Seismological Constraints on Global Mantle Anisotropy

University Of California-Los Angeles, Los Angeles CA

Investigators

Abstract

Technical Description: The goal of the proposed research is to determine the three-dimensional variations of azimuthal anisotropy in the upper 1000 km of the mantle, at the global scale. Seismic anisotropy, that is the dependence of seismic wave velocity on the direction of propagation or polarization, offers a more complete description of Earth's elastic structure than isotropic velocities alone, and may be a signal of mantle deformation. Therefore, it constitutes a unique way of understanding and constraining Earth?s interior. However, aside from the base of the mantle and the top of the upper mantle, little is known about mantle seismic anisotropy, mostly because of the reduced resolution of commonly used seismic data below ~250 km. The proposed work will take advantage of a new global surface wave dataset to model the three-dimensional changes in azimuthal anisotropy in the upper 1000 km of the mantle. These data are azimuthally anisotropic fundamental mode and overtone surface wave phase velocity maps for Love and Rayleigh waves, and because higher mode measurements are included these data have high sensitivity to the targeted deep structure. In order to constrain anisotropy, we will combine a traditional least-squares inverse technique with a model space search approach. This forward modeling method will enable us to determine the common properties of all the models that satisfy the data. It will therefore allow us to ascertain which model features are robust and which parameters trade-off with others, a key-element in making meaningful interpretation of the results. This project will enable us to determine the likelihood of presence of azimuthal anisotropy in the transition zone and the top of the lower mantle, how azimuthal anisotropy changes with depth, and how it differs beneath oceans and continents. To the extent we can relate seismic anisotropy to deformation, these results will be used to shed new light on (1) the geometry of mantle deformation, (2) whether and where sub-lithospheric deformation couples with plate motion, (3) the contribution to shear-wave splitting of present-day asthenospheric deformation versus fossil lithospheric deformation, and (4) the passive vs. active nature of mid-ocean ridges. Non-Technical Description: Our current knowledge of Earth?s deep mantle is poor, but essential to comprehend surface plate tectonics and how they relate to deformation at greater depths. The goal of this project is to improve our understanding of mantle deformation at large depths by mapping the three-dimensional directional dependence of seismic wave velocities, i.e. seismic anisotropy, down to depths of 1000 km. Seismic anisotropy is probably a signal of mantle deformation and therefore constitutes a unique tool to constraint its mineralogy, composition, and dynamics. Combined with mineral physics data and geodynamic modeling, it can help us understand the evolution of our planet. To reach our goal we will combine a new global seismic dataset with an innovative forward modeling method, which will help us explore many different types of models and select those that explain the data the best. With this computationally intensive approach we will be able to estimate model uncertainties, a key-element to make sensible interpretation of the results in terms of composition, mineralogy, and deformation. The models obtained will help constraining geodynamic models, and will guide mineral physics experiments trying to understand mantle deformation. Funding of this project will thus benefit a large part of the Earth science community. The results, methods and models will be shared with the broader audience and the seismological research community through scientific publications, conference presentations, and a website. In addition, the proposed research will provide support for a graduate student who will be trained and acquire knowledge in the fields of global seismic tomography, modeling techniques, and parallel computer programming.