Geophysical and geological perspectives on accretionary wedge evolution, splay faulting, and megathrust earthquake processes at the Cascadia Subduction Zone

Abstract

Accretionary wedges, submarine fold-and-thrust belts that accumulate at subduction zones, are the main locus of brittle deformation during megathrust earthquakes, and this deformation is in turn the main driver of tsunamis. Though accretionary wedges were initially thought aseismic, recent megathrust earthquakes have revealed that coseismic slip on the shallow megathrust and accompanying high-angle splay faults is common and can be devastating. In this Dissertation, I apply geophysical and geological methods to better constrain accretionary wedge structure and deformation as it applies to megathrust earthquake hazards at the Cascadia Subduction Zone. In Chapter 2, I use nested scales of seismic reflection imaging to better constrain splay fault activity in the landward vergence zone, which reveals that active splay faults concentrate near the deformation front. These splay faults are likely to slip coseismically along with the shallow megathrust during earthquakes. In Chapter 3, I then reexamine the position of the megathrust fault in the landward vergence zone by applying structural geological techniques to newly-collected crustal-scale seismic reflection imaging. I find that the megathrust fault is never co-located with the sediment-basement interface, and instead facilitates on average ~500 m of sediment underthrusting in the active outer wedge. These studies have distinct implications for better understanding the characteristics of future megathrust earthquakes at Cascadia, including informing fault slip scenarios and fault physical properties. Next, I turn to the portion of Cascadia's accretionary wedge exposed onshore, known as the Olympic Subduction Complex. This area has been identified as an accretionary complex for decades, but there has been little modern geological work in the region. In Chapter 4, I focus on a site known as Lakes of the Gods, where I identify a paleomegathrust interface comprised of multiple anastomosing fault strands contained in a ~450 m thick section of deformed mélange. In Chapter 5, I design an experiment to test how strain partitions across the various structural domains of the paleomegathrust interface. I find that interseismic, mélange-forming processes account for negligible amounts of plate boundary strain, while localized, coseismic, fault zone processes must account for the remainder. This is consistent with the geophysical observations of modern subduction zones, but geologically counterintuitive because mélange is vastly volumetrically dominant in the rock record compared to fault rocks.