• 1D ground response analysis
• Po Plain
• PSHA
• seismic bedrock
• seismo-stratigraphic model
• soil amplification model

The characterization of seismic site response represents one of the most important issues of seismic hazard assessment and risk mitigation strategies (e.g., microzonation studies). This is particularly important for regions that are characterized by widespread sedimentary cover, with soft soil deposits that overlie rock or rock-like (i.e., stiff soil) formations. The Po Plain sedimentary basin in northern Italy represents a particular case study, which is investigated here.
In this thesis, a new seismo-stratigraphic model of the Po Plain sedimentary basin is provided together with a new long-period (> 1 s) site response modelling on the regional scale. Finally, probabilistic seismic hazard assessments (PSHA) are provided with the aim to evaluate the impact of the computed site amplifications on the Po Plain seismic hazard. This research includes two main innovative features: (1) the large-scale application of studies usually performed on the urban-scale; (2) the complex case study in terms of soil amplification, wave propagation and geophysical interpretation.
In the first part of this dissertation, the bedrock issue has been addressed. In earthquake engineering, the so-called ‘engineering bedrock’ is regarded as a stiff material (i.e., rock or rock-like geological formation) that is characterized by a shear-wave velocity greater than a target value (e.g., 800 m/s in the current Italian and European seismic codes). In the case of deep basins, the identification of the engineering bedrock is problematic, as it can lie well below the penetration depth of most common prospecting methods (i.e., a few tens of meters). Moreover, the depth of the engineering bedrock might not represent an effective proxy of the sedimentary thickness responsible for site amplification. Indeed, the deep seismic impedance contrasts may influence the ground motion amplification at the medium and long periods, causing the resonance of vertically travelling shear-waves at specific frequencies. The Po Plain sedimentary basin is one of the deepest and widest worldwide, and presents such problems. The aim of this work is to estimate the sedimentary thickness responsible for ground-motion amplification at medium and long periods. To this end, passive seismic prospecting methods based on ambient-vibration measurements using single-station and array configurations are considered to map the ‘seismic bedrock’ depth. This corresponds to a marked seismic impedance contrast where the shear-wave velocity approaches, or exceeds, 800 m/s. In the latter case, seismic and engineering bedrocks coincide.
The second part of this thesis regards the modelling of ground response at long periods, exploiting the seismic bedrock deep geometry retrieved in the first part. To address this task, a regional shear-wave velocity model of soft sediments above the seismic bedrock is provided through the interpolation of several S-wave velocity profiles obtained from 1-D joint inversion of H/V and array data. To compute the soil amplification functions, the velocity model is discretized into a grid. For each grid node, a 1-D soil model is defined, and a numerical ground response analysis is carried out. The soil model is verified at those sites with borehole seismic stations, where recordings of the same earthquake are available both at the surface and bedrock depth. In this second part of the thesis, a new methodology for the selection of ground motion time histories for numerical ground response analyses over vast areas is developed. The proposed procedure takes advantage of unsupervised machine learning techniques to define areas for which it is reasonable to assume the same sets of time histories based on the similarity of the seismic hazard.
The last part of this dissertation deals with the probabilistic seismic hazard assessment (PSHA) of the study area. The PSHA is carried out adopting the partially non-ergodic approach (e.g., Rodriguez-Marek et al., 2011; Kotha et al., 2017), and considering the updated assumptions about the seismogenic model and ground motion prediction equations. As one of the scopes of this study is to evaluate the impact of the computed site amplifications on the Po Plain seismic hazard rather than performing the best PSHA possible, only the uncertainty in the soil amplification is considered in the calculations.
The results of this thesis are compared to those provided by previous studies, where soil amplification factors are computed. Finally, the results of this thesis are also compared to those provided by the current Italian building code, so as to evaluate the impact of the different factors on the probabilistic seismic hazard assessment of the study area.