Upper-Mantle Anisotropy beneath California and the Surrounding Region
Prasoon Gupta and James B. Gaherty
School of Earth and Atmospheric Sciences, Georgia Institute of Technology, Atlanta, GA 30332-0340; email email@example.com
Observations of seismic anisotropy provide critical constraints on the nature of the structural fabric associated with flow and deformation in the Earth’s mantle. Important mechanisms of deformation illuminated by anisotropy might include large-scale flow associated with plate motion, and/or small-scale deformation associated with regional tectonism. Single-station analyses of vertical shear-wave splitting and global studies of long-period surface-wave dispersion have successfully mapped anisotropy at both small (~100 km) and large (>2000 km) length scales, respectively. At a regional scale of several hundred km, the nature of anisotropy is very poorly known, however. Lack of knowledge of anisotropy at these intermediate scale lengths severely limits our ability to definitively associate seismic anisotropy with the dominant patterns of mantle flow.
In this study we attempt to estimate regional-scale anisotropic heterogeneity using phase and polarization observations of fundamental mode surface waves across "array" geometries of closely spaced stations in California and neighboring regions. Seismic sources are 120 earthquakes that occurred between 1990-1999 in the western US, central America, the northeast Pacific, and the western Pacific, recorded at up to 36 broad band seismic stations from a variety of networks, including the GSN, USNSN, BDSN, TriNet, and Anza. We have measured frequency-dependent phase delays (travel-time residuals) of the observed fundamental- and multi-mode Love and Rayleigh waves, across the 10-40 mHz band, relative to synthetics constructed using IASP91 and the Crust5.1 crustal model. To date, we have extracted over 8000 delay times from over 2000 seismograms. For regional events within the western US, the phase delays can be used directly to constrain anisotropic heterogeneity along the entire source-receiver path. For the teleseismic observations, we localize the delay times to the region of interest using two-station or other intra-array techniques. These phase delays will be inverted for a 3D regional tomographic model of isotropic and anisotropic heterogeneity using a Bayesian approach that allows us to explicitly test a variety of a priori assumptions on the nature of the anisotropy. Ultimately, we aim to map the upper-mantle strain associated with the San Andreas fault and Basin and Range tectonism. In the process, we hope to (1) constrain and distinguish intra-array isotropic and anisotropic heterogeneity; (2) evaluate the dominant length scales of anisotropic heterogeneity in the mantle beneath western North America; and (3) test the compatibility between inferences of anisotropy obtained from surface wave, Pn, and shear-wave splitting analysis.