Solid mantle upwelling rate beneath the Mid Atlantic Ridge

The upwelling velocity of the solid mantle beneath mid ocean ridges affects processes of melting and generation of the oceanic crust, and constrains models of spreading of oceanic plates. Models of passive flow require that the solid mantle rises beneath a ridge at a speed similar to the half spreading rate of the plates. However, increased buoyancy due to melt depletion and melt retention may cause the sub-ridge mantle to rise at a speed much faster than the half spreading rate. In order to constrain these models it is important to estimate the sub- ridge mantle up-welling rate. Such estimates have been attempted in a few cases by measuring in zero-age basalt disequilibria in short lived isotopes produced by the U-decay series. Melt upward migration from the sub-ridge melting zone to the crust is much faster than its parent solid mantle upward flow; therefore, a time lag is created between melt emplacement as basalt in upper lithosphere and emplacement of the parent residual peridotite in the lower lithosphere. We were able to estimate this time lag along a lithospheric sliver, generated at a 80 km long segment (EMAR) of the Mid Atlantic Ridge located between 10° N and 11°N, just south of the Vema transform. The northern edge of this lithospheric sliver exposes crustal and upper mantle units along a 320 km long spreading flowline equivalent to a 26 Myr long time interval. Comparing temporal variations of crustal thickness, inferred from the Residual Mantle Bouguer Anomaly along a flowline starting from the centre of the EMAR segment, to temporal variations of mantle degree of melting, estimated from mineral chemistry of peridotites exposed along the Vema lithospheric section, allowed us to estimate an average solid mantle rising rate of 16.5 mm/yr below the EMAR segment for a time interval from 26 to 2 Ma. This rate is slightly higher than the average half spreading rate of 15.6 mm/yr for the same period. The similarity between average up-welling rate and spreading rate at 11° N on the Mid Atlantic ridge is in line with up-welling estimates obtained by U-series disequilibria, supporting in general a mostly passive mantle flow model. However, temporal variations of spreading rate at 11° N for the last 26 Myr are decoupled from mantle degree of melting and geothermometry estimates, suggesting variations of mantle upwelling rates due to an active component.