GNGTS 2015 - Atti del 34° Convegno Nazionale

deconvolve the ground motion by considering the mono-dimensional seismo-stratigraphical model obtained by the passive seismic study. Ambient vibration monitoring and dynamic characterization of the sites. The ambient vibration monitoring has been carried out by performing, in each of the above-mentioned accelerometric sites, a number of single-station as well as a multiple-station (seismic array) measurements. In one site, two arrays have been realized, while in another one, an active MASW measurement has also been carried out ( vide infra ). As concerns single-station acquisitions, ambient vibrations have been recorded using a three-directional digital tromograph Tromino ® Micromed (http://www.tromino.eu/ ); for seismic arrays, vertical geophones (4.5 Hz) and a BrainSpy 16 channel digital acquisition system by Micromed have been used. Acquisition duration of single-station measurements was of 20 or 30 minutes, with a sampling frequency of 128 Hz, while passive seismic arrays recorded at 128, 256 or 512 Hz, in any case for 20 minutes. Each acquisition by single-station measurements has been processed in order to obtain the HVSR curve (Nakamura, 1989; SESAME, 2004), which allow to evaluate the possible presence of seismic resonance phenomena. Acquisitions by seismic arrays have been analysed by both ESAC (Ohori et al. , 2002; Okada, 2003) and f-k (Lacoss et al. , 1969; Capon, 1969) techniques, in order to obtain relative Rayleigh-wave dispersion curves. Jointly with each ambient-vibration monitoring, a geological/geomorphological survey has been carried out at a local scale (1:5000 or 1:10000), with the aim to better identify the geometry and to characterize the main lithological units where seismic impedance contrasts may occur (cf. Pileggi et al. , 2011). In the cases where the accelerometric station is located on outcropping stiff soil (Cagli, Cascia, Rincine), it has not been possible to deploy seismic arrays in exact correspondence of the stations, due to the rough topography: in these situations, geologic survey was particularly useful to identify alternative sites characterized by a strict geologic correspondence with the site of interest. Topographical conditions have caused a similar problem for the station of Peglio. As fully described in Pileggi et al. (2011), passive seismic surveys on stiff-rock sites suffer of significant drawbacks. In fact, in these geological settings, ambient vibrations are characterized by very low powers of surface waves (particularly evident in vertical ground motion) in the whole frequency range of interest (1-20 Hz). Moreover, when relatively high phase velocities exist (such as the ones expected at stiff-rock sites), ambient vibrations can be characterized by large wavelengths with respect to the overall dimension of the array (Foti et al. , 2011; Pileggi et al. , 2011). These effects may hamper retrieving clear dispersion curves in some of the analysed rock sites. At Cagli, a reasonable dispersion curve has been obtained for relatively high frequencies (10-25 Hz) only, while at Rincine it has not been possible to obtain any kind of dispersion curve, both using ESAC and f-k procedures. In this latter case, an active seismic prospecting has been necessary: by using the MASW technique (Park et al. , 1999), a dispersion curve that confirm the presence of Rayleigh-wave velocity values greater than 800 m/s at high frequencies (about 20 Hz) has been obtained. As concerns the seismic array analysis, dispersion curves obtained by ESAC and f-k technique are, in general, very similar. Just in one case (Sirolo) the curves produced by these two techniques have been merged to “build” a single curve, because the ESAC curve is acceptable for relatively low frequencies (3-6 Hz) while the f-k one is so for higher frequencies (6-14 Hz). In the remaining cases, the choice of the dispersion curve to be used in the joint inversion procedure ( vide infra ) has been done by considering the shape regularity and the frequency coverage of the obtained ones. To assess the V S profile for the sites whose Rayleigh-wave velocity values are lower than 800 m/s, a joint inversion of a site-representative HVSR curve and of the Rayleigh-wave dispersion curve has been performed in each of them, by using a Genetic Algorithm procedure ( e.g. , Albarello et al. , 2011). Since other geophysical measurements and borehole data are not available in the studied places, information provided by geological/geomorphological surveys GNGTS 2015 S essione 2.2 149