Author: Catherine Ross

Publisher:

Published: 2018

Total Pages:

ISBN-13:

"Exhumed fault rocks contain records of past earthquakes and provide insights into deformation processes associated with seismic slip. Static stress changes caused by fault displacement may be of significant magnitude around fault bends, ends, intersections, and have been shown to partially explain aftershock distributions (Poliakov et al., 2002; Savage et al., 2017). Post-seismic relaxation around the faults may change the recurrence interval of large events along certain fault strands in a fault system (Freed, 2005; Felzer and Brodsky, 2005; Richards-Dinger et al., 2010). In the brittle-ductile transition zone, these stress concentrations may be relaxed after earthquakes by ductile flow. I used an outcrop of pseudotachylyte faults and nearby deformed wallrock as a small-scale model of a seismic fault system to test whether wallrock deformation is the result of staticstress changes associated with earthquake displacements. To do this, I used a newly developed technique of measuring strain across off-fault deformation features and comparing the strain to static stress change maps. I used Coulomb3, a fault modeling software, to model the static stress changes (with model inputs constrained by field observations) and compared the orientation and relative magnitude of compressive and tensile predicted stress changes with the shortening and elongation represented by wallrock deformation features.In the Fort Foster Brittle Zone (Kittery, Maine; Swanson, 2006), I mapped a 5.6 m-long area with two interconnected, sub-parallel pseudotachylyte fault veins cutting an ultramylonite zone, and associated wallrock deformation features including pseudotachylyte injections, pseudotachylyte-filled voids, mm-[mu]m subsidiary faults, and folds. High-resolution photos, orientation measurements, and coseismic offsets were used to generate a simplified fault model in Coulomb3. I measured strain across the off-fault deformation features in the fault-parallel and perpendicular directions. Using microstructural analysis, I also determined the deformation mechanisms involved in the formation of each type of wallrock feature. As some deformation mechanisms are rate-limited, this information can also be used to infer whether features could have formed co-seismically, post-seismically, or require longer timescales of creep. I used spatial statistics (Local Indicators of Spatial Association or LISA) to test whether the areas of significant stress change predicted by the Coulomb3 model correlate with areas of substantial wallrock deformation on the strain map interpreted from field observations.The correlation of orientation and magnitude between static stress change and strain is strongest at the major bends in the bounding faults. The correlation confirms my hypothesis that stress changes caused by co-seismic displacements were at least partially relieved by off-fault deformation. The correlation between static stress change and strain is strongest for pseudotachylyte melt-related features and strike-slip faults, but weakest for the folds. The strong association implies that the melt-related features and strike-slip faults are the most likely features to have formed in the co- to post-seismic interval to facilitate post-seismic relaxation.Additionally, a microstructural analysis reveals that the deformation features (injection veins, dilational zones, and strike-slip faults) either involve pseudotachylyte melt production and quenching, brittle fracturing, or cataclasis with the exception of the folds. Cataclasis and frictional sliding are not rate-limited, but because strike-slip faults are strongly correlated with areas of static stress increase, and are therefore interpreted to form in response to co-seismic displacements, they can be constrained to the co- to post-seismic period. Features that are melt-filled represent coseismic deformation because of the short quench times of pseudotachylyte. The folds in the wallrock deformed by a combination..." --