| Summary: | In this thesis, I investigate the seismic structure of oceanic lithosphere and asthenosphere with a particular focus on seismic anisotropy, using high-resolution surface waves recorded on ocean-bottom seismometers (OBS) in the Pacific and Atlantic Oceans. The NoMelt (~70 Ma) and Young OBS Research into Convecting Asthenosphere (ORCA) (~43 Ma) OBS experiments located in the central and south Pacific, respectively, provide a detailed picture of ``typical'' oceanic lithosphere and asthenosphere and offer an unprecedented opportunity to investigate the age dependence of oceanic upper mantle structure. The Eastern North American Margin Community Seismic Experiment (ENAM-CSE) OBS array located just offshore the Eastern U.S. captures the transition from continental rifting during Pangea to normal seafloor spreading, representing significantly slower spreading rates. Collectively, this work represents a diverse set of observations that improve our understanding of seafloor spreading, present-day mantle dynamics, and ocean basin evolution. At NoMelt, which represents pristine relatively unaltered oceanic mantle, we observe strong azimuthal anisotropy in the lithosphere that correlates with corner-flow induced shear during seafloor spreading. We observe perhaps the first clear Love-wave azimuthal anisotropy that, in addition to co-located Rayleigh-wave and active source Pn constraints, provides a novel in-situ estimate of the complete elastic tensor of the oceanic lithosphere. Comparing this observed anisotropy to a database of laboratory and naturally deformed olivine samples from the literature leads us to infer an alternative ``D-type'' fabric associated with grain-size sensitive deformation, rather than the commonly assumed A-type fabric. This has vast implications for our understanding of grain-scale deformation active at mid-ocean ridges and subsequent thermo-rheological evolution of the lithosphere. At both NoMelt and YoungORCA we observe radial anisotropy in the lithosphere with Vsh > Vsv indicating subhorizontal fabric, in contrast to some recent global models. We also observe azimuthal anisotropy in the lithosphere that parallels the fossil-spreading direction. Estimates of radial anisotropy in the crust at both locations are the first of their kind and suggest horizontal layering and/or shearing associated with the crustal accretion process. Both experiments show asthenospheric anisotropy that is significantly rotated from current-day absolute plate motion as well as rotated from one another, at odds with the typical expectation of plate-induced shearing. This observation is consistent with small-scale density- or pressure-driven convection beneath the Pacific basin that varies in orientation over a length scale of at most ~2000 km and likely shorter. By directly comparing shear velocities at YoungORCA and NoMelt, we show that the half-space cooling model can account for most (~75%) of the sublithospheric velocity difference between the two location when anelastic effects are accounted for. The unaccounted for ~25% velocity reduction at YoungORCA is consistent with lithospheric reheating, perhaps related to upwelling of hot mantle from small-scale convection or its proximity to the Marquesas hotspot. While lithospheric anisotropy is parallel to the fossil-seafloor-spreading direction at both fast-spreading Pacific locations, it is perpendicular to spreading at the ENAM-CSE in the northwest Atlantic where spreading was ultra-slow to slow. Instead, anisotropy correlates with paleo absolute plate motion at the time of Pangea rifting ~180–195 Ma. We propose that ultra-slow-spreading environments, such as the early Atlantic, primarily record plate-motion modified fabric in the lithosphere rather than typical seafloor spreading fabric. Furthermore, slow shear velocities in the lithosphere may indicate that normal seafloor spreading did not initiate until ~170 Ma, 10–25 Myr after the initiation of continental rifting, revising previous estimates. Alternatively, it may shed new light on melt extraction at ultra-slow spreading environments.
|
|---|