Effects of high-impedance-contrast boundary upon multi-modal seismic surface wave data

Synthetic studies and analyses of an experimental dataset were conducted to address the use of Rayleigh-type surface waves for estimation of shear wave velocity (VS) profiles of shallow bedrock sites. The shallow bedrock presents a high impedance contrast boundary which causes surface wave energy to be partitioned to higher modes. Idealized studies of hypothetical datasets and root-mean-squared calculations of error surfaces showed that if reliable dispersion data are available over a broad frequency spectrum, the VS profile can be recovered using the fundamental mode alone. However, when dispersion data are limited to a relatively narrow frequency band representing what might be commonly sampled in practice, the fundamental mode alone is insufficient to resolve the VS profile, but joint consideration of higher modes along with the fundamental mode will be successful in some cases. Outcomes of similar tests on experimental data from a shallow bedrock site were instructive but less definitive. Tests repeatedly demonstrated that a good quality starting model is necessary to converge upon a reasonable solution. To improve the starting model, the dispersion dataset can be scrutinized to estimate depth to bedrock. Dispersion curves for the shallow bedrock case are particularly susceptible to misinterpretation. Tests revealed that the consequences of a likely form of misinterpretation are strong velocity fluctuations and overestimation of VS of the bedrock.