Geothermal processes at the Galapagos Spreading Center

Abstract

The measurements of the heat flow field of the Galapagos Spreading Center in an area of about 570 km2 reveal the planform of the conductive flux and permit the first truly areal estimate of the near-axis heat flux for comparison with theoretical plate cooling models. The intrusion process and associated hydrothermal circulation dominate the surface heat flow pattern, with circulation apparently continuing beyond the limits of our survey. The areal average of the conductive heat flux is 7.1 ± .8 HFU (295± 33 mW/m2), about one-third the heat flux predicted by plate models. The remaining heat is apparently removed by venting of hydrothermal waters at the spreading axis and through basalt outcrops and hydrothermal mounds off-axis. The pattern of surface heat flux is lineated parallel to the axis and the strongly lineated topography. Sharp lateral gradients in heat flow, greater than 10 HFU/km near escarpments and commonly expressed as high heat flow at the tops of scarps and lower heat flow in the valleys, may indicate a local concentration of the circulation by surface fault systems and/or variable sediment thickness. The mounds of the Galapagos Rift are spectacular hydrothermal features. Their internal temperatures have been mesured at up to 13°C above the bottom water temperature and total heat flow (conducted plus convected) can be several hundred times the normal oceanic value. Fluids, when they discharge from the mound, do so at a very slow rate and at temperatures probably quite near the bottom water temperature. The mounds are principally composed of iron silicates intermixed and encrusted with lesser amounts of manganese oxides. They are generally found in rows, in a uniformly sedimented area above faults or fractures in the crustal rocks which permit fluids to escape from a deep hydrothermal aquifer. The mounds field, covering an area of at least 200 square kilometers and consisting of thousands of individual mounds, is probably less than 300,000 years old; and many of the mounds may be only a few tens of thousands of years old or less. Numerical modeling of two-dimensional convection within the oceanic crust is constrained by these observations of the detailed heat flow field. By comparing the estimated heat and mass flow rates with model flows, the maximum permeability-depth distribution is limited to no greater than .1 mD (10-12 cm2) at depths greater than 2 km. The bulk permeability of the upper 2 km appears constrained to greater than 1.0 mD (10-11 cm2); bulk permeabilities in excess of 5. mD are limited to the uppermost km. One major consequence of the convection is the reduction of the overall temperature of the upper crust relative to conductive models. We found that convection with variable fluid properties responds to thermal forcing in a predictable fashion based on the dimensionless Rayleigh number evaluated at elevated temperature. Correlation of model fluxes and comparison to results for slow spreading rate models indicates a depth of fluid penetration shallower for the Galapagos than other ridges. The heat input at the ridge axis forces the wavelength of the resulting circulation to reflect the depth of strong circulation at the axis, and indicates 60% or more of the "missing" heat flux near mid-ocean ridges is removed by hydrothermal upwelling at the axis of spreading.