An empirical model for estimating horizontal acceleration Fourier spectra for the Imperial-Mexicali Valley region

Abstract

An empirical model for estimating the horizontal acceleration spectra is obtained using strong motion records from stations located in the Imperial Valley, California and the Mexicali Valley, Baja California regions. The functional form of the model is based on the regression relation used by Joyner and Boore (1981) to derive attenuation relations for peak horizontal velocity and acceleration. The model adopted scales the acceleration spectra in terms of the magnitude, and the closest distance to the surface projection of the fault that generated the earthquake. To account for the effect of the local site conditions, a frequency-dependent correction term for each station is also incorporated in the model. The coefficients of the model were determined using a two-stage regression procedure. The data set used consists of 44 events, in the magnitude range between 2.9 and 6.8, and distances between 0.2 and 136 km. The dependence of the spectral amplitudes on magnitude was tested using linear and squared dependence. ·The use of a squared term to scale the amplitudes improved the accuracy of the model. In addition, the site corrections show a considerable variation with frequency and between sites. This observation emphasizes the importance of using site specific corrections for predicting ground motion spectra if.they are available.