Ground Motion Prediction Equations for Cumulative Absolute Velocity (CAV) Using the PEER-NGA Strong Motion Database

摘要

It has been shown that cumulative absolute velocity (CAV), defined as the sum of the absolute acceleration time series, is a more stable predictor of ground motion than peak or response spectral parameters. We present empirical ground motion prediction equations (GMPEs) for two variants of CAV using the database and functional forms we used in the PEER Next Generation Attenuation (NGA) Project. We consider our CAV relationships to be valid for magnitudes ranging from 5.0 up to 7.5-8.5 (depending on fault mechanism) and distances ranging from 0-200 km for shallow crustal earthquakes in active tectonic regimes. We find that the standard deviations of these GMPEs are smaller than those of any ground motion parameter we have studied thus far, including Arias intensity. We propose that a variant of CAV could serve as a single duration-based ground motion index to aid in the rapid assessment of ground shaking severity.