Oceanic asthenosphere subduction and its geological implications

Abstract:

We investigate the evolution of oceanic asthenosphere during subduction by exploring various scenarios including plate kinematics and plausible values of asthenosphere viscosity and density. We find that the oceanic asthenosphere will always subduct with the down-going slab as long as its average viscosity value is no smaller than 1×10¹⁸ Pa s. In order to allow slabs to subduct into the deep upper mantle, a maximum oceanic asthenosphere density reduction relative to the underlying mantle should be no larger than 0.7%, assuming a 200-km-thick asthenosphere channel. We find that a significant portion of the asthenosphere buoyancy should result from its excess temperature from the long-term thermal evolution of mantle convection. Our results are in contrast to an earlier suggestion that negligible amount (<30 km thick) of asthenosphere could get subducted, which is likely due to over-simplicity of subduction geometry and model boundary conditions.

The recycling of a weak and hot asthenosphere provides a novel mechanism for the formation of slow seismic anomalies within the deep mantle. This, in turn, questions the commonly believed deep mantle plume origin of intra-plate volcanism, with a typical example being the Yellowstone volcanic system. Our current results suggest that a buoyant asthenosphere can be dragged down into the lower mantle and then moves upward due to its buoyancy when the overlying slab barrier is removed. To further test our hypothesis, we construct a 4D subduction model for western North America during the Cenozoic. We use data assimilation techniques to incorporate plate kinematics and sea floor ages as boundary conditions, and seismic anomalies converted density structure as internal buoyancy source. The subduction history is calibrated through a hybrid of forward and adjoint simulations satisfying multiple observational constraints. Some preliminary results will be presented.