Using Surface Observations to Constrain the Direction and Magnitude of Mantle Flow Beneath Western North America

While the motions of the surface tectonic plates are well determined, the accompanying horizontal mantle flow is not. Observations of surface deformation (GPS velocities and Quaternary fault slip rates) and upper mantle seismic anisotropy are combined for the first time, to provide a direct estimate of this flow field. We apply our investigation to western North America where seismic tomography shows a relatively thin lithosphere. Here the likely source of shear wave anisotropy results from a deformation fabric associated with the differential horizontal motion between the base of the lithosphere and the underlying mantle. For a vertically propagating shear wave recorded at a single station, and for mantle strains of order unity, the fast polarization direction, φ , of a split shear wave will be parallel to the direction of progressive simple shear, defined by this differential motion between lithosphere and underlying mantle. If the motion of the overlying lithospehre is known both within and across a plate boundary zone, such as western North America, then the direction and magnitude of mantle flow beneath the plate boundary zone can be uniquely determined with three or more observations of fast polarization directions. Within the Pacific-North American Plate boundary zone in western North America we find that the mantle velocity is 5.0+/-1.5 cm/yr and directed E-NE in a hotspot frame, nearly opposite to the direction of North American plate motion (WSW). The flow is only weakly coupled to the motion of the surface plates, producing a weak drag force. This flow field is most likely due to mantle density heterogeneity associated with the sinking of the old Farallon slab beneath North America. The last few decades have seen the development of two basically incompatible views of the plate-mantle system. The tectonophysical view assumes effective decoupling between the plate and a stationary mantle by a well developed asthenosphere. The plates are essentially 'self-driving'. In the mantle-dynamics view, the plates are strongly-coupled to a mantle flow field that is driven by sources of buoyancy in the mantle. Our results suggest that both views are partly correct. We do observe plate-mantle decoupling, but also find a mantle flow field that is likely driven by deep mantle density heterogeneity. If this description is correct, then beneath the oceanic two-thirds of the Earth, where a well developed asthenosphere is most likely present, we may be completely surprised by the motions of the mantle. More importantly, however, the direct measurement of these motions, as done here, hold the possibility of dramatically increasing our understanding of the dynamic mantle.