A Failed Search for Regional Anisotropic Domains in Northern Appalachians
V. Levin,
Dept. of Geology and Geophysics,
Box 208109, Yale University,
New Haven, CT 06520
email: vadim@hess.geology.yale.edu;
W. Menke, LDEO, Palisades, NY
email: menke@ldeo.columbia.edu;
J. Park,
Dept. of Geology and Geophysics,
Box 208109, Yale University,
New Haven, CT 06520
park@hess.geology.yale.edu:
There are two competing explanations for the shear-wave birefringence that is commonly observed in stable continental regions. A fabric "frozen" within the continental lithosphere should lead to "regional anisotropic domains" with close correlation between tectonics and seismic anisotropy. Correlation between shear-wave splitting and shear-wave traveltime delays would be likely in such domains. An alternative hypothesis requires seismic anisotropy to be maintained through active deformation processes, most likely in sub-lithospheric mantle. It would predict broad consistency in shear-wave splitting over large regions and a low level of correlation between the splitting, surface tectonics and the traveltime delays. Observations of shear-wave splitting on a network of stations in the Northeastern US form a remarkably coherent pattern. The region of study - covering large parts of New York and the bulk of New England - spans the width of the Appalachian Orogen, and also includes areas underlain by the Grenvillian basement. The direction of fast shear-wave propagation, as measured from split core-refracted phase (SKS, SKKS, PKS), varies systematically with event backazimuth at all sites examined. The pattern is fit well by a two-layer system with sub-horizontal anisotropic symmetry axes. Relative traveltime delays of shear waves sampling the region vary over distances on the order of 100 km, and correlate with surface geology. The tomographic image constructed using these delays displays lateral variation of shear velocity on the order of 3%. The most robust feature of the tomographic inversion is a slow anomaly beneath the region of exposed Grenvillian basement - the Adirondack Mountains. There is little correlation between this horizontally rough seismic velocity and the horizontally smooth anisotropic model consistent with shear-wave splitting. We therefore conclude that in the northeastern US the concept of "regional anisotropic domains" - i.e. regions within the lithosphere characterized by coherent anisotropic properties reflective of present or past deformation - does not apply.