SITE AMPLIFICATION ESTIMATES IN THE GARIGLIANO VALLEY, CENTRAL ITALY,BASED ON DENSE ARRAY MEASUREMENTS OF AMBIENT NOISE

Citation
L. Malagnini et al., SITE AMPLIFICATION ESTIMATES IN THE GARIGLIANO VALLEY, CENTRAL ITALY,BASED ON DENSE ARRAY MEASUREMENTS OF AMBIENT NOISE, Bulletin of the Seismological Society of America, 83(6), 1993, pp. 1744-1755
Citations number
21
Categorie Soggetti
Geosciences, Interdisciplinary
ISSN journal
00371106
Volume
83
Issue
6
Year of publication
1993
Pages
1744 - 1755
Database
ISI
SICI code
0037-1106(1993)83:6<1744:SAEITG>2.0.ZU;2-F
Abstract
A frequency-domain formulation of the Aki (1957, 1965) autocorrelation method has been applied to seismic noise recorded by a 100-m wide cir cular array deployed on soft Holocene sediments in the Garigliano rive r valley, where a large amplification of ground motion during earthqua kes was experienced (Rovelli et al., 1988). The application of this me thod to ambient noise recordings demonstrates that microtremors in the valley are dispersive and dominated by surface waves. By assuming tha t the vertical component reflects Rayleigh wave motion, we obtain a di spersion curve that is interpreted in terms of a layered shear-wave ve locity structure. Layer thicknesses are constrained by the stratigraph ic information provided by a deep hole drilled in the area, and shear velocities are estimated by means of a trial-and-error approach to ach ieve a satisfactory fit of the ambient noise dispersion. The best-fit velocity model is used to compute a theoretical transfer function, whi ch is then compared with an average spectral ratio obtained from earth quake weak ground motions recorded at two stations, one in the valley and the other on a limestone reference site. An overall agreement is f ound between the theoretical curve and the observed spectral ratios. T he discrepancies that do exist may be ascribed to the assumption of 1- D inhomogeneity which considerably simplifies the theoretical transfer function. Our results show that the spatial-correlation method can be useful to infer velocity structure down to depths of hundreds of mete rs, when generalized geological informations are available, and can th us provide useful constraints for theoretical methodologies for the pr ediction of site response.