HYDROTHERMAL HEAT AND MASS FLUXES THROUGH MID-OCEAN RIDGE AXIS AND FLANKS

Although both chemical and physical constraints have been applied to estimate hydrothermal fluxes through oceanic crust, the physical methods are at present more robust and less uncertain. The strongest constraint on the total hydrothermal power output from mid-ocean ridge axis and flanks is the discrepancy between measured heat flow and that predicted theoretically for a cooling lithospheric plate. The current best estimate is 10 ± 2 TW, a value that has not changed for 25 yrs. This represents nearly 1/4 of the 44 TW lost from the solid Earth. The strongest constraint on hydrothermal power output at the axis is the rate of supply of magmatic heat to form the crust, coupled with observations that support rapid crystallization and possibly cooling of the crust to full depth within a few km of the axis. A lower limit of 1.5 ± 0.2 TW would crystallize the entire crust and cool the sheeted dikes to 350°C, whereas an upper limit of 2.8 ± 0.3 TW would cool as well the layer-3 gabbros to 350°C. (Power output would exceed 2.8 TW in the latter case because the mantle would contribute substantial heat by conduction, as noted below.) Applying the lower limit to medium- and fast-spreading ridges, which typically have a shallow crustal magma lens, and the upper limit to slow-spreading ridges, which have deep crustal earthquakes, results in a global best estimate of 1.8 ± 0.3 TW, which would support a seawater flux of 3.7 ± 0.5 x 10e16 g/yr at 350˚C. These calculations assume that all magmatic heat is removed by hydrothermal advection and none by conduction, i.e., the Nusselt number for axial convection is >10, which seems likely. They ignore input of heat to the crust by conduction from the mantle, a factor that becomes important if hydrothermal cooling extends to full crustal depth near the axis. Maclennan et al. (2005, Geology) found that such deep cooling was required to match subsidence even for the fast-spreading northern E. Pacific Rise, requiring a hydrothermal power output of 2.9-4.2 TW within <30 km of the axis from crust <0.5 Ma-old. Such deep cooling implies that the near-axial region, defined as crust 0.1-1 Ma-old, would experience mainly reheating by conduction across the Moho. Hydrothermal power output on ridge flanks, through crust 1-65 Ma in age, can be calculated by difference as (10 ± 2) – (3 ± 1) = 7 ± 2 TW, which would support a seawater flux of 5 x 10e18 g/yr at 10˚C, nearly 1/7 of the river input of 3.8 x 10e19 g H2O/yr.