Fracture-induced hydrothermal convection in the oceanic crust and the interpretation of heat-flow data

Small-scale seafloor heat-flow variations with a characteristic wavelength of about 1200 m have been observed on a profile over a sediment-sealed ridge flank. Two theories have been advanced to explain them - low aspect ratio hydrothermal convection in a highly permeable basement layer 600 m thick, or high aspect ratio convection in an anisotropic permeable layer 200 m thick, induced by 20 m topographic variations with a 1000 m half wavelength. Neither theory is totally satisfactory. The first requires a very thick highly permeable zone while the absence of any resolvable basement relief on a complementary seismic survey limits the credibility of the second. We hypothesize that sub-critical hydrothermal convection in normal upper oceanic crust, driven by discrete fractures, can cause the observed heat flow anomalies. We test the hypothesis by using a finite-element code to solve the coupled time-dependent heat transport and fluid flow differential equations. Discrete fractures are incorporated explicitly. We show that the inclusion of fractures in layers 2A and 2B promotes convection. Fluid flow through fractures causes horizontal thermal gradients, and initiates and maintains sub-critical convection within the upper basalts. The predicted heat flow variations are comparable with the observed data.