GNGTS 2022 - Atti del 40° Convegno Nazionale

GNGTS 2022 Sessione 2.2 321 SELECTION OF THE SEISMIC INPUT FOR SITE RESPONSE NUMERICAL SIMULATIONS WITHIN THE SEISMIC MICROZONATION IN THE APULIAN REGION A. Romeo 1 , P. Pierri 1 , D. Casarano 2 , V. Del Gaudio 1 1 Dipartimento di Scienze della Terra e Geoambientali, Università degli Studi di Bari “Aldo Moro”, Bari, Italy 2 Consiglio Nazionale delle Ricerche – Istituto di Ricerca per la Protezione Idrogeologica - Bari, Italy Introduction. Seismic Microzonation (SM) studies are ongoing in the Apulian region, where seismic hazard depends on two types of active seismogenic structures: 1) autochthonous faults, mainly located in northern Apulia, capable of causing highly energetic earthquakes with rather long return times; 2) tectonic structures located in neighbouring regions, which can generate very strong earthquakes. This region presents a considerable heterogeneity of site conditions due to different geological-structural domains: orogenic belt (Daunia Mounts), foredeep (Tavoliere Plain) and foreland (Gargano, Murge). This geologic variability implies a heterogeneity of seismic scenarios potentially capable of affecting local seismic hazard. The site response properties can be obtained by numerical simulations that require the selection of waveforms to be used as seismic inputs. Accelerometric databases collect real waveforms for a variety of seismic scenarios, however often not representative of the studied area; so we combined real and synthetic accelerograms. In this paper we propose an approach to optimise the selection of the seismic inputs for seismic response numerical simulations. The results of the first tests of its application appear promising and the method will be implemented in the third level studies of SM planned for the Apulia region. Method. The proposed approach for the selection of the accelerograms to be used in numerical simulations of site response consists in the following stages: 1) determination of spectral acceleration values (SA) having, on soils of category A, an exceedance probability of 10 % in 50 years, which correspond to the ordinates of the elastic design spectrum for safety verification at the Ultimate Limit State; 2) definition of the parameters (earthquake magnitude, distance and style of fault) of seismic scenarios contributing to 2/3 of the total basic seismic hazard expressed by PGA values having an exceedance probability of 10 % in 50 years; 3) extraction, from accelerometric databases, of real accelerograms as compatible as possible with the seismic scenarios defined at the stage 2); 4) selection, among the accelerograms found at stage 3), of the 4 accelerograms best fitting, in the period range 0.05-2 s, the SA values determined at the stage 1); 5) calculation of the differences between the median values of the SA values of the accelerograms selected at the stage 4) and the design spectrum obtained at the stage 1); 6) definition of a target spectrum in the range 0.05-2 s for the generation of simulated accelerograms, by subtracting the differences obtained at the stage 5) from the design spectrum obtained at the stage 1); 7) generation of 20 simulated accelerograms for each scenario defined at stage 2) and selection of the 3 accelerograms, relative to different scenarios, best fitting the target spectrum calculated at the stage 6). The spectra to be determined at the stage 1) can be derived for any point of the Italian territory following the Italian Building Code (MIT, 2018 – hereinafter NTC18). In the implementation of the proposed method, the seismic scenarios to be defined at the stage 2) are identified from a disaggregated analysis of the seismic hazard (Bazzurro and Cornell, 1999). The results of this type of analysis for PGA values having 10 % exceedance probability in 50 years are provided by INGV for a grid of points spaced by about 5 km, covering