GEODETIC EXPRESSION OF UPPER MANTLE FLOWS IN SOUTHERN CALIFORNIA

An important outstanding question is whether geodetic measurements in continental plate boundary zones can sense active small-scale convection within the lithospheric mantle. Negatively buoyant rocks deriving from eclogite facies metamorphism or melt residues may accumulate at the base of the crust in collisional orogens. Foundering of these dense materials through the mantle entrains mantle lithosphere, and may lead to significant loss of lithosphere to the underlying less dense asthenosphere. Convective flows associated with these processes generate tractions on the base of the overlying crust. Geologic observations and numerical explorations suggest that crustal deformation associated with these tractions may be appreciable over relatively long periods of time. New geodetic velocity and seismic tomography data for southern California reveal a strong correlation between dilatational strain rates and upper mantle wave speeds, suggesting that small-scale upper mantle convection may be actively contributing to upper crustal strain accumulation there. We developed an analytical model in which a viscoelastic crust is driven by tractions induced by negatively buoyant cylinders sinking in a viscous mantle halfspace. For models in which the crust is composed of a single viscoelastic layer with viscosity that is higher than that of the upper mantle, we found that tractions on the base of the crust associated with Stokes-like flow may result in horizontal motions of the crustal surface at strain rates of order 20 nanostrain/yr, depending on crustal thickness, cylinder size and density contrast, and crust and mantle viscosities. The model results are similar in magnitude and style of deformation to the geodetic observations for southern California. For models in which the lower crust behaves as a weak, low viscosity layer sandwiched between stronger upper crust and mantle, the horizontal components of upper crust are decoupled from mantle flow. Thus, the geodetic strain rate and seismic tomography data for southern California may suggest that the lower crust of southern California is relatively strong. This inference is consistent with long-time scale rheologic properties recently inferred by other researchers from short-time scale postseismic deformation data for the southern California region.