Recent measurements from Ocean Drilling Program leg 110 and Deep Sea Drilling Project leg 78a indicate that pore pressures near the toe of the Barbados accretionary prism may be dose to lithostatic and that the d6collement is a zone with relatively high rams of fluid flow and methane transport. We used a numerical model of fluid flow to estimate intrinsic permeabilities, pore pressures, and flow velocities that are consistent with these observations. Model results suggest that the permeability of the d6collement may be 3-5 orders of magnitude greater than that of adjacent prism sediments. ff permeabilities in the prism vary with depth in a manner similar to those in sedimentary basins, the average intrinsic permeability of the d6collement, k d , must be about 10 -14 m 2. When k a is 10 -13 m 2, high pore pressures do not develop near the deformation front in the model. ff k a is 10 -15 m 2, simulated pressures are unrealistically high in both the prism and underthrust sediments arcward of the deformation front. Water originating from compaction in the d6collement and underthrust sediments flows laterally seaward, while water expelled from prism sediments flows upward to the ocean floor. However, flow velocities are small, and the net motion of pore water in prism and undenhrust sediments is arcward relative to the deformation front because of tectonic transport. Pore water migrates seaward in spite of tectonic transport only in discrete zones with higher permeability, in this case the d6collement. brmODUCT•ON 1982]. Both Ocean Drilling Project (ODP) leg 110 and Deep Sea Drilling Project (DSDP) leg 78a recovered sediment samples In accretionary complexes, gravitational and tectonic forces from the complex. These samples document decreases in porosity deform highly porous sediments as they are either accreted onto as basin sediments are incorporated into the accretionary the overriding plate or carried downward with the underthrust complex. Observations from DSDP leg 78A also suggest that plate. Fluids expelled from the compacting sediments influence pore pressures may reach near-lithostatic values just above the many aspects of subduction zone geology, including heat d6collement within 5 km of the deformation front [Moore and transport [e.g., Langseth and Hobart, 1984], diagenesis and Biju-Duval, 1984]. In addition, some pore water samples metamorphism [e.g., Etheridge et al., 1983], and benthic biology collected during ODP leg 110 contained anomalous methane, [e.g., Kulm et a/.,1986]. In addition, relative rates of sediment chloride, and temperature distributions, suggesting preferential loading and fluid dissipation determine the magnitude and fluid flow along fault zones, especially within the d6collement distribution of excess fluid pore pressures. These pore pressures [Moore et al., 1987; Gieskes et al., 1989]. affect the shape of the accretionary wedge [e.g., Davis et al., In this paper, we present a nmerical model of fluid flow in the 1983] as well as thrust fault and fold geometries [Hubbert and Barbados R...

Sediment compaction at convergent margins expels pore fluids, which in turn influence many aspects of subduction zone geology. Drilling in the Barbados Ridge complex during ODP Leg 110 and DSDP Leg 78A provided information about sediment types, porosities, and the geometry of the complex. In this paper, we use these observations to estimate the rates of sediment porosity loss and accompanying fluid expulsion from the prism, the decollement, and the underthrust sediments. Rates of porosity loss depend on how rapidly the sediments move arcward through the complex. We compute rates of sediment movement in the prism and decollement as a function of distance from the deformation front. This calculation assumes that the rates of sediment movement decrease arcward as the result of porosity loss in the prism and decollement and thickening of the prism. According to computed rates of sediment movement for the prism, sediment deceleration is greatest within the first 3 to 5 km arcward from the deformation front. Similarly, most dewatering of the prism sediments takes place in this region. Beyond 5 km, rates of sediment dewatering become great est in the underthrust sediments, assuming these sediments are carried downward at a uniform rate with the oceanic plate.

Measurements of sediment physical properties and pore-water chemistry gathered during ODP Leg 110 and DSDP Leg 78A suggest that (1) fluid flow in the decollement is predominantly updip, and (2) near-lithostatic pore pressures may exist just above the decollement at Site 542. We use these observations to constrain a numerical model of fluid flow in the toe of the complex. Gravitational and tectonic forces drive flow within the complex and are incorporated in the numerical model by esti mating the rate and distribution of fluid generation from sediment compaction. Modeling results reveal that near-litho static fluid pressures could form at Site 542 if the equivalent prism permeability is between I0-18 and I0-19 m 2 (hydrau lic conductivity of I0-9 to I0-10 cm/s). Predominantly horizontal flow in the decollement can occur if the decollement permeability exceeds the equivalent prism permeability by three to four orders of magnitude. Under these conditions, average linear fluid velocities in the decollement range between 3 and 10 cm/yr (1.5 to 5 times the convergence rate). Fluid velocities in the prism are about 1000 times slower. Depending on the permeability contrast between the decolle ment and the underthrust sediments, between 65% and 90% of the fluids expelled beneath the accretionary prism flow out of the complex through the decollement.