Methodology for validation of simulated ground motions for seismic response assessment: Application to cybershake source-based ground motions

A comprehensive methodology for the validation of simulated ground motions is presented. The suggested methodology can be geared toward any ground-motion simulation method and seismic response assessment, in a target engineering application. The methodology is founded on the comparison between conforming groups of ground-motion waveforms from recordings and simulations and their effect on a representative collection of structures that represent the engineering application. The comparison considers the statistics of earthquake scenarios at the level of the event and site parameters, the resulting waveform characteristics, and the subsequent structural responses. Regression models are developed at three levels (between structural responses and waveform characteristics, structural responses and event and site parameters, and waveform characteristics and event and site parameters). Similarities between the models from groups of recorded and simulated ground motions guide the validation process. The validation methodology is applied to CyberShake (v.15.12) simulations and for the estimation of the column drift ratio of a bridge structure. It is shown that CyberShake (v.15.12) can be used to assess the median seismic response of the used bridge. Some discrepancies between simulations and recordings are observed, which could be attributed to the basin and site-response models used for simulations. Further implementation and refinement of the suggested methodology are recommended to make broader conclusions.