Abstract

A large number of ground-motion time histories is required for a reliable probabilistic seismic-hazard analysis. However, insufficient ground-motion records has been a major issue in earthquake engineering even in high-seismicity areas. Stochastic and seismological ground-motion models are used to simulate time histories in order to overcome this issue. Many times these models require a large number of records for calibration, are available only for regions with a wealth of data, or fail to characterize well specific sites.The current paper introduces a novel stochastic ground-motion model that allows simulation of realistic non-stationary, non-Gaussian time histories that are statistically consistent with data even at sites with very low numbers of records. This goal is achieved by combining the prior knowledge from a seismological model, calibrated to global data, with the site-specific data within a Bayesian framework. The proposed model uses the specific barrier model as the seismological model and numerical results consistent with records from a nuclear powerplant site in Romania are presented.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call