An empirical estimate of near-source ground motion for a major, mb = 6.8, earthquake in the eastern United States

Abstract

Abstract Three empirical techniques are used to estimate ground-motion amplitudes in the near-source region of a major earthquake hypothesized to occur along the east coast of the United States. These techniques include the estimation of strong-motion parameters from the statistical analysis of selected strong-motion records, from empirical and theoretical attenuation relations, and from Modified Mercalli intensity. Source parameters for the hypothesized earthquake were selected to be similar to those inferred for the 1886 Charleston, South Carolina, earthquake. The epicentral region is characterized by a Modified Mercalli intensity of X, from which magnitudes of mb = 6.8, Ms = 7.2, and M = 7.2 were estimated. Inferred fault mechanisms for the hypothesized event vary from strike-slip on a vertical fault to reverse or reverse-oblique slip on a fault dipping as shallow as 45°. The median estimate of peak horizontal acceleration for a point located 15 km from the surface projection of fault rupture is 0.30 g, independent of site geology. Corresponding estimates of peak horizontal velocity are 35 cm/sec for soil and soft sedimentary rock sites and 18 cm/sec for crystalline basement rock sites. Strong-motion parameters for a point located 15 km from the epicenter, corresponding to a point located directly above the rupture zone of the hypothesized earthquake, are estimated to be approximately 80 per cent higher than these estimates, or about 0.55 g for peak acceleration and about 63 and 33 cm/sec for peak velocity on soil/soft rock and crystalline rock, respectively. Corresponding velocity-acceleration ratios are 115 cm/sec/g for soil/soft-rock sites and 60 cm/sec/g for crystalline rock sites. Standard deviations associated with the natural logarithm of peak horizontal acceleration and velocity are estimated to be 0.5 and 0.7, respectively, representing an 84th percentile estimate (or equivalently an upper 68 per cent confidence limit) that is approximately 65 per cent higher than the median for acceleration and 100 per cent higher than the median for velocity. While these strong-motion values are generally higher than those currently used in the seismic design of most structures in the Eastern United States, their impact on the seismic safety of specific sites in the region will depend on an assessment of the probability of occurrence of an earthquake of this size within the near-source region of these sites.