Developing a Near-fault Non-ergodic Ground Motion Model for the Ridgecrest, CA, Area

Xiaofeng Meng, Camilo Ignacio Pinilla Ramos, Albert R. Kottke, & Yehuda Ben-Zion

Published September 8, 2024, SCEC Contribution #13873, 2024 SCEC Annual Meeting Poster #179

One major obstacle in earthquake research is the lack of near-fault recordings (<10 km). Abundant near-fault recordings could transform our understanding of the physics of earthquakes and generation of ground motion at very short distances. In the current ground motion models (GMM), the near-fault motions (and hazards) are extrapolated from the data recorded at moderate and large distances assuming linear behavior which is not necessarily correct. However, near-fault seismic motions are crucial for near-fault structures, as they impose the most severe hazards due to the strongest shaking and possible significant displacement.

Starting at ~8 days following the M7.1 Ridgecrest, CA, earthquake, 15 1D or 2D dense arrays (461 sites) were deployed in the area of the rupture zone, which include four 1D arrays deployed across the surface rupture of the mainshock. The dense arrays recorded continuously for ~30 days and captured numerous aftershocks between magnitude 0 and 5.2. The majority of recordings have epicentral distances smaller than 20 km, with the smallest distance being ~0.1 km, which make it an unprecedented dataset for studying near-fault ground motions.

In this work, we first compute PGA and PGV caused by aftershocks listed in a relocated earthquake catalog, after removing signals from overlapping aftershocks and cars. Next, we combine the dense array ground motion dataset with an existing, regional Ridgecrest ground motion dataset, and develop a GMM with the mixed-effects regression. The total residuals are decomposed into non-ergodic source, site and path terms. Then, we investigate several aspects of the near-fault ground motions that are caused by source, site or path effects. Our current main findings include: 1) clear radiation patterns across all magnitudes with both PGA and PGV; 2) significant variations in site responses within the fault zone in both fault-normal and fault-parallel directions; 3) strong amplifications of ground motions within the fault zone by trapped and other damage zone waves, which are more prominent with PGV. We are currently computing the PSA and FAS at different periods and will investigate the period dependency of the aforementioned results. The updated results will be presented at the meeting.

Key Words
Ground motion model

Citation
Meng, X., Pinilla Ramos, C., Kottke, A. R., & Ben-Zion, Y. (2024, 09). Developing a Near-fault Non-ergodic Ground Motion Model for the Ridgecrest, CA, Area. Poster Presentation at 2024 SCEC Annual Meeting.

Related Projects & Working Groups
Ground Motions (GM)