Accretion and Subduction of Oceanic Lithosphere: 2D and 3D Seismic Studies of Off-Axis Magma Lenses at East Pacific Rise 9°37-40'N Area and Downgoing Juan de Fuca Plate at Cascadia Subduction Zone

Two thirds of the Earth's lithosphere is covered by the ocean. The oceanic lithosphere is formed at mid-ocean ridges, evolves and interacts with the overlying ocean for millions of years, and is eventually consumed at subduction zones. In this thesis, I use 2D and 3D multichannel seismic (MCS) data to investigate the accretionary and hydrothermal process on the ridge flank of the fast-spreading East Pacific Rise (EPR) at 9°37-40'N and the structure of the downgoing Juan de Fuca plate at the Cascadia subduction zone offshore Oregon and Washington.;Using 3D multichannel seismic (MCS) data, I image a series of off-axis magma lenses (OAML) in the middle or lower crust, 2-10 km from the ridge axis at EPR 9°37-40'N. The large OAMLs are associated with Moho travel time anomalies and local volcanic edifices above them, indicating off-axis magmatism contributes to crustal accretion though both intrusion and eruption (Chapter 1). To assess the effect of OAMLs on the upper crustal structure, I conduct 2-D travel time tomography on downward continued MCS data along two across-axis lines above a prominent OAML in our study area. I find higher upper crustal velocity in a region ~ 2 km wide above this OAML compared with the surrounding crust. I attribute these local anomalies to enhanced precipitation of alteration minerals in the pore space of upper crust associated with high-temperature off-axis hydrothermal circulation driven by the OAML (Chapter 2).;At Cascadia, a young and hot end-member of the global subduction system, the state of hydration of the downgoing Juan de Fuca (JdF) plate is important to a number of subduction processes, yet is poorly known. As local zones of higher porosity and permeability, faults constitute primary conduits for seawater to enter the crust and potentially uppermost mantle. From pre-stack time migrated MCS images, I observe pervasive faulting in the sediment section up to 200 km from the deformation front. Yet faults with large throw and bright fault plane reflections that are developed under subduction bending are confined to a region 50-60 km wide offshore Oregon and less than ~45 km wide offshore Washington. Near the deformation front of Oregon margin, bending-related faults cut through the crust and extend to ~6-7 km in the mantle, whereas at Washington margin, faults are confined to upper and middle crust, indicating that Oregon margin has experienced more extensive bend faulting and related alteration. These observations argue against pervasive serpentinization in the slab mantle beneath Washington and suggest mechanisms other than dehydration embrittlement need to be considered to explain the intermediate depth earthquakes found along the Washington margin (Chapter 3). Using MCS images of a ~400 km along-strike profile ~10-15 km from the deformation front, I investigate the along-trench variation of the structure of downgoing JdF plate and its relation to the regional segmentation of Cascadia subduction zone. I observe that the propagator wakes within the oceanic plate are associated with anomalous basement topography and crustal reflectivity. Further landward, segment boundaries of ETS recurrence interval and relative timing align with the propagator traces within the subducting plate. I propose while the upper plate structure or composition may determine the threshold of fluid pore pressure at which ETS occur, the propagators may define barriers for ETS events that occur at the same time. I also observe a change in crustal structure near 45.8°N that is consistent with an increase in bend-faulting and hydration south of 45.8°N;. In addition, four previously mapped oblique strike-slip faults are associated with changes in Moho reflection, indicating that they transect the entire crust and may cause localized mantle hydration (Chapter 4).