CAUSES OF LOW-FREQUENCY GROUND MOTION AMPLIFICATION IN THE SALT LAKE BASIN - THE CASE OF THE VERTICALLY INCIDENT P-WAVE

We use simulations of 1-D, 2-D and 3-D wave propagation to identify th e major causes of low-frequency (0.2-1.2 Hz) seismic amplification in the Salt Lake Basin. For a simple two-layer basin model and a vertical ly incident P wave, we examine how amplification is influenced by mode conversion, surface-wave generation, impedance effects at the sedimen t-bedrock boundary, resonance, and 2-D and 3-D focusing and scattering . Results show the following. (1) Approximately 30 per cent of the tot al cumulative kinetic energy at the Salt Lake Valley floor consists of shear-wave energy generated by P-to-S converted waves and surface wav es. The surface waves appear to be generated primarily along the edges of the basin, and the instantaneous S/P energy ratio in the sedimenta ry layer is as large as 3. (2) The largest peak particle velocity at t he free surface is due to the direct P wave. The value is roughly pred icted by the transmission coefficient of 1.46 at the sediment-bedrock interface, i.e. a normally incident P wave in the stiff bedrock will b e magnified in amplitude by 1.46 times as it enters the softer sedimen ts. (3) The low-frequency elastic response of the two-layer Salt Lake Basin model is characterized by surface-wave propagation and resonance from vertically interfering waves. (4) The peak particle velocities, cumulative kinetic energies, and mean spectral magnitudes computed fro m the 2-D (1-D) synthetics underestimate the values computed from the 3-D synthetics by up to 40 per cent (48 per cent) along a profile abov e the deepest part of the basin model. The 2-D and 1-D signal duration times underestimate the 3-D values by up to 59 and 94 per cent, respe ctively. Our results suggest that 2-D basin modelling may yield good a pproximations to the 3-D ground motion amplification above the deepest part of the Salt Lake Basin. Our results show that several mechanisms contribute significantly to low-frequency seismic amplification in th e semi-consolidated sediments of the Salt Lake Basin-P-to-S wave conve rsion, surface-wave generation, impedance effects at the sediment-bedr ock boundary, and resonance. Future attempts to estimate ground motion amplification in the Salt Lake Basin should therefore account for the amplification effects of all these mechanisms.