Ground-Motion Prediction Equations for Central and Eastern North America Using the Hybrid Empirical Method and NGA-West2 Empirical Ground-Motion Models

The hybrid empirical method (HEM) of simulating ground-motion intensity measures (GMIMs) in a target region uses stochastically simulated GMIMs in the host and target regions to develop adjustment factors that are applied to empirical GMIM predictions in the host region. In this study, the HEM approach was used to develop two new ground-motion prediction equations (GMPEs) for a target region defined as central and eastern North America (CENA), excluding the Gulf Coast region. The method uses five new empirical GMPEs developed by the Pacific Earthquake Engineering Research Center for the Next Generation Attenuation-West2 (NGA-West2) project to estimate GMIMs in the host region. The two new CENA GMPEs are derived for peak ground acceleration and response spectral ordinates at periods ranging from 0.01 to 10 s, moment magnitudes (M) ranging from 4.0 to 8.0, and rupture distance (R-rup) as far as 1000 km from the site, although the GMPEs are best constrained for R-rup < 300-400 km The predicted GMIMs are for a reference site defined as CENA hard rock with V-S30 = 3000 m/s and kappa(0) = 0.006 s. The seismological parameters for the western North America host region were adopted from a point-source inversion of the median GMIM predictions from the NGA-West2 GMPEs for events and sites with M <= 6.0, R-rup <= 200 km, V-S30 = 760 m/s, a generic (average of strike slip and reverse) style of faulting, and earthquake-depth and sediment-depth parameters equal to the default values recommended by the NGA-West2 developers. The two CENA GMPEs are based on two fundamentally different approaches to magnitude scaling at large magnitudes: (1) using the HEM approach to model magnitude scaling over the entire magnitude range and (2) using the HEM approach to model magnitude scaling for events with M <= 6.0 and using the magnitude scaling predicted by the NGA-West2 GMPEs for the larger events.