This repository contains the Python scripts associated to the paper:
Ducellier A., Creager K.C. Depth and thickness of tectonic tremor in the northeastern Olympic Peninsula, to be submitted.
The tremor catalogs used in the paper are the mat files in data.
Figure 1 of the paper is done using the files in figures.
The cross-correlation of the seismic recordings are done with src/stack_ccorr_tremor.py. The corresponding output files should be stored in src/cc but are not on the Github repository due to their large volume. They are available on request.
The clusterings of the time windows are done with src/cluster_select.py. This Python script has also been used to make Figures 2 and 3.
The computation of the corresponding time lags is done with src/get_timelag.py.
The files containing the time lags for all the tremor-array locations are the src/XX_timelag.pkl, where XX is the two-letter code for the array.
The scripts to compute the depth given a timelag and a 1D velocity model are in misc.py. The corresponding velocity models are stored in the src/ttgrid_*.pkl.
The corresponding depths of the tremor are computed using src/plot_envelopes.py. Figure 4 and the figures in the supplement are also made with plot_envelopes.py.
The difference in timelags and depths for the EW and NS components are also computed with src/plot_envelopes.py, and then plotted using plot_diff_EW_NS.py.
The variations of the depth with the Poisson's ratio is plotted using src/plot_variations.py.
The corresponding thicknesses of the tremor zone are computed using src/plot_thickness.py.
The input files for the maps are then created using src/write_files_depth.py and src/write_files_thickness.py.
The GMT scripts to plot the maps are in src/map_depth and src/map_thick.
The linear regression to fit the plane is done with src/fit_plane.py.
The two notebooks in notebooks explain in more detail how we have downloaded and preprocessed the seismic data, and how we have computed and stacked the cross-correlations.