From the command line

SCEDC event 38695658, the M4.5 South El Monte earthquake of 19 September 2020. DB is the Green’s-function database directory, provided by the operator. One command does every step:

xcap3d run 38695658 --gf DB --mt full --refine --report

or one step at a time:

xcap3d download 38695658        # the event, StationXML and raw waveforms from SCEDC
xcap3d process 38695658         # the instrument responses removed, rotated to Z, R, T
xcap3d invert 38695658 --gf DB --mt full --refine --report

--refine follows the 5-degree grid search at each depth with a 1-degree search near its minima. Everything goes to ./xcap3d-out/38695658/: result.json, report.html and the report’s figures, shown below. The run ends with:

xcap3d-out/38695658/report.html
38695658 production: depth 15.5 km  FM 263/43/92  Mw 4.56  VR 67.8

FM is strike/dip/rake in degrees and VR the variance reduction in percent. xcap3d report 38695658 writes the report again from result.json.

Stations

The stations within 100 km in four groups, counted in the key: used (blue, the area the summed weight of their five components), deselected (filled gray: removed by the QC or the source-margin rule, each reason in the report’s station table), not in the database (open circles) and without data (gray crosses). The solution is drawn at the epicenter.

Depth

The misfit at each trial depth, in percent above the minimum, with the best double couple of each depth along the top. The chosen depth, 15.5 km, is filled and the runner-up hollow; the shaded band holds the depths whose misfit is within 10 % of the best, and the dashed line is the catalog depth.

Waveform fits

Observed (gray) and synthetic (blue) velocity of the 16 nearest of the 63 stations fitted (56 used); the report’s figure has one row per station. Of the 65 stations loaded, RUS (too close to the source) and MUR (the source near the edge of its store) are left out before the fit. The windows are Pnl (Z, R) and surface waves (Z, R, T); gray rows are stations the QC removed. Above each trace is the time shift of the synthetic in seconds, below it the cross-correlation and, where the QC reduced it, the weight.

Full moment tensor

With --mt full the run also searches the isotropic and CLVD parts at the chosen depth, around the double-couple solution. The lune of Tape and Tape (2012) shows the best variance reduction of each source type, and the star is the solution: ISO 1.3 %, CLVD 0.4 %, DC 98.3 %, close to a double couple. The run prints it as:

full: FM 263/43/92  iso 0.116  clvd 0.064  Mw 4.56  VR 68.2

The signs and the percentages are defined on the Conventions page.