Exciting news! We're transitioning to the Statewide California Earthquake Center. Our new website is under construction, but we'll continue using this website for SCEC business in the meantime. We're also archiving the Southern Center site to preserve its rich history. A new and improved platform is coming soon!

Group B, Poster #048, Seismology

How robust are repeating earthquake detections? A comparison of different detection strategies

Norma A. Contreras, Gareth J. Funning, Taka'aki Taira, & Nader Shakibay Senobari
Poster Image: 

Poster Presentation

2023 SCEC Annual Meeting, Poster #048, SCEC Contribution #13193 VIEW PDF
Repeating earthquakes (‘repeaters’) are sequences of periodic or near-periodic collocated earthquakes that produce nearly identical waveforms. Repeaters are commonly found on, and associated with, regions of creep (aseismic slip) on a fault, and can be used to estimate creep rates. Conversely, an absence of repeaters can be indicative of fault locking. Repeater identification typically relies on measures of waveform similarity, such as waveform cross-correlation or coherence, where a specified similarity threshold must be met. In addition, choices of filter pass-band and station coverage can affect the number and quality of the identified events. In this study, we examine the effects of diff...erent detection strategies on repeater detection by comparing different candidate repeating sequences identified on California’s Hayward and Calaveras faults by two different groups using different approaches.
The first (‘HFRC’) approach uses mean waveform coherence to identify repeating event pairs, grouping together pairs that share a common event (e.g. Materna et al., 2018). The final catalog (‘HFRC’) uses a 8-24 Hz frequency band. In the second approach, we apply the Full Automated Repeating Earthquake Search (FARESearch) code (Shakibay Senobari and Funning, 2019). This automated code divides the area of interest, typically a fault-bounding polygon, into subregions and cross-correlates all events within each subregion. Events are grouped into clusters on the basis of pairwise similarity and hierarchical clustering, which accounts for station network changes. We use this approach with two different frequency bands, 1-15 Hz (‘FS15’) and 8-24 Hz (‘FS24’), and compare all three catalogs.
We find the FS24 and HFRC catalogs identify a comparable number of candidate repeaters and clusters. However, when we compare numbers of events in the clusters we find that only one-sixth are identical. The majority of shared events are grouped into different clusters in both catalogs. Compared to the FS24 and HFRC catalogs, the FS15 catalog identifies almost twice as many candidate events. This catalog includes more off-fault events, many of which are two-event clusters. Overall, candidate repeaters in all three catalogs cluster along the same fault segments, with the exception of the off-fault clusters. Validating repeater clusters using relative relocation and source radii overlap is needed.