Detecting Very-Low-Frequency Earthquakes in Parkfield, California: Evidence for Slow-Slip Events

Andres Felipe Felipe Peña Castro, McGill University, Montreal, QC, Canada, Rebecca M Harrington, Ruhr University Bochum, Institute of Geology, Mineralogy and Geophysics, Bochum, Germany, Elizabeth S Cochran, U.S. Geological Survey, Earthquake Science Center, Pasadena, United States, Xiangfang Zeng, Institute of Geodesy and Geophysics, Chinese Academy of Sciences, State Key Laboratory of Geodesy and Earth’s dynamics, Wuhan, WI, China, Clifford H Thurber, University of Wisconsin Madison, Department of Geoscience, Madison, WI, United States and David R Shelly, U.S. Geological Survey, Geologic Hazards Science Center, Golden, United States

Contact First Author: Andres Felipe Felipe Peña Castro; andresfpenacastro@gmail.com

Abstract ID#: 33703

 

English Abstract:
Geodetically observed slow-slip events (SSEs) occur in many subduction zones around the world, and are indicative of slip on the plate interface below the brittle-ductile transition, at depths of roughly 30-50 km. SSEs are commonly associated with tremor and low-frequency earthquakes (LFEs) at the up-dip boundary of fault patches hosting SSEs. Another type of seismic signal, termed very-low-frequency earthquakes (VLFs), may potentially represent the seismic signature of SSEs in the region of a fault where tremor and LFEs occur (~20-30 km depth). VLF signals have been observed in association with tectonic tremor and LFEs in the Nankai through subduction zone, and their focal mechanism solutions suggest they result from slip along the plate interface. Here we search for VLF events on the San Andreas Fault near Parkfield, California, where vigorous tremor and LFEs occur in the absence of geodetically observed slip. We test a procedure that analyzes seismograms in the LFE band (2-8 Hz), the VLF band (0.02-0.05 Hz) and the waveform envelope, and stack multiple time windows containing co-located tremor events in the VLF band. We base our stacking time windows on LFE arrival times, and use waveform envelopes to determine peak amplitudes for aligning waveforms in the VLF band. The objective of stacking multiple records is to increase the signal-to-noise ratio of the VLF signal, which is often not visible above the noise level on a single waveform. We use a combination of data from permanent stations and seismic data collected from a temporary deployment of 13 broadband stations installed < 30 km from the San Andreas Fault near Cholame, California in 2010-2011. The postulated VLF origin times result from a LFE catalog calculated using a high resolution 3D velocity model. Our procedure will test wheter VLFs can be detected in the Parkfield region, which will support the existence of slow-slip events that may be below the limits of detection by geodetic methods.
Back to: Understanding Slow Slip and Tectonic Tremor Process Along a Variety of Deformation Boundaries I