Utility of Stochastic Ground Motion Models in Conditional Spectrum-Based Selection Consistent with the Causal Earthquakes

ABSTRACT

Conditional Spectrum (CS), characterized by a conditional mean and conditional standard deviation, serves as a target spectrum linking seismic hazard information to ground motion selection for seismic demand analysis. Current Ground Motion Selection and Modification (GMSM) methods aim to align with this target spectrum but face challenges due to the limited availability of recorded ground motions and potential modifications, leading to difficulties in ensuring selected ground motions accurately reflect site characteristics. Addressing this issue, the study presents an algorithm to select hazard-consistent CS-based records utilizing a stochastic ground motion model with a dual objective: (1) alleviate the need for scaling during the selection process, and (2) select records consistent with the target hazard and the contributing causal scenarios. To achieve both objectives, the study generates a database of hazard-targeted ground motions utilizing the kriging surrogate with scenarios sampled from the disaggregation matrix. Then, a postprocessing step explicitly considers causative parameters in the selection process. The algorithm’s potential to select site-specific ground motions is demonstrated using sites in the Western United States, providing insights into the computational cost and accuracy. Additionally, statistical comparisons are conducted to explore circumstances where similar hazard consistency can be achieved without the postprocessing step, reducing the overall computational cost. Finally, the selected ground motions are favorably compared to those selected through traditional conditional spectrum-based approaches based on their spectral shape and ground motion intensity measures.

KEYWORDS:

• Ground motion selection
• conditional spectrum
• hazard consistency
• causal earthquake scenarios
• stochastic simulations
• surrogate modeling

Acknowledgments

The authors thank the Graduate School at Texas Tech University for providing financial support through the AT&T Chancellor Fellowship to N.S. The computations presented in the work were performed using the Red Raider computing cluster at Texas Tech University. The authors thank the High-Performance Computing Center (HPCC) team at the Texas Tech University for their generous support. In addition, the equipment support from the Vice President for Research & Innovation for T.L.’s Multi-Hazard Sustainability (HazSus) Research Group is gratefully acknowledged.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Data Availability Statement

Data sharing is not applicable to this article as no new data were created or analysed in this study. The results can be reproduced using the illustrative stochastic ground motion model by Rezaeian and Der Kiureghian (2010).

Additional information

Funding

This work is supported in part by the Edward E. Whitacre, Jr. College of Engineering at Texas Tech University, under funds awarded to the Principal Investigator T.L. – corresponding (second) author. The funds provide conference participation and programming bootcamp training for the lead student author N.S.