Constraints on the Lost City Hydrothermal System from borehole thermal data; 3-D models of heat flow and hydrothermal circulation in an oceanic core complex.

Abstract:

A perennial problem in near-ridge hydrothermal circulation is that the only directly measurable data to test models is often vent fluid temperature. Surface heat flow measurements may be available but without the underlying thermal structure it is not known if they are transient and affected by local hydrothermal flow, or conductive.

The Atlantis Massif oceanic core complex at 30 °N on the mid-Atlantic Ridge, offers a unique opportunity to better constrain hydrothermal circulation models. The temperature profile in gabbroic rocks of IODP Hole 1309D was measured in IODPExpedition 340T, and found to be near-conductive, but with a slight inflexion at ~750 mbsf indicating downward advection of fluid above that level. The lack of deep convection is especially remarkable given that the long-lived Lost City Hydrothermal Field (LCHF) is located only 5km to the south.

We have modelled hydrothermal circulation in the Massif using Comsol Multiphysics, comparing 2-D and 3-D topographic models and using temperature-dependent conductivity to give the best estimate of heatflow into the Massif. We can constrain maximum permeability in gabbro below 750 mbsf to 5e-17 m². The thermal gradient in the upper part of the borehole can be matched with a permeability of 3e-14 m² in a 750 m thick layer parallel to the surface of the massif, with upflow occurring in areas of high topography and downflow at the location of the borehole. However in 3-D the precise flow pattern is quite model dependent, and the thermal structure can be matched either by downflow centred on the borehole at lower permeability or centred a few hundred metres from the borehole at higher permeability.

The borehole gradient is compatible with the longevity (>120 kyr) and outflow temperature (40-90 °C) of the LCHF either with a deep more permeable (1e-14 m² to 1e-15 m²) domain beneath the vent site in 2-D or a permeable fault slot 500 to 1000m wide and parallel to the transform fault in 3-D. In both cases topography exerts a strong control on vent location, and steep boundaries between permeable and impermeable domains stabilise long term steady venting.

Work is in progress to model the thermal evolution of the massif during fault exhumation over the last 1.2 million years, comparing this with the present day heatflow estimated from measurements.