Expanding the Seismicity Catalog in San Fernando Valley with Dense Array Recordings of Micro-Earthquakes

Joses B. Omojola, & Patricia Persaud

Published September 8, 2024, SCEC Contribution #13721, 2024 SCEC Annual Meeting Poster #022

The San Fernando Valley (SFV) is a major population center close to downtown Los Angeles. Following the 1971 (Mw 6.7) San Fernando event, significant changes in construction practices and regulations were implemented to make the SFV area more resilient to future earthquakes. However, a new blind fault was identified when the 1994 (Mw 6.7) Northridge earthquake hit, causing significant damage across the surrounding area. Prior studies of industry seismic profiles have mapped a network of normal and blind thrust faults beneath the basin, and seismicity has been typically linked to thrust fault zones. Fifty-two years after the San Fernando event, we seek to identify contemporary seismic hazards that have not been captured by regional networks, which can help to improve earthquake preparedness. Urbanization limits the possibility of active seismic surveys for imaging the subsurface, and we opted for monitoring of the local seismicity using a dense array installed by volunteers. Our array, comprising of 140 three-component nodal seismic instruments, was installed in October 2023 and recorded continuously for ~30 days.

Earthquake clusters are useful for identifying and characterizing blind faults; however, high-noise levels in urban settings can suppress the signal-to-noise ratio across long distances limiting the effectiveness of regional networks in detecting low-magnitude events. Our nodal array with interstation distances ranging from 300 m along the dense profile to 2 km across the distributed array is better suited at closing data gaps for events not detected by regional networks. To improve the detection of small events in noisy data, we applied a pretrained U-Net model to our dataset. P- and S-phase probabilities were predicted at individual stations using 6-s long waveforms with overlap. Detections were associated using constant seismic velocities and a 1D velocity model with PyOcto. Preliminary event locations were computed with hypoinverse and hypoDD. We detected hundreds of small earthquakes across the array and relocated 53 high-quality events. Relocated events are clustered within proximity of major faults. We also identified additional event clusters which were not detected by the regional network, that could indicate unmapped blind faults. Our study aims to improve the characterization of seismic hazards using modern instrumentation and fills data gaps not currently addressed by regional networks.

Citation
Omojola, J. B., & Persaud, P. (2024, 09). Expanding the Seismicity Catalog in San Fernando Valley with Dense Array Recordings of Micro-Earthquakes. Poster Presentation at 2024 SCEC Annual Meeting.

Related Projects & Working Groups
Seismology