Robust Analysis of Stress Drop Variation along San Andreas Fault at Parkfield Using Multiple Local Networks

Jiewen Zhang, Xiaowei Chen, & Rachel E. Abercrombie

Published December 10, 2018, SCEC Contribution #9044

Earthquake stress drop measures the stress status changes on a fault before and after an earthquake, which is related to ground motion prediction and source physics. Estimation of stress drop can be subject to several factors of influence, e.g. the sampling capability of instruments to resolve corner frequencies, earthquake source complexity, method limitations, etc. Accumulated research on San Andreas Fault at Parkfield area has revealed the general structure and rock properties of the major fault, but stress drop accuracy still remains questionable as a result the factors above. In our research, we collect seismic records from both surface and borehole networks within a long-time span (1984-2016), and perform analysis on borehole stations using both an improved stacking-based spectral decomposition method that solves an empirical-correction-spectrum (ECS) based on a group of selected events, and spectral ratio analysis that is based on empirical-green’s-function (EGF) with carefully selected event pairs to address stress drop uncertainties. For the stacking-based method, we examine the stress drop stability by synthesizing results from both fixed and overlapping spatial windows. Our results suggest that the depth range has significant influence on the resolved stress drop pattern. Apparent depth dependent trend can be removed by implementing depth-dependent ECS correction. Lateral variation is relatively stable in respect of the spatial window selection. A weak magnitude-dependent trend in stress drop is seen for earthquakes with M<=1.5 with borehole stations. Further analysis with high-sampling rate SAFOD borehole stations will be utilized to examine scaling relationship for smaller earthquakes. The current results show that the stacking-based analysis results with stress drops that are systematically lower than spectral ratio based analysis. Further investigation on the surface station data, influence of source complexity and data quality control on stress drop consistency will be conducted.

Citation
Zhang, J., Chen, X., & Abercrombie, R. E. (2018, 12). Robust Analysis of Stress Drop Variation along San Andreas Fault at Parkfield Using Multiple Local Networks. Poster Presentation at American Geophysical Union.