Imaging northern Cascadia wave speed structure and slow slip

Thesis (Ph.D.)--University of Washington, 2018I calculate tremor source amplitudes for the northern Cascadia episodic tremor and slip (ETS) events from 2007-2010 and find they exhibit similar spatiotemporal patterns of radiated energy from tectonic tremor. In the initiation phase of each event, during which tremor starts downdip and moves updip over ~8 days, the tremor area and tremor amplitudes increase quasi-linearly, implying a constant radiated energy rate per unit area and a diffusional process for tremor migration. During this time, tremor amplitudes do not exhibit a strong sensitivity to tidal stress fluctuations. Once the tremor fills the downdip width of the tremoring region, the ETS events begin to propagate to the north and south at a constant rate, with the amplitudes being strongly modulated by tidal stresses. This implies a generally low effective normal stress or low effective friction along the plate interface, and that stress or friction begins higher during the initiation of an ETS event and decreases as the ETS grows to the point where small tidal stress fluctuations can modulate the energy released during slow slip. Using a 2-year deployment of 70 broadband seismometers, and several other seismic data sets, I invert local earthquake travel times to obtain 3-D P- and S-wave velocity models of the Mount St. Helens (MSH) region. Principal features of the 3-D models include: (1) Low P- and S-wave velocities along the St. Helens seismic Zone (SHZ), striking NNW-SSE north of MSH from near the surface to where we lose resolution at 15–20 km depth. This anomaly corresponds to high conductivity as imaged by magnetotelluric studies. The SHZ could represent a zone of crustal weakness with the presence of fluids, fractured rock, and/or sediments from the accretion of the Siletzia terrane; (2) A 4-5% negative P- and S-wave velocity anomaly beneath MSH at depths of 6-15 km with a quasi-cylindrical geometry and a diameter of 5 km, probably indicating a magma storage region. Based on resolution testing of similar-sized features, it is possible that this velocity anomaly is narrower and slower. Assuming approximately 1% partial melt per % velocity variation, this region could contain up to 5-10 km3 of partial melt; (3) A broad, very low P-wave velocity region below 10-km depth extending between Mount Adams and Mount Rainier along and to the east of the main Cascade arc, which is likely due to high-temperature arc crust and the possible presence of melt; (4) Several anomalies associated with surface-mapped features, including high-velocity igneous units such as the Spud Mountain, Spirit Lake, McCoy Creek, Silver Star, and Tatoosh plutons and low velocities in the Chehalis sedimentary basin and the Indian Heaven volcanic field. This dissertation includes two sets of supplementary files: (1) a set of 3-D P- and S-wave velocity models; and (2) a catalog of earthquakes relocated using 3-D velocity models