reducers before being incorporated into a carbonate cement. Carbonate precipitation occurs below the oxic layer, probably no deeper than several centimetres to a few metres below the seabed. Cementation may be induced by three factors: (1) increased carbonate alkalinity resulting from microbial sulfate reduction, (2) decreased σCO 2 solubility resulting from a pressure decrease when the pore water escapes the prism, and/or (3) the addition of Ca 2+ and Mg 2+ ions from sea water near the sediment/water interface. The convergent margin setting engenders precipitation of authigenic carbonates in several ways. Compressive stresses induce anomalously rapid compaction and dewatering rates, and they may cause overpressuring in migrating pore water, thereby delaying precipitation of carbonates until pressure is released near the sediment-water interface. Structural deformation of the accretionary prism creates pathways (such as fault zones), secondary fracture porosity, and dipping permeable layers (often exposed by mass movement) for efficient advection and expulsion of methane-enriched pore water. These characteristic conditions, which lead to the precipitation of methane-derived carbonates, may be found at other convergent margins.

Fluid flow in accretionary prisms: Evidence for focused, time‐variable discharge

Accretionary prisms are wedges of saturated sediment that are subject to intense deformation as a result of lithosphere convergence. Compressive stress and rapid burial of the accreted deposits result in sediment compaction and mineral dehydration. These latter processes, in conjunction with fermentation or thermal maturation of entrained organic matter, yield hydrocarbon-bearing pore fluids that are expelled from the prism. Regional fluxes of heat and a number of dissolved chemical species, most notably carbon, are controlled by the advective expulsion of the pore waters. Numerical modeling, observation and monitoring of flow patterns and rates, and recent in situ hydrogeological tests quantify the conditions that control rates of fluid flow. Dispersed, intergranular flow (10−8 to 10−11 m/s), controlled by the vertical permeability of the prism (10−14 to 10−20 m²), is limited by low-permeability lithologies and seems not to vary much from margin to margin. Focused flow (10−1 to 10−8 m/s) above the decollement is controlled by fault zones or sedimentary intrusions (diapiric structures). At low-fluid pressures, fault zone permeability may be similar to that of adjacent wall rock, but as fluid pressure increases from hydrostatic (λ* = 0) to near lithostatic levels (λ* ≈ 1.0), fault zones dilate, and (fracture) permeability increases by 2–4 orders of magnitude (10−10 to 10−16 m²). Similarly, mud volcanoes and diapirs provide high-permeability fluid conduits to the sediment-water interface. As a result, faults and intrusions become primary flow paths and support surface vents at which syntectonic deposits (carbonate and gas hydrates) accumulate and chemosynthetic organisms cluster. Models of thermal and chemical anomalies and epigenetic deposits indicate that flow is temporally variable. That conclusion has been quantified by extended (1–10 months) seafloor and borehole experiments that measured temperature anomalies associated with flow events. On the Cascadia prism, flow (estimated velocity ∼3 × 10−5 m/s; 950 m/yr), confined to a thrust fault that cuts upward through the prism, has brought thermogenic hydrocarbons from depths >1.5 km to the surface within the last 400 years.

Fluid expulsion sites on the Cascadia accretionary prism: Mapping diagenetic deposits with processed GLORIA imagery

Point-discharge fluid expulsion on accretionary prisms is commonly indicated by diagenetic deposition of calcium carbonate cements and gas hydrates in near-surface (<10 m below seafloor; mbsf) hemipelagic sediment. The contrasting clastic and diagenetic lithologies should be apparent in side scan images. However, sonar also responds to variations in bottom slope, so unprocessed images mix topographic and lithologic information. We have processed GLORIA imagery from the Oregon continental margin to remove topographic effects. A synthetic side scan image was created initially from Sea Beam bathymetric data and then was subtracted iteratively from the original GLORIA data until topographic features disappeared. The residual image contains high-amplitude backscattering that we attribute to diagenetic deposits associated with fluid discharge, based on submersible mapping, Ocean Drilling Program drilling, and collected samples. Diagenetic deposits are concentrated (1) near an out-of-sequence thrust fault on the second ridge landward of the base of the continental slope, (2) along zones characterized by deep-seated strikeslip faults that cut transversely across the margin, and (3) in undeformed Cascadia Basin deposits which overlie incipient thrust faults seaward of the toe of the prism. There is no evidence of diagenetic deposition associated with the frontal thrust that rises from the decollement. If the decollement is an important aquifer, apparently the fluids are passed either to the strike-slip faults which intersect the decollement or to the incipient faults in Cascadia Basin for expulsion. Diagenetic deposits seaward of the prism toe probably consist dominantly of gas hydrates.

Tectonically induced deformation of deep-sea sediments off Washington and northern Oregon: Mechanical consolidation

Abstract Convergent motion of the North American and Juan de Fuca plates has resulted in deformation of Cascadia Basin sediments and accretion of these deposits to the North American continental margin. The accreted deposits, which occur as anticlinal ridges and thrust blocks, constitute the lower continental slope or borderland off Washington and northern Oregon. Over the past 2.0 m.y., approximately 30 km of this deformed material has been added to the lower slope, removing undeformed deposits from Cascadia Basin at a rate of 2.3–2.9 cm/yr. Near-surface sediments involved in this accretionary process are mechanically consolidated: mudstones dredged from the lower slope exhibit physical properties (water contents, 20–47%; void ratios, 0.4–1.2; preconsolidation pressures, 0.8–8.2 MPa) which differ significantly from properties of similar, but undeformed sediments (water contents, 50–250%; void ratios, 1.1–1.9). While some consolidation may be attributable to prior burial ( MPa ) or carbonate precipitation, neither mechanism can wholly account for the values observed. It appears that most of the consolidation has occurred in response to tectonically induced overpressures. Initial consolidation occurs rapidly across a narrow ( km ) front, defined by the base of the continental slope. Further consolidation and dewatering appears to take place, at a much reduced rate, over the entire width of the lower slope. Development of foliation is nearly ubiquitous in the deformed mudstones. This property limits the strength of the deposits (shear strengths, 90–416 kPa) and movement along these planes probably accommodates much of the strain after initial consolidation. The physical properties characteristic of Washington—Oregon deformed sediments may represent limiting values for mechanical consolidation of near-surface terrigenous sediments under horizontal stress.

Modern sediment dispersal and accumulation in Quinault submarine canyon — A summary

Abstract Quinault Canyon, off the coast of Washington, intersects the dispersal path of modern Columbia River sediments on the continental shelf. Study of the dynamics, dispersal, and accumulation of these deposits in the canyon indicates that no systematic down-canyon transport occurs. Rather, sediment dispersal is controlled primarily by the regional flow pattern. Suspended particulate matter is resuspended on the shelf, primarily during winter storm events, and advected northward as a bottom nepheloid layer (BNL) over the shelf. Cross-canyon advection produces an intermediate nepheloid layer (INL) over the upper slope and canyon from which the suspended particulate matter settles rapidly, probably in amorphous aggregates. Rapid short-term deposition ( ∼ 10 −1 g cm −2 yr −1 ) and enhanced modern accumulation (generally 10 × 10 −2 g cm −2 yr −1 ) are confined to the upper canyon (1600 m), where fluctuations in deposition rates and grain size are related to specific shelf sediment resuspension events. In the lower canyon (1600 m), accumulation is slow and temporally uniform.

Initial Deep-Sea Sediment Deformation at the Base of the Washington Continental Slope: A Response to Subduction

Compression of flat-lying Cascadia Basin (continental rise) sediment against the Washington continental slope, resulting from underthrusting of the Gorda–Juan de Fuca plate beneath the North American plate, progresses from passive uplift and anticlinal upwarping to thrust faulting. When folding and thrusting at the continental slope-rise boundary can no longer accommodate the rate of compression, the locus of deformation shifts offshore and a new anticline begins to form. Repetition of this deformational cycle has produced the multiple ridges of the lower continental slope and westward progradation of the continental margin. The most recent anticlinal upwarping apparently began 0.11 to 0.35 m.y. ago. Consolidation of the ridge sediment during the deformation process may be tectonically induced.

Hydrogeologic properties of a thrust fault within the Oregon Accretionary Prism

Two sets of hydrogeologic tests conducted at Ocean Drilling Program (ODP) Hole 892 on the Oregon Accretionary Prism provided the opportunity to determine hydrogeologic properties of an active accretionary prism fault zone. The first set of tests consisted of shipboard packer tests conducted during ODP Leg 146 (fall 1992), while the second set of tests were constant-drawdown and constant-discharge tests conducted in fall 1993 using the submersible Alvin. Pressure response during the first set of tests suggests that fractures remained open until excess fluid pressure (relative to hydrostatic) dropped below 0.315 to 0.325 MPa (λ* ∼ 0.53 to 0.54, where λ* = (pore pressure - hydrostatic)/(lithostatic-hydrostatic)). Analysis of the packer test data suggested an apparent background pressure of 0.25 MPa (λ* ∼ 0.42 to 0.50). Because the borehole had been open for 12 hours prior to the packer tests, formation pore pressures may have exceeded this value prior to drilling of the borehole. These overpressures dissipated by the time the second set of tests were conducted. One possible explanation for this decay is that the borehole may provide a vertical conduit between the overpressured zone and overlying or underlying sediments that had previously been hydraulically separated from the overpressured zone. The second set of tests were conducted at pressures (≤0.019 MPa or λ* ∼ 0.03) below that estimated to maintain open fractures and yielded transmissivities 1 to 2 orders of magnitude less than estimated for the packer tests (when fractures were open). Constraints on fluid flow rate along the fault are provided by observed displacement in a bottom-simulating reflector (BSR) at its intersection with the fault zone. The closed-fracture transmissivities are insufficient to produce flow rates capable of displacing the BSR; therefore open-fracture transmissivities under conditions of elevated pore pressure are inferred to be necessary for the observed BSR displacement. In addition, calculated rates of specific discharge through the fault zone are 2 to 3 orders of magnitude lower than discharge measured at an associated seafloor vent site; fluid flow must become spatially or temporally focused as it moves up the fault zone toward the seafloor.

Rethinking turbidite paleoseismology along the Cascadia subduction zone

A stratigraphic synthesis of dozens of deep-sea cores, most of them overlooked in recent decades, provides new insights into deepsea turbidites as guides to earthquake and tsunami hazards along the Cascadia subduction zone, which extends 1100 km along the Pacific coast of North America. The synthesis shows greater variability in Holocene stratigraphy and facies off the Washington coast than was recognized a quarter century ago in a confluence test for seismic triggering of sediment gravity flows. That test compared counts of Holocene turbidites upstream and downstream of a deep-sea channel junction. Similarity in the turbidite counts among seven core sites provided evidence that turbidity currents from different submarine canyons usually reached the junction around the same time, as expected of widespread seismic triggering. The fuller synthesis, however, shows distinct differences between tributaries, and these differences suggest sediment routing for which the confluence test was not designed. The synthesis also bears on recent estimates of Cascadia earthquake magnitudes and recurrence intervals. The magnitude estimates hinge on stratigraphic correlations that discount variability in turbidite facies. The recurrence estimates require turbidites to represent megathrust earthquakes more dependably than they do along a flow path where turbidite frequency appears limited less by seismic shaking than by sediment supply. These concerns underscore the complexity of extracting earthquake history from deep-sea turbidites at Cascadia.

BARBADOS RIDGE HYDROGEOLOGIC TESTS : IMPLICATIONS FOR FLUID MIGRATION ALONG AN ACTIVE DECOLLEMENT

Hydrogeologic tests were conducted at a sealed borehole penetrating the decollement of the Barbados Ridge accretionary complex. At low excess pore pressures [λ * = ( P p – P h )/ ( P l – P h ) = 0.0 to 0.36, where P p = pore pressure, P h = hydrostatic pressure, and P l = lithostatic pressure], estimated permeabilities were comparable to those of similar, unfractured sediment. These tests complement shipboard packer tests completed at higher fluid pressures (λ * = 0.5 to 1.0) during Ocean Drilling Program (ODP) Leg 156. Together, the test results suggest a 4- to 5-order-of-magnitude permeability increase as fluid pressure varied from hydrostatic (λ * = 0) to lithostatic (λ * = 1). However, unlike the results of the shipboard packer tests, the test results presented here exhibit no evidence of a relationship between permeability and pore pressure. The combined findings from the two sets of hydrogeologic tests indicate that significant permeability increases can occur within the decollement at pore pressures below lithostatic pressure.

Fluid expulsion from the Cascadia accretionary prism: evidence from porosity distribution, direct measurements, and GLORIA imagery

Fluid expulsion from the Cascadia accretionary prism off Oregon results from porosity reduction by compaction, and by cementation as methane-rich pore waters precipitate diagenetic carbonate deposits near the sediment-water interface. Porosity changes suggest that dewatering begins 5-6 km west of the base of the slope, in a proto-deformation zone, GLORIA imagery of surficial carbonate deposits confirms that fluid is actively expelled from this zone; there is no such evidence further west in Cascadia Basin. Within the uncertainties of the data, porosities do not decrease landward beneath the prism. This pattern is consistent with imbricate thrust faulting on the slope which provides the vertical load to induce compactive dewatering, and may physically import as much as 50% of the total fluid volume in the section. A simple vertical compaction model suggests that significant pore water volumes have been expelled from the lower slope, but at flux rates (10-11 -10-12 m3 m-2 s-1) which are orders of magnitude less than those measured at individual vent sites (10-6 m3 m-2 s-1). Faulting clearly controls some fluid expulsion, but GLORIA data suggest that repeated local discharge, cementation, and abandonment lead to dispersed accumulations of diagenetic carbonate.