GNGTS 2023 - Atti del 41° Convegno Nazionale

Session 2.1 GNGTS 2023 mainly NNW-SSE to N-S oriented, that present a variable relationship with the underlying main thrust faults. The uppermost sedimentary strata within the anticlines are affected by numerous small-scale extensional faults, not rooted at depth, likely due to outer-arc extension above uplifted depocenters. In places, the inversion of basin-bounding faults is also visible. More regularly spaced and cylindrical NW-SE anticlines are also observed in the Gulf of Taranto, in the outer sector of the accretionary prism to the North, where a thrust/back-thrust tectonic style is evident. The origin of the anticlines varies within the overall set and reflects the complex long-term tectonic evolution of the accretionary prism, with the oblique docking of the Calabrian accretionary prism onto the Apulian Escarpment as a key feature. Defining and mapping the main tectonic structures allows the modeling of shaking scenarios related to the identified faults. The ground motion prediction is one of the main goals in seismic hazard assessment. Empirical ground motion prediction equations may fail to reproduce the complexity of ground shaking in complex 3D media and therefore the use of full waveform modelling is increasingly adopted to model ground shaking. The knowledge of the 3D crustal structure in terms of geometries of the main discontinuities and velocities is fundamental to model wave propagation. For these reasons, based on the collected extensive subsurface dataset, both onshore and offshore, a regional 3D crustal velocity model of this sector of the Calabrian accretionary prism was built. We integrated geological and velocity information into a regional 3D velocity and geological model at crustal scale, extended c.a. 120 km x 130 km x 60 km in depth. We implement our crustal model in the spectral-element code SPECFEM3D_Cartesian to simulate wave propagation in the 3D velocity model honoring surface topography. This allows reconstructing the low-frequency part of the waveforms (up to ~1 Hz), which is then combined with high-frequency seismograms obtained with a stochastic method following the hybrid broadband simulation approach by Graves and Pitarka (2010). We evaluate the goodness of our model by simulating real earthquakes and comparing simulated and recorded waveforms at the available seismic stations in the area. We also compare the results from our 3D model with the ones obtained using a local tomography model and the European crust model EPcrust (Molinari& Morelli, 2010). We then compute the shaking that would occur if some of the larger faults identified completely slip. The source of these scenarios is modelled as a bilateral finite fault with constant rupture velocity. The maps of ground motion obtained from the simulated broadband waveforms are then compared with empirical ShakeMaps, which have been calculated with a new VS30 map covering the offshore area. Faults, 3D model and scenarios have been used as input data for subsequent steps of the project, such as geomechanical modeling, as well as earthquake and tsunami impact scenarios modelling based on vulnerability evaluations and including the modeling of human and economic theoretical losses.