Underground and ridge site effects: A comparison of observation and theory

Abstract This paper compares experimental observations and theoretical predictions of the effects of underground and ridge sites on the amplitude of seismic waves. Observations by Tucker et al. (1984) of narrow-band, earthquake-independent site effects of a factor of 3 at underground sites down to a depth of 40 m can be predicted well using a one-dimensional theoretical model and information about seismic velocity as a function of depth. The observations can be explained as being due to the interference of upcoming, incident, and downgoing, surface-reflected waves within a thin, low-velocity, near-surface layer. An important effect of this layer is to make the interference pattern at shallow depths (i.e., within the layer) insensitive to the incidence angle of the incoming signal and, thus, stable from earthquake to earthquake. Our model suggests that at depths of about 100 m, site effects as great as a factor of 6 would be observed at about 5 Hz, while at much greater depths, site effects would be small. Currently available theoretical techniques are sufficient to assess the expected ground motion at underground sites provided that the velocity profile is known. This is useful for buried pipes, tunnels, and critical storage facilities. Observations by Tucker et al. (1984) and others of narrow-band, earthquake-independent site effects of about a factor of 8 on ridges cannot be predicted completely using theoretical techniques currently available to us. Accurate predictions could probably be made with a model that included the effects of ridge topography, neighboring topography, and near-surface velocity gradients. Use of such a model, if it were available, would require detailed, expensive knowledge of regional topography and velocity structure. The observed insensitivity of ridge site effects on ridges to incidence angle can, however, be explained with existing theoretical models as being due to a near-surface, low-velocity layer. The size of site effects on ridges and the stability of these effects from earthquake to earthquake suggests that they should be considered when interpreting seismograms obtained from ridge sites and when designing structures, such as communication antennae and power transmission towers, that will be located on mountain ridges. Accurate determination of the site effects of ridges and mountain tops can probably best be done using experimental techniques.