This electronic supplement contains descriptions of the validation events and scenarios, fixed and region-dependent parameters, additional goodness-of-fit plots for ground-motion prediction equations (GMPEs) versus data, simulations versus GMPEs, and pseudospectral acceleration (PSA) bias versus distance, and comparison of rupture models with associated PSA bias and time histories for Landers.
Table S1. Description of part A events. Ztor is depth to the top of the fault. Mechanisms are reverse oblique (REV-OBL), strike slip (SS), and reverse (REV).
Table S2. Description of part B scenarios. Abbreviations are as for Table S1; SOCAL, southern California model; NOCAL, northern California model.
Table S3. Fixed parameters for BBtoolbox V1.5.
Table S4. Region-dependent parameters for BBtoolbox V1.5. Here, c0 is chosen as 2.1 for active regions and as 1.1 for stable regions (Somerville et al., 2009).
Figure S1. PSA bias plots for the observed data versus the four GMPEs used in the validation for Alum Rock: AS08, Abrahamson and Silva, 2008; BA08, Boore and Atkinson, 2008; CB08, Campbell and Bozorgnia, 2008; and CY08, Chiou and Youngs, 2008. The GMPEs overpredict the data in a manner similar to but more pronounced than the SDSU synthetics (see Fig. 3 in the main article), suggesting the presence of a relatively large negative event term for Alum Rock.
Figure S2. Part B of the validation comparing PSAs for 50 realization ensemble synthetics against leading GMPEs for Mw 5.5, 6.2, and 6.6 scenarios, using the southern California 1D velocity model. The plots show the means (squares) and standard deviations (boxes), while error bars show extrema for all realizations.
Figure S3. PSA bias for selected periods versus distance for the Alum Rock scenario. The dots and error bars depict the median and range of values obtained for the 50 realizations.
Figure S4. Same as Figure S3, but for Chino Hills.
Figure S5. Same as Figure S3, but for Landers.
Figure S6. Same as Figure S3, but for Loma Prieta.
Figure S7. Same as Figure S3, but for North Palm Springs.
Figure S8. Same as Figure S3, but for Northridge.
Figure S9. Same as Figure S3, but for Whittier.
Figure S10. Same as Figure S3, but for Niigata.
Figure S11. Same as Figure S3, but for Tottori.
Figure S12. Same as Figure S3, but for Mineral.
Figure S13. Same as Figure S3, but for Riviere du Loup.
Figure S14. Same as Figure S3, but for Saguenay.
Figure S15. Best-fit line (green) and 95% confidence regions (red dashed lines) for period bins of 0.01–0.1 s, 0.1–1 s, 1–3 s, and 3–5 s, including all 12 part A validation events. The y axis is the mean bias in natural log units. Values are for each distance bin plotted with respect to the natural log of the central distance of each bin. Data are weighted by the number of stations and discrete periods in each distance bin. The ratio is the absolute value of the slope of the trend lines (green) divided by its standard error. Values less than 1 indicate that the zero slope lies within the 95% confidence and constitutes a passing criterion. All SDSU results met this criterion.
Figure S16. (Left) Slip and rupture time contours and (right) goodness of fit for the Landers realization with the (top) smallest and (bottom) largest bias. The rupture and slip pattern for the source generating the best fit generally appears in better agreement with the kinematic inversion models obtained for Landers, as compared to the source associated with the poorest fit.
Figure S17. (Left) Locations of stations (circles), the fault (black line), and epicenter (star). (Right) Broadband velocity time histories and Arias intensities for station YER (yellow circle in map) for the Landers realization with the smallest bias (data are in black, synthetics are in red).
Abrahamson, N. A., and W. J. Silva (2008). Summary of the Abrahamson & Silva NGA Ground-Motion Relations, Earthq. Spectra 24, no. 1, 67–97.
Boore, D. M., and G. M. Atkinson (2008). Ground-motion prediction equations for the average horizontal component of PGA, PGV, and 5%-damped PSA at spectral periods between 0.01 and 10.0 s, Earthq. Spectra 24, no. 1, 99–138.
Campbell, K. W., and Y. Bozorgnia (2008). NGA ground motion model for the geometric mean horizontal component of PGA, PGV, PGD and 5% damped linear elastic response spectra for periods ranging from 0.01 to 10 s, Earthq. Spectra 24, no. 1, 139–171.
Chiou, B. S.-J., and R. R. Youngs (2008). An NGA model for the average horizontal component of peak ground motion and response spectra, Earthq. Spectra 24, no. 1, 173–215.
Somerville, P., R. Graves, N. Collins, S. G. Song, S. Ni, and P. Cummins (2009). Source and Ground Motion Models for Australian Earthquakes, Proc. 2009 Annual Conference of the Australian Earthquake Engineering Society, Newcastle, Australia, 11–13.
[ Back ]