gamma-phase-associator

You can associate seismic phase picks by using a Gaussian Mixture Model to cluster P and S phase picks across multiple stations. This process estimates earthquake event locations, origin times, and magnitudes from raw phase pick and station data inputs.

Bayesian phase association is an unsupervised clustering technique that uses a Bayesian Gaussian Mixture Model to group seismic phase picks. It constructs earthquake events by estimating location, origin time, and magnitude from multi-station pick data.

You can cluster unknown quantities of earthquake phase picks using unsupervised Gaussian Mixture Models. The association function processes phase picks and station metadata to automatically group picks into distinct earthquake events.

Yes, you can use a Gaussian Mixture Model for earthquake clustering across multiple stations. The method supports flexible configuration for dimensions and velocity models to associate phase picks and reveal earthquake events.

You can tune DBSCAN pre-clustering parameters by using the estimate_eps utility. This auxiliary function helps configure the pre-clustering step before applying Bayesian or standard Gaussian Mixture Models for phase association.

GaMMA phase association requires a picks DataFrame, a stations DataFrame, and a config dictionary. These inputs allow the association function to cluster phase picks and output earthquake locations, origin times, and magnitudes.