Low-Frequency Earthquakes and the Subduction Zone Plate Boundary

Abstract:

Low frequency earthquakes (LFEs) in subduction zones are commonly thought to represent slip on the plate interface. However, owing to large depth errors associated with absolute LFE hypocenters and the lack of a precisely defined structural marker, the plate stratigraphic context for LFEs and tremor remains poorly constrained. For example, LFEs may align on a single, thin and geographically expansive plate interface at the top of the subducting place, but it is also conceivable that they occur along fault strands within a distributed deformation zone with low shear strength. We combine high precision relative locations of LFEs from southern Vancouver Island and northern Washington with observations of source-side P-to-S conversions within LFE template waveforms to constrain LFE locations relative to plate boundary stratigraphy.

Relative tremor and LFE locations have been previously determined using either "cross-detection" methods, which compare similar waveforms of different events at the same station, or "cross-station" methods, which exploit similar waveforms of the same event at different stations. We combine both approaches by including all station-detection pairs in one large inversion for optimum delay times using the VanDecar and Crosson multi-channel cross-correlation method. These times are employed within HypoDD to generate relative locations for individual LFE detections. We plan to investigate the consistency of these high precision relative locations with timing of scattered phases within stacked LFE waveforms. These scattered phases represent conversions from the top and base of a ubiquitous low-velocity zone widely inferred to represent upper oceanic crust. Although the scattered phases are weak and observable only in a limited range of source-receiver geometries, they provide the structural marker necessary to uniquely establish the plate boundary stratigraphic context for LFEs and tremor.