Towards Improved stress drop measurement: a detailed comparison of contrasting approaches
Rachel E. AbercrombiePublished December 10, 2018, SCEC Contribution #9035
The large uncertainties and scatter in stress drop estimates affect strong ground motion prediction and limit our understanding of the physics of earthquake rupture. We study two complementary approaches to investigate sources of consistency and discrepancies in stress drop estimates, and quantify uncertainties: (1) the spectral decomposition method of Shearer et al. [2006] and Trugman & Shearer [2017], a large-scale approach involving stacking and averaging spectra to obtain parameters for large catalogs of events; and (2) the smaller-scale empirical Green's function (EGF), spectral-ratio method of Abercrombie [2014] and Abercrombie et al. [2017], designed to obtain optimal results for the best-recorded earthquakes, analyzed individually. In comparison to the second method, the first has the advantage of stability, from using larger numbers of events (both by M range and spatial volume) in one inversion, but the disadvantage of using only one global EGF to correct for relatively large regions, potentially leading to bias. Both methods are dependent on the appropriateness of the assumed source model.
We apply the two methods independently to clusters of earthquakes in Southern California. We find strong correlation between the results, but also significant differences. We compare the EGF spectral ratios with the corresponding ratios of event-terms from the spectral decomposition and find them to be very similar, implying that the discrepancies result from the different modeling approaches.
We then analyze the results in detail to investigate the most likely sources of discrepancy including whether (a) spatially varying global EGFs are needed even within a cluster to remove systematic bias in the spectral decomposition approach, (b) the magnitude and frequency range of the data are sufficient to resolve observed trade-offs between source model and stress drop scaling, and (c) the lack of constraints on the EGF corner frequency in the spectral-ratio method causes systematic bias and increased uncertainties. Preliminary results using synthetic data suggest that a relatively wide magnitude range is needed to resolve source scaling. We also find clear depth dependence in the high frequency amplitudes of the event terms, implying that using a global EGF will produce an artificial depth dependence in stress drop.
Citation
Abercrombie, R. E. (2018, 12). Towards Improved stress drop measurement: a detailed comparison of contrasting approaches. Oral Presentation at 2018 Fall Meeting, American Geophysical Union.
Related Projects & Working Groups
Testing and Reconciling Stress Drop and Attenuation Models for Southern California, Seismology
