SCEC collaborators from Harvard and Caltech have developed a new, statistical way to rapidly identify the most likely fault (or sets of candidate faults) in the Community Fault Model (CFM) that generated an earthquake. This information is now available near real-time through an automated web-based service with email notifications.

For the first time ever, the information provided by sophisticated, near real-time seismic networks is automatically connected with the comprehensive 3D Community Fault Model developed by SCEC researchers. The method of earthquake-to-fault association was developed using comprehensive earthquake hypocenter and focal mechanism datasets in California to assess what properties of earthquakes serve as the best predictors of the fault on which they occurred. The development team used a series of training datasets for earthquakes known to have occurred on faults within the fault model, and established that proximity (distance), focal mechanism (nodal plane orientation), and earthquake history (spatial and temporal clustering) can be combined to assign a probability that a given earthquake was associated with one or more source faults in the fault model (or on a fault not included in the model – such as the 2019 Ridgecrest earthquakes). Notably, these training datasets were comprised of earthquakes that occurred in the decade since the release of CFM 2.0, to ensure that they did not influence the modeled fault geometries.

The method is implemented as an R script that calculates distances between earthquakes and CFM faults, compares nodal plane orientations, and evaluates recent earthquake clustering to produce association probabilities. The approach has been applied to the full SCSN catalog (M ≥ 3.0) using CFM (5.2), and probabilities of association between every earthquake in the catalog and each fault in the CFM are available at the Southern California Earthquake Data Center (SCEDC). For each earthquake in the catalog, the code outputs the five highest probabilities of association with a CFM fault, as well as the probability that the earthquake is not associated with any source within the fault model. The majority of earthquakes (> 80%, above M 3) have a high probability of association with one or two faults in the model. In addition, this tool has been implemented as a wrapper script (Eq2CFM) integrated into the AQMS seismic data processing system at Caltech. This enables the Eq2CFM script to run automatically when a M≥3.0 event with a focal mechanism is processed or post-processed. The wrapper script sends out an email a few minutes after the event with an abbreviated message stating the probability of association with the nearest fault.

Objective earthquake-to-fault associations provide a direct measure of the activity of faults within the southern California plate boundary. The fault associations can facilitate studies of the completeness of the CFM and identify potential sources that should be added to the model. Rapid implementation of the tool can help to identify clusters of small earthquakes that may be foreshocks of a larger, imminent event on a major fault. Earthquake-to-fault associations can be used to communicate objective information about the faults that source earthquakes to interested audiences. More accurate identification of source faults for large and small earthquakes will improve understanding these hazards and help us to mitigate their risks.

Subscribe to SCEC CFM Fault Association Notifications

Members of the SCEC community can now sign up to receive SCEC CFM association information through email. This service will notify you each time there is an ≥M3 event and provide fault associations. We have implemented an aftershock suppression system, so your inbox doesn’t get overloaded! You’ll need a SCEC.org account to receive notifications. To subscribe to this service, update your SCEC user profile to include “CFM Fault Association Notifications”. Check the box next to the Mailing List Subscriptions and click “SAVE” at the bottom of the page.

Development Team

Walker Evans, Harvard University
Andreas Plesch, Harvard University
John H. Shaw, Harvard University
Men-Adrin Meier, California Institute of Technology
Egill Hauksson, California Institute of Technology
Ellen Yu, California Institute of Technology
Edric Pauk, University of Southern California
Tran Huynh, University of Southern California

More Information

Evans, W.S., A. Plesch, J. H. Shaw, N. L. Pillai, E. Yu, M. Meier, and E. Hauksson, A statistical method for associating earthquakes with their source faults in southern California, Bulletin of the Seismological Society of America, 110 (1): 213–225, doi 10.1785/0120190115 SCEC Contribution 9057.

Acknowledgements

This research was supported by the Southern California Earthquake Center (Contribution No. 9057). SCEC is funded by NSF Cooperative Agreement EAR-1600087 and USGS Cooperative Agreement G17AC00047.