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 #054, Seismology

Connecting the Co- and Post-seismic Fault Behavior in Ridgecrest-Coso Region with the Pre-earthquake Seismicity (2008-2019): a Revisit with AI Phase Picker

Yijian ZHOU
Poster Image: 

Poster Presentation

2022 SCEC Annual Meeting, Poster #054, SCEC Contribution #12387 VIEW PDF
The co- and post-seismic processes cause most seismic hazard, but it is unclear how much they are connected to the pre-seismic phenomena. The 2019 Mw 7.1 Ridgecrest earthquake provides a unique opportunity to investigate the connection between the two processes, because the Southern California Seismic Network maintains high-quality data for a decadal scale. We construct a high-resolution seismic catalog during 2008-2019 with a hybrid AI picker, which is composed of a CNN for event detection and an RNN for phase picking. The picking performance of our AI method is tested on the aftershocks, which realizes comparable detectability with that of matched filter, but is more generalized and effici...ent.

Our final catalog of the pre-seismic period since 2008 consists of ~100,000 events of ML -1~4.5. On the ruptured segment, sparse seismicity was observed by previous catalogs, while our new catalog reveals this unmapped conjugate fault system. The depth distribution concentrated at 12-15 km, which is consistent with that of the aftershocks, indicating a stable locked patch. The Ridgecrest earthquake triggers a seismic swarm and a slow-slip events on the central Garlock fault. The pre-earthquake seismicity shows that the western end of central Garlock has long-term swarm activity, and is partially locked at 0-12 km; the eastern end is decoupled down to 15 km, and two repeating sequences are detected surrounding the creeping area. Thus, the seismic / aseismic nature of Garlock fault does not alter before and after the earthquake. The Coso geothermal area on the north of Ridgecrest is featured by intense and shallow (<5 km) microseismicity. The Mw 7.1 mainshock enhanced the seismic rate in the surrounding area of Coso field, while seismic activity within the field remains more or less constant. This different triggering behavior is explained by a depletion of shear stress caused by geothermal production. We calculate the spatiotemporal b-value in Coso area to indicate its stress level distribution. Results show that the area outside and within Coso area have a significant contrast in b-value, and that the inner area show a gradually increasing b-value, which is consistent with a depleting shear stress. In summary, our results show that the co- and post-seismic phenomenon can be better interpreted in light of the pre-earthquake seismicity.