Melt extraction in the Earth's mantle: Constraints from U–Th–Pa–Ra studies in oceanic basalts

Abstract

U-series studies in oceanic basalts are critical for understanding melting and melt extraction in the Earth's mantle. The combined results of a detailed geochemical study of melting and melt extraction at Theistareykir, northern Iceland, provide a strong case for melt extraction via channeled melt flow at an active spreading ridge. It has often been argued, however, that widely used melting and melt extraction models, which simulate channeled melt extraction (i.e. fractional and/or dynamic melting), can only partially explain the global U-series systematics in oceanic basalts. As a consequence, more complicated models have been invoked, which suggest different styles of melt extraction at different depths/pressures in the mantle, so-called “two-porosity models”. Alternatively, diffusion-controlled mechanisms have been proposed. Here we show that U–Th–Pa–Ra systematics in oceanic basalts can indeed be explained by models where melt transport occurs without chemical equilibrium between melt and solid when variations in all three critical melting parameters (residual porosity, upwelling rate of the solid mantle and melt velocity) are taken into account. Melting at ridges requires systematic variation of at least two critical melting parameters, most likely upwelling and melt extraction rate. Melts generated with increasing lateral distance to the ridge axis are generated with slower upwelling rates and are extracted with lower velocities than melts created closer to the ridge axis. Melting at ocean islands, on the other hand, can successfully be explained by variations in upwelling rate only. Global U-series systematics in OIB originate from superimposed global variations in upwelling velocity due to different buoyancy fluxes and from local variation in upwelling velocity as a function of radial distance to the plume center. The model proposed here is consistent with other geochemical data for oceanic basalts and strongly supports melt extraction via high-porosity channels as a general means of melt extraction from the Earth's upper mantle.