Ground-Motion Variability from Kinematic Rupture Models on the Wasatch Fault Zone, with Implications for Probabilistic Seismic Hazard Analysis

Grace A. Parker, Morgan P. Moschetti, & Eric M. Thompson

Published September 11, 2022, SCEC Contribution #12227, 2022 SCEC Annual Meeting Poster #225

We consider the effect of changing kinematic rupture parameters on simulated ground motion variability for M7 earthquakes on the Salt Lake City segment of the Wasatch Fault Zone. We use a mixed-effects analysis to estimate the contribution to ground motion variability from each of five kinematic parameters used in rupture generation—the correlation length and random seed of the slip distribution, average rupture speed, rise-time scaling or slip-velocity relations, and hypocenter location. Ruptures were computed with the kinematic rupture generator of Frankel et al. (2014) and convolved with 3D Green’s functions computed from the Wasatch Front community velocity model. We find that the total variability from the simulations is approximately equivalent to that from the NGA-W2 ground motion models (GMMs), although the median simulated ground motions are larger than median GMM predictions by about a factor of 1.3. We estimate contributions from the kinematic parameters to the variability, which suggest approaches to improving future simulations and reducing the uncertainty in site-specific earthquake hazard studies. Specifically, we find that the chosen slip velocity values have a large effect on the ground motion variability, and result in a between-event variability larger than that of the NGA-W2 GMMs. This misfit could arise from use of values that are inconsistent with observations from earthquakes in similar regions (i.e., slip velocities that are too high or low), or from the implementation of variations in slip velocity as uniform adjustments of the average. Additionally, the simulations have strong spatial heterogeneities in variability due to directivity effects. We illustrate the impact of this using a partially nonergodic probabilistic seismic hazard analysis (PSHA), where information about the earthquake source, path attenuation, and site response are paired with reductions in the GMM variability. We find that two sites equidistant from the finite fault can have significantly different hazard curves due to large variations in ground motion variability between the locations. These results highlight the need for directivity effects on the median and standard deviation of ground motions to be included in nonergodic PSHA. If directivity effects are not accounted for in the median GMMs, branches with increased variability should be included in the PSHA logic tree to appropriately capture the full epistemic uncertainty.

Key Words
kinematic rupture models, ground motion variability, nonergodic PSHA, directivity effects

Citation
Parker, G. A., Moschetti, M. P., & Thompson, E. M. (2022, 09). Ground-Motion Variability from Kinematic Rupture Models on the Wasatch Fault Zone, with Implications for Probabilistic Seismic Hazard Analysis. Poster Presentation at 2022 SCEC Annual Meeting.

Related Projects & Working Groups
Ground Motions