Earthquake ground-motion prediction equations for northern Iran

Earthquake ground-motion prediction equations for soil and rock sites in northern Iran have been developed based on stochastic models and Bayesian updating. Due to a lack of recorded data, the well-known simulation methodology, finite-fault model, including estimates of the inherent uncertainty of ground-motion parameters, has been used for generating more than one thousand strong motions as input data. The Bayesian approach is an effective approach that allows the combination of knowledge of seismological theory with recorded data. Estimation of the prior information is one of the most controversial issues in a Bayesian approach. In this study, generated data based on the stochastic simulation model is first used to derive the prior coefficient of earthquake ground-motion prediction equations. The prior coefficients are updated within the Bayesian approach framework by using the recorded ground motion in northern Iran. The residual plots show that the updated prediction equations agree well with available northern Iran ground-motion data. Additionally, the proposed prediction equation is validated by comparing the estimated ground motion with those of recorded data at the observed stations. Copyright Springer Science+Business Media Dordrecht 2013