The proposed program of investigation of deep earth structure aims at providing seismological constraints on global geodynamical models of the earth's mantle as well as the core, to help further our understanding of the mechanisms, deep inside the earth, that drive plate tectonics, causing earthquakes and volcanic eruptions. The resulting information on the location, shape and strength of major structures such as upwellings and downwellings will benefit a broader community of non seismologists in geosciences (geodynamicists, mineral physicists and geochemists. This program will directly contribute to the education of 6 graduate students and 2 post-docs. We estimate that it will result in at least 8 publications in refereed journals.
This program is aimed at investigating the deep structure of the earth using seismic waveforms collected at global and regional arrays of digital broadband stations using forward and inverse methodology. A central goal is to implement a novel approach based on the combination of accurate seismic wavefield computations in complex three dimensional (3D) structures, as afforded by the efficient C-SEM ("Coupled spectral element method"), for the forward problem, and approximate estimation of Frechet derivatives for the inverse problem. This hybrid approach will allow us to keep the computational costs down while significantly improving resolution through the use of the complete wavefield. We plan, in particular, to: + Develop models of 3D elastic, anelastic and anisotropic structure of the earth's upper mantle using C-SEM + Address the issue of parametrization of large scale tomographic models of the mantle to better constrain the boundaries of major heterogeneous provinces both in the upper and in the lower mantle. + Combine long period waveform and body wave data to map the anisotropic structure of the lithosphere/asthenosphere system at the global scale. + Continue our program of inner core sensitive data collection and interpretation + Investigate further the mechanisms of generation of the earth's low frequency "hum"
The main goal of this project was to improve the sharpness of images of the earth's upper mantle at the global scale, in order to inform our understanding of the dynamics that control tectonic plate motions. One of the fundamental questions that have remained unanswered since the plate tectonics revolution in the late 1960's is the discrepancy, in the oceans, between the expected relationship of seafloor depth and heat flow, with age of the plate (or distance from the mid-ocean ridges where new crust is formed) with the observations. The latter imply the presence of other sources of heat under the ocean basins, in addition to that provided by upwelling under ridges. Two models have been proposed: 1) the "plume" model, where heat is supplied through narrow conduits originating deep in the mantle and emerging in mid-plate "hotspot" volcanoes; and 2) secondary scale convection (ie. in addition to the main circulation defined by upwelling at ridges and downwelling at subduction zones) in the form of what has become known as "Richter Rolls", convective rolls spanning the upper mantle whose axes align with the direction of plate motion. We have adapted new tools for the numerical computation of the seismic wavefield in complex 3D models of the earth and developed an imaging technique based on the analysis of seismic waveforms observed globally. In particular, this approach allows an accurate capture of the effect of low velocity regions on the wavefield, and therefore better resolution of these features in the resulting images than was possible previously with approximate wavefield computations. We have applied this technique in an iterative process, starting from a spherically symmetric (1D) model of the Earth, to develop two generations of global three dimensional shear velocity models of the upper mantle. We are the first to have achieved this at this global scale. While the large scale features of these models confirm 20 years of global seismic tomography performed using ray theoretical approximations, our images provide a more focused view of structures in the upper mantle, in particular in regions of lower than average velocity extending deep under ocean basins (Figure 1).
