GNGTS 2024 - Atti del 42° Convegno Nazionale

Session 2.2 GNGTS 2024 Fig. 3: Comparison of the results of HVIP and HVNR analysis of Daunia noise recordings: (a) peak frequencies and (b) peak amplitudes. Green dots mark the cases for which the main resonance frequency idenCfied by the two techniques differs by more than 1 Hz. The amplitudes of the HVIP peaks (which represent Rayleigh wave ellipCcity) are usually greater than those of the HVNR peaks. The main peak frequency idenCfied by the two techniques differs by more than 1 Hz in about 40% of sites. This seems related to the complex resonance pabern of flysch units, with mulCple peaks of similar amplitude. Future work The results of ambient noise analysis will be compared with the outcomes of numerical modeling of the local seismic response in order to seek possible correlaCons useful to esCmate the amplificaCon factors from ambient noise data. Such modelling will rely on the detailed auxiliary data collected for Seismic MicrozonaCon of the Daunia Mts urban and peri- urban areas, including borehole straCgraphies and results of geophysical surveys. A major contribuCon is expected from the study sites which, thanks to a large availability of data, will allow a 3D modelling of the local seismic response, to be compared with the results of simplified (1D and 2D) modelling; this will help evaluaCng uncertainCes introduced by the simplificaCon. The objecCve is to expand as much as possible the data set for the calibraCon and validaCon of empirical relaCon providing esCmates of amplificaCon factors in terms of Arias Intensity. Where the presence of site amplificaCon effects will be recognised through the noise measurements, the resistance demand values (A c ) x will be accordingly modified, introducing the esCmated amplificaCon factors in the calculaCon of exceedance probability of Arias Intensity. Finally, the (A c ) x values corrected for site amplificaCon effects will be compared with the slopes’ criCcal acceleraCon values a c based on local topography and geotechnical properCes of slope material. This should allow us to idenCfy slopes most suscepCble to co-seismic failure.