GNGTS 2013 - Atti del 32° Convegno Nazionale

To test the new technique two sites were taken into consideration: one site (named CAR2) is located on the head of a now quiescent landslide, that in 1989 mobilised an about 40 m thick Quaternary colluvial deposit overlying Pliocene mudstone; the other site (CAR5) is located on the same kind of material, but about 150 m upslope of the landslide crown, on a weakly inclined area not involved in recent mass movements. The analysis of recordings of seismic events with different location and mechanism, compared with recordings at a nearby reference site, revealed that the site CAR2 is affected by an anisotropic amplification with a maximum approximately parallel to the maximum slope direction (about EW): along this direction amplification presents multiple spectral peaks, the major being observed at a frequency of about 2.5 Hz. On the contrary, CAR5, although affected by a comparable amplification factor, does not show evidence of site response directivity (Del Gaudio and Wasowski, 2011). Fig. 2 – 3D histograms relative to the site CAR2 showing, for each filtering central frequency, the percentage of data samples having polarisation azimuths within intervals of 10°. The colour scale represent the mean H/V ratio amplitude for data samples belonging to each frequency-azimuth bin: dark columns represent H/V ratios exceeding the colour scale maximum. The different histograms are relative to the entire data set (left), the Rayleigh type wave packets (centre) and the SH-type wave packets (right). Data relatives to measurements carried out in 2007 (top), 2011 (middle) and 2013 (bottom) are reported. 230 GNGTS 2013 S essione 2.2