On the scale of mantle convection

Abstract

Observational evidence from glacial rebound and gravity anomalies shows that the mantle has a relatively uniform viscosity. Such a structure is consistent with the effects of temperature and pressure on microscopic mechanisms for plastic flow of solids. Thus the asthenosphere may extend to the base of the mantle. Mantle phase changes do not inhibit mantle-wide convection, and any compositional differences between the upper and lower mantle are not resolved at present. Thus there is no reason not to expect convection extending into the lower mantle. The stress state of descending slabs is consistent with mantle-wide convection, and although the slabs may encounter an increased resistance to sinking at ∼700 km depth there is no evidence that they are stopped by the increased resistance. Seismic heterogeneity in the lower mantle suggests large-scale convection, which is perhaps driven by heat from the earth's core. Plate motions may interact with large-scale flow in the mantle, and are probably not completely decoupled from flow in the lower mantle. Lateral variations in flow properties in the upper mantle may be more important than vertical variations, and may strongly influence plate motions.