Magma Supply at the Arctic Gakkel Ridge: Constraints from Peridotites and Basalts

Abstract:

Crustal thickness in global ridge systems is widely believed to be nearly uniform (~7 km) at slow- and fast-spreading mid-ocean ridges, but appears significantly thinner (< ~4 km) at ultraslow-spreading ridges. At the slowest-spreading Arctic Gakkel Ridge, the crust becomes extremely thin (1.4 – 2.9 km; [1]). The thin crust at the Gakkel and other ultraslow-spreading ridges, has been attributed to lithosphere thickening, ancient mantle depletion, lower mantle temperature, ridge obliquity, and melt retention/focusing. To better understand the magma supply at ultraslow-spreading ridges, we examined melting dynamics by linking peridotites and basalts dredged along the Gakkel Ridge. We analyzed rare earth elements in clinopyroxene from 84 residual peridotites, and estimated melting parameters for individual samples through nonlinear least squares analyses. The degrees of melting show a large variation but mainly center at around 7% assuming a somewhat arbitrary but widely used depleted MORB mantle starting composition. Thermobarometry on published primitive basaltic glasses from [2] indicates that the mantle potential temperature at the Gakkel Ridge is ~50°C cooler than that at the East Pacific Rise. The ridge-scale low-degree melting and lower mantle potential temperature place the final depth of melting at ~30 km and a melt thickness of 1.0 or 2.9 km for a triangular or trapezoidal melting regime, respectively. The final melting depth is consistent with excess conductive cooling and lithosphere thickening suggested by geodynamic models, while the estimated melt thickness is comparable to the seismic crust (1.4 – 2.9 km; [1]). The general agreement among geochemical analyses, seismic measurements, and geodynamic models supports that lower mantle potential temperature and thick lithosphere determine the ridge-scale low-degree melting and thin crust at the Gakkel Ridge, while melt retention/focusing and excess ancient mantle depletion are perhaps locally important at short length scales (e.g., < 50 – 100 km).

[1] Jokat and Schmidt-Aursch (2007) Geophys. J. Int. (2007) 168, 983-998. [2] Gale et al. (2012) J. Petrology, 55, 1051-1082.