Exciting news! We're transitioning to the Statewide California Earthquake Center. Our new website is under construction, but we'll continue using this website for SCEC business in the meantime. We're also archiving the Southern Center site to preserve its rich history. A new and improved platform is coming soon!

Path and site effects in GMPEs: Incorporating crustal physical properties for region-specific ground motion estimation using small magnitude data from Southern California

Valerie J. Sahakian, Annemarie S. Baltay, Tom C. Hanks, Janine Buehler, & Frank L. Vernon

Published July 31, 2017, SCEC Contribution #7367, 2017 SCEC Annual Meeting Poster #084

The purpose of this work is to reduce aleatory uncertainty in ground motion prediction equations (GMPEs) by including knowable, physical processes or properties of path effects; in effect, reclassifying this uncertainty as epistemic, and improving both the accuracy and precision of region-specific GMPEs. Relationships between ground motions and path physical properties are sought in an empirical sense, by taking a large (>100,000 recordings) database of small M events (mostly 1 < M < 3.5) from Southern California in the year 2013, and investigating correlations between ground motions and material properties along each recording’s raypath. For every recording we compute: an empirical GMPE residual (difference between observation and predicted ground motion) attributable to the path with a mixed effects approach; a 3D raypath from source to receiver; and integrated material properties along the raypath. We then compare the metric of material property with each recording’s path residual. Both magnitude- and distance-dependence are tested in these comparisons. The metric of the gradient of velocity along the raypath correlates weakly to moderately with path residuals, with correlations increasing at longer source-to-receiver distances (> 80km). We find that although there is no correlation with near-field paths, the site residuals correlate with the gradient of velocity metrics. Along with the observation that the site residuals show no correlation with Vs30 values, this indicates that near-field path effects may be synonymous with site effects, and in fact the “site” affects a larger volume that simply the top 30m of the Earth’s crust. Finally, to pursue this distinction between these regional near-field path and site effects, we consider the same material properties within 5, 10, 15, and 20km of the site. Increasing correlations with site residuals at shorter distances indicates that that near-field ground motions are highly affected by more than just near-surface effects. The physical correlation of empirical path terms with independent velocity models implies that regionalized GMPEs can be built to include propagation path information in new regions or areas with sparse seismicity, resulting in more accurate and more precise region-specific GMPEs.

Key Words
GMPEs, Path Effects, Ground Motion, Seismic Hazard

Citation
Sahakian, V. J., Baltay, A. S., Hanks, T. C., Buehler, J., & Vernon, F. L. (2017, 07). Path and site effects in GMPEs: Incorporating crustal physical properties for region-specific ground motion estimation using small magnitude data from Southern California. Poster Presentation at 2017 SCEC Annual Meeting.


Related Projects & Working Groups
Seismology