Splitting intensity measurements of North America and finite-frequency modeling of upper mantle anisotropy

The central theme of this dissertation is to investigate the interaction between the overlying lithospheric plates and the hotter and more deformable asthenosphere to examine how they are coupled. The answer will have very significant implications because the coupled lithosphere-asthenosphere setting predicts shear beneath the lithosphere which drives large-scale mantle convection, a widely-accepted schematic for modern mantle convection research. In our research, the primary tool to examine this coupled interaction is anisotropy in the upper mantle observed via shear wave splitting. The first research component involves measuring splitting intensity (SI) of the core-refracted shear waves (SKS) observed at 1,436 USArray Transportable Array (USArray-TA) seismic stations which cover most of the contiguous U.S. By fitting a sinusoidal to the back-azimuthal dependence of splitting intensity, traditional splitting parameters, the polarization angle between the radial direction and the fast axis, &phis;, and the delay time between the fast and slow polarizations, deltat, are obtained and used in comparison with absolute plate motion (APM), geological basement provinces, magnetic, gravity anomalies and lithospheric thickness to reveal the interactions between asthenospheric flow and lithospheric anisotropic structures of several geological regions of North America. Preliminary results shows that the Rocky Mountain front has a complicated flow due to transition with thickness along APM flow, the Gulf Coast has a strong APM asthenospheric signature in region of thin lithosphere and the northern Central U.S. has complicated interactions between asthenosphere and lithosphere. We observe a notable contrast between the Superior Province vs. Trans-Hudson where lithospheric texture alignment plays an important role in adding vs. subtracting the splitting signals.;The next component of our research focuses on the Idaho-Oregon (IDOR) region. This region is an assemblage of several micro terranes with both continental and oceanic origins throughout its accretion history making it a very complex geological setting including the presence of the north-striking western Idaho shear zone (WISZ) in the middle. We deployed 85 temporary seismic stations with station-spacing of ∼30 km during 2011--2013 and passively recorded seismic data for an average duration of 1.5 years. The SKS phase of the seismogram is used to obtain splitting intensity, which we use to model realistic 3-D upper-mantle anisotropy. There are two parts in this study, first SKS splitting intensity measurements were made from seismograms recorded at 83 IDOR seismic stations and 45 USArray-TA stations, which consist of analyzing more than 75,000 individual traces. As a result, we obtain high-resolution and spatially coherent shear-wave splitting dataset of the IDOR region. Second, we use back-azimuthal variations of splitting intensity at all stations to model for 3-D anisotropy using the finite-frequency approach. Preliminary models show depth-dependent behaviors of both fast polarization direction and strength of anisotropy down to ∼150 km where the model starts to show poor resolution due to the size of the SKS fresnel zone.;Last, we show preliminary inverted models for 3-D upper-mantle anisotropy of North America as well as our progress of spherical coordinate inversion of the USArray-TA splitting measurements. This will set up a starting point for performing a joint-inversion with surface wave dataset that will be measured at exact seismic stations. This last task will be exercised by the help of 3-D finite-frequency Frechet sensitivity kernels for surface waveforms based on the Born approximation with a model parametrized for hexagonal symmetry. Their formulation will provide a complementary approach to invert surface wave data in combination with our SI data for upper mantle anisotropy model of North America with highest resolution for the first time.