Description: My work with tectonic tremor in Cascadia involves calculating the radiated energy rate, or source amplitude, of the tremor signal at the subducting plate interface. This provides more information about the strength of the plate interface downdip of where large, damaging earthquakes can occur.
Techniques: envelope functions, seismic energy inversion, fourier transforms, spatial and temporal clustering, time series analysis
Tools: MATLAB, Bash, IRIS-WS java library
The northern Cascadia subduction zone experiences episodic tremor and slip (ETS) events every ~14 months, during which slow slip occurs along the subduction interface at depths of 35-50 km. Over the course of several weeks, this slip releases seismic energy equivalent to a magnitude 6.1-6.7 earthquake. By examining the seismic signal (tremor), we can gain a better understanding of conditions on the plate interface. We find that tremor source amplitudes increase in size and become much more susceptible to tidal stress modulation during the propagation phase of ETS events, implying a low effective normal stress or low effective friction along the plate interface.
Data: The data that go into this are surface seismic velocities measured at 22 sites along the Olympic peninsula, and tremor locations calculated by Aaron Wech at the University of Washington.
Surface amplitudes (station location)
Method: The raw seismic data are processed with the following steps:
- Deconvolve instrument response
- Bandpass filter
- Calculate envelope function
- Resample to standard, lower sample rate
- Stack arrays (several of the ‘sites’ involved arrays of 10-15 seismometers placed within 1-2 km)
- Stack vertical and horizontal components of motion
- Calculate rms mean, median, and peak amplitudes in 5-minute periods, to correspond to the tremor location time periods
Inversion: Once 5-minute ‘surface amplitudes’ are calculated, we invert these along with the tremor locations to obtain a measure of the ‘source amplitude,’ or the radiated energy rate at the plate interface, taking into account geometric spreading and attenuation effects on the seismic waves.
Analysis: With our source amplitudes calculated, we perform various tests to see how the amplitudes change in space and time, and infer properties of the plate interface from this behavior. For example, in the plots below we see that there are higher peaks at tidal periods (red bars in the lower plots) during the latter, propagation phase of the ETS events. This suggests that the tremor amplitudes are susceptible to modulation by tidal stresses, and implies a low effective stress or low effective friction along the plate interface.