Physics-Based Approach to Define Energy-Based Seismic Input: Application to Selected Sites in Central Italy

For relevant engineering purposes a viable reliable alternative to standard estimates of seismic hazard is represented by the use of scenario earthquakes, characterized not only in terms of magnitude, distance and faulting style, but also taking into account the complexity of the kinematic source rupturing process. Multi-scenario-based NDSHA (Neo-Deterministic Seismic Hazard Assessment) effectively accounts for the tensor nature of earthquake ground motions, formally described as the tensor product of the earthquake source functions and the Green’s functions of the transmitting anelastic medium, naturally supplying realistic time series, readily applicable to engineering analysis. Furthermore, it does not rely on scalar empirical ground motion attenuation models (GMPEs), as these are often both weakly constrained by available observations and fundamentally unable to account for the tensor nature of earthquake ground motions. This methodology has been successfully applied to many urban areas worldwide for the purpose of seismic microzoning, to strategic buildings, lifelines and cultural heritage sites. Some examples for selected sites in Central Italy, affected by the seismic sequence started in 2016, are here discussed, analyzing the energy content of the synthetic strong motion time histories, including the ones that predicted the spectral characteristics of the ground shaking due to the 24 August 2016 event. Specifically, the energy content obtained from (1) linearly scaled real records, (2) spectral matched records, (3) artificial accelerograms and (4) physics-based NDSHA ground motions, selected for a bridge hypothetically located at Amatrice, is compared. The comparison shows that the generation of artificial accelerograms (3) and partially also the spectral matching process (2) are characterized by an increase of the ground motion energy parameters with respect to the physics-based synthetic signals (4) and the linearly scaled records (1).