We provide ground-motion prediction equations (G MPEs) for the computation of median peak ground motions and response spectra for shallow crustal earthquakes in active tectonic regions. The equations were developed as part of the NGA-West 2 project and are based on a composite data set [Ancheta et al . 2013] that includes global events from 1935 to 2011 spanning a wide magnitude range, plus a large number of small-to-moderate magnitude events from California principally from 1998 to 2011. This data set comprises more than 21,000 records, of which we use a subset of ~16000 records in our analys is. The equations follow a philosophy, functional form, and regression method that is similar to the Boore and Atkinson [2008] (BA08) ground- motion model from the original NGA project (i.e., the “NGA-West 1” project). That approach, which is continued here, provides for what we refer to as base-case GMPEs accounting for ground motion scaling with magnitude, distance, and site condition.
Aside from updating the coefficients based on new data, a principal motivation for the revised regression is that the enhanced database enables improved evaluations of numerous factors, including: the distinction between mainshock and aftershock motions; the influence of depth to top of rupture; basin depth effects; and regional variations of site effects and distance attenuation. Of these, only basin depth and regional apparent anelastic attenuation were found to be statistically significant and were incorporated into the median model. We treat these additional variables as optional adjustment factors to the base-case GMPEs, which maintains the simplicity of the base-case GMPEs for general applications. The enhanced data also enables evaluation of the effects of magnitude, distance, and site condition on standard deviation terms, which are now incorporated into the GMPEs.
The predicted ground-motion amplitudes from this study are generally similar to those from BA08 for larger magnitudes. The most significant differences are for motions from M < 6 earthquakes, where the new predicted motions are less than those from BA08 as modified by Atkinson and Boore [2011], particularly at distances within about 20 km. We were unable to identify significant differences between motions from a mainshock and its aftershocks, and hence we consider our equations applicable for both event types. We find that over the magnitude range of engineering interest ( M > 5) the equations are unbiased with respect to source depth, as represented by either depth to top-of-rupture or hypocentral depth. Hence, a depth term is not introduced. With regard to apparent anelastic attenuation, the base-case GMPEs are unbiased for data from California and Taiwan, but other regions exhibit faster (Italy and Japan) or slower (China and Turkey) attenuation rates for which adjustment factors are introduced. Our equations are found to be centered for basin-depths compatible with an empirical relationship between depth parameter 1z and VS30 , to be negative for depths shallower than the mean of such models, and positive for depths greater than the mean. These changes in ground motions with depth occur for long period ground motions, and applicable adjustment factors are introduced.
As with BA08, the standard deviation is segregated into between-event and within-event components. However, unlike BA08, the between-event component is now M-dependent, decreasing as M increases beyond 5.5. The within-event component now has variable trends with M for different spectral periods (decreases for short periods, increases for long period), increases with distance for RJB > 80 km, and is reduced for soft soil sites.
Our GMPEs are considered applicable over a magnitude range of 3 to 8.5 for strike-slip or reverse-slip events ( M 3 to 7 for normal-slip events), distances up to 400 km, and site conditions ranging from VS30 = 150 to 1500 m/sec and Z1 = 0.0 to 3.0 km. The equations are useful for prediction of the ground-motion intensity measures (IMs) PGA, PGV, and PSA at periods T = 0 to 10 sec.
Full List of PEER Reports: click here.
