Pore fluids are passively convecting through young oceanic sediments and crust around Deep Sea Drilling Project (DSDP) site 504 on the southern flank of the Costa Rica Rift, as inferred from a variety of geological, geochemical, and geothermal observations. The pressence of a fluid circulation system is supported by new data collected on Ocean Drilling Program (ODP) leg 111 and a predrilling survey cruise over the heavily sedimented, 5.9 Ma site; during the latter, elongated heat flow anomalies were mapped subparallel to structural strike, with individual measurements of twice the regional mean value, and strong lateral and vertical geochemical gradients were detected in pore waters squeezed from sediment cores. Also, there is a strong correlation between heat flow, bathymetry, sediments thickness, and inferred fluid velocities up through the sediments. On an earlier DSDP leg, an 8-bar underpressure was measured in the upper 200 m of basement beneath thick sediment cover. Although the forces which drive passive circulation are not well understood, it has generally been thought that the length scale of heat flow variations provides a good indication of the depth of hydrothermal circulation within the oceanic crust. This assumption was based on analytical and numerical analyses of relatively simple porous media models.

Deep crustal convection had been inferred near site 504 based on the geometry of surface heat flow anomalies with a wavelength of 4--7 km but appears to be precluded by low crustal permeability, as measured in DSDP hole 504B. The widely varied geothermal and hydrogeological observations near site 504 are readily explained by a model which combines (1) basement relief, (2) irregular sediment drape, (3) largely conductive heat transfer through the sediments overlying the crust, and (4) thermal and geochemical homogenization of pore fluids at the sediment/basement interface, which results from (5) topographically induced, passive hydrothermal circulation with large aspect ratio, convection cells. This convection involves mainly the permeable, upper 200--300 m of crust; the deeper crust is not involved. This convection is induced through a combination of buoyancy fluxes, due to heating from below, and topographic variations on the seafloor and at the basement-sediment interface. This model was designed to incorporate data from both near the sediment surface and deep within boreholes; it is sucessfull in duplicating numerous field observations. ¿American Geophysical Union 1990