Reliable information concerning the predominant site effects on ground motions can be obtained from low-cost shear-wave refraction surveys using a sledgehammer as an energy source. For the south and southeast Ontario region, the velocity structure can be determined to a depth of approximately 70 m. Near-surface shear-wave velocities of hard-rock sites range from 1.7 to 3.1 km/sec, with an average value of approximately 2.6 km/sec. Typical soil sites have shear-wave velocities of 250–700 m/sec near the surface.
Empirical methods of determining the relative values of frequency-dependent amplification are commonly employed. These methods fit the observed earthquake spectra with a regression model that decomposes the recorded spectrum into source, path, and site terms. By regression analysis of large amounts of recorded data, the average site term for a particular station can be determined. We calibrated such an empirical technique against the theoretical responses based on the velocity structure obtained from a detailed field survey. The empirical-regression approach and the theoretical-response approach provide reasonably consistent estimates of amplification at hard-rock sites. We conclude that the amplification for an average hard-rock site is about a factor of 1.3. At soil sites, empirical-analysis and theoretical-response results agree as to the frequency of the fundamental resonance peaks. The maximum theoretical amplification values generally exceed those indicated from the empirical analyses; theoretical amplifications by as much as factors of 6 and 7 were calculated, while empirical amplifications were below a factor of 3. Thus use of theoretical site responses may be conservative.