GNGTS 2015 - Atti del 34° Convegno Nazionale

has been used to constrain the inversion process. At each site, the inversion procedure has been repeated several times (>10) in order to estimate possible uncertainties affecting the resulting profiles. The overall variability for the S-wave velocity profile of each site has been assessed by using the extreme values of the set of solutions (profiles obtained by the inversion procedure) whose misfit is not greater than two times the misfit value associated to the best fitting model. These variability ranges have then been used in the deconvolution procedure ( vide infra ). In the sites where placing the seismic array in the same position of the accelerometric station was not possible, obtained best profiles have been manually adapted in order they were able to reproduce the HVSR curves obtained in correspondence of the stations. As these adapted profiles have always shown to be included in the correspondent variability ranges given by the inversion procedures, these variability ranges have been reputed valid for the relative accelerometric stations too. In any case, the best profile of each site has been used to estimate the V S,30 and therefore to reclassify the site. A summary of obtained results is shown in Tab. 1. Deconvolution analysis. When the stations are located, according with the realized reclassification, on reference sites, the ground-motion recorded by them can be directly used for validating seismic hazard estimates. For the other stations, recorded ground-motion has been deconvolved by considering the seismic response of shallow low-velocity layers from the surface to the seismic bedrock where the ground motion has to be assessed. The deconvolution has been performed, in a 1D scheme (in which the shallow subsoil is constituted by a stack of horizontal homogeneous and isotropic viscoelastic layers), by means the free software STRATA (https://nees.org/resources/strata ), doing both linear and equivalent linear analysis. STRATA requests the definition of a stratigraphy where each stratum is characterized by: thickness, some statistical properties of S-wave velocity (extreme values, means and variance), the kind of the material and its viscosity properties. Results exposed in previous section have been used to fix these parameters. First of all, a seismo-stratigraphical profile has been built by correlating the geological section of the area around each station with the V S profiles derived from the relative inversion process. This step has been useful in order to link each geological unit to a density value as well as to a damping and a bulk modulus reduction curves. Density values and dynamic curves have been chosen from data derived from seismic microzoning studies in similar lithological units, Tab. 1 - Synthesis of the results of the dynamical characterization of the studied acceleorometric stations. station municipality V S,30 new old seismic bedrock average V S to f0 code (m/s) soil soil depth the seismic (Hz) class class min – max (m) bedrock (m/s) CGL Cagli (PU) > 800 A B 0 - 20.2 CSC Cascia (PG) 698 B B 8 - 13 530 3.7 CSD Castel Viscardo (TR) 484 B B 22 - 189 454 3.9 GLD Gildone (CB) 470 B B 18 - 74 423 2.8 MTL Matelica (MC) 579 B C 33 - 82 578 2.0 PGL Peglio (PU) 358 C B 23 - 68 365 2.9 RNC Rincine (FI) 871 A B 9 - 12 570 no SNG Senigallia (AN) 258 C B 65 - 159 420 1.4 SRL Sirolo (AN) 270 C B 54 - 296 515 5.8 150 GNGTS 2015 S essione 2.2