Observation of 1- to 5-second microtremors and their application to earthquake engineering. Part III. A two-dimensional study of site effects in the San Fernando Valley

Abstract

A two-dimensional study of the influence of deep sediment on seismic ground motions was conducted by recording and analyzing long-period microtremors in the San Fernando Valley, California. The recordings were made at 50 regularly distributed sites in the valley. Three reference sites were employed at the baserock outcrop around the valley so as to observe the time-dependent characteristics of the microtremors. Amplitudes in a period range of several seconds correlate with the thicknesses of the sedimentary layers. A site amplification effect, which was evaluated in terms of sediment-to-rock spectral ratios, is qualitatively consistent with available geological and strong motion data. The practical field observation procedure designed for this study, the two-dimensional approach, can be easily used in other surveys for estimating the spatial characteristics of earthquake ground motions in the period range of several seconds.

Site response on seismic motion in the Granada Basin (southern Spain) based on microtremor measurements

“…However, our results in the Granada basin are not an exception. Similar results as ours have been obtained in the Niigata plain in Japan (Kagami et al, 1982), San Fernando valley and Flushing Meadows in USA (Kagami et al, 1986;Field et al, 1990) and the Zafarraya basin in Spain (Morales et al, 1991a). This behavior in the non-variation of the predominant period could either be the result of a small impedance contrast between the basement and the sedimentary fill (Kagami et al, 1982) or else could be the result of a response dominated by a two-dimensional resonance model (Kagami et al, 1986;Field et al, 1990;Morales et al, 1991).…”

“…Similar results as ours have been obtained in the Niigata plain in Japan (Kagami et al, 1982), San Fernando valley and Flushing Meadows in USA (Kagami et al, 1986;Field et al, 1990) and the Zafarraya basin in Spain (Morales et al, 1991a). This behavior in the non-variation of the predominant period could either be the result of a small impedance contrast between the basement and the sedimentary fill (Kagami et al, 1982) or else could be the result of a response dominated by a two-dimensional resonance model (Kagami et al, 1986;Field et al, 1990;Morales et al, 1991). Field et al (1990) incline towards this last solution to explain the response in Flushing Meadows by considering the numerical solutions given by Bard and Bouchon (1985), who predict this type of behavior in the response of a sediment-filled basin in terms of its shape.…”

“…The results obtained in the Granada basin agree with those obtained by other authors in the study of site response in sedimentary basins. The variation of the spectral amplitudes for long period microtremors and their increase with thickness of sediments may give us information about site amplification of seismic ground motion in this range of periods resulting from extremely thick deposits (Kagami et al, 1982(Kagami et al, , 1986. We note that there is no variation of the predominant period in the Granada basin in the long period spectra (1 to 5 s), a product of the site transfer function under each point, unlike in Hachinohe and Aomori in Japan (Ohta et al, 1978) and Mexico valley (Lermo et al, 1988).…”

“…However, the analysis of microseisms can be a useful tool for the study of the influence of deep structure such as depth to basement, on ground motion if there are no strong changes in the meteorological conditions. Clear examples of these types of studies are those of Ohta et al (1978) and Kagami et al (1982Kagami et al ( , 1986.…”

“…Another procedure is to take the spectral ratio of microtremors motion at soil site with respect to the motion recorded at a hard-rock reference site in order to retrieve the site transfer function (Kagami et al, 1982(Kagami et al, , 1986Field et al, 1990). The lack of knowledge of the origin and the exact location of the sources that generate short period microtremors (Aki, 1988) and the variation in temporal and spatial spectral characteristics make interpretation difficult, thus it should be applied carefully, especially when the spectral ratio is used to avoid source and attenuation effects, since they may not be totally canceled.…”