Resolving the spatiotemporal variability of small earthquake source parameters at Parkfield and their relationship with 2004M6 Parkfield earthquake

Jiewen Zhang, Xiaowei Chen, & Rachel E. Abercrombie

Published August 15, 2019, SCEC Contribution #9803, 2019 SCEC Annual Meeting Poster #088

Parkfield has long been a natural, well-studied laboratory for investigating spatial and temporal changes in fault properties and their relation to large earthquakes. Previous studies have found possible changes in stress drop for small events with time and space but are limited by using only a small number of earthquakes or only surface stations that lack high-frequency signal, leading to significant uncertainties in the results. In our research, we perform a comprehensive analysis of the spatiotemporal variation in earthquake stress drops using both surface and borehole data, and investigate the effects of using different data, frequency range, and modeling assumptions on the stress drop estimates.

We follow the large-scale stacking approach developed by Shearer et al., (2006), and previously applied to regional Northern California Seismic Network (NCSN) at Parkfield by Allmann and Shearer (2007). We use ~5000 earthquakes (2001-2016) recorded by the High-Resolution Seismic Network (HRSN) in 4-60Hz range and ~6000 earthquakes (1984-2016) recorded by regional surface network in 4-40Hz range, with wider frequency band above noise for higher ability to resolve small earthquakes.

To account for spatial variation in path effects from heterogeneity, we partition the fault both along strike and with depth, and invert for independent EGFs within each bin. Our results show that the smallest earthquakes (M<1 for borehole network and M<1.5 for surface network) appear to scale with magnitude, which may be ascribed to strong trade-off between attenuation and corner frequency. Then we try an alternative approach to calculate the EGF for each bin, and fix the stress drop of the lowest magnitudes, resulting in considerably less magnitude-scaling stress drops. And the assumption of depth-dependent rupture velocity helps decrease the scatter of stress drop estimates. Based on above, we find consistency of stress drop spatial pattern between our results and Allmann and Shearer (2007) using only surface stations on larger stress drops around 2004 M6 and smaller stress drops around 1966 M6, and agreement between borehole and surface results only for low corner frequencies (1/3-1/2 of upper bound) due to frequency limitation. We also observe some temporal change of stress drop before and after 2004 M6 by generating time-varying EGFs, consistent with previous results, where NW section of the fault has increased stress drops and those in SE section is decreased.

Key Words
stress drop spatiotemporal change, sampling rate limitation, fault heterogeneity, spectral stacking method

Citation
Zhang, J., Chen, X., & Abercrombie, R. E. (2019, 08). Resolving the spatiotemporal variability of small earthquake source parameters at Parkfield and their relationship with 2004M6 Parkfield earthquake. Poster Presentation at 2019 SCEC Annual Meeting.

Related Projects & Working Groups
Seismology