Long-period amplification in deep alluvial basins and consequences for site-specific probabilistic seismic-hazard analysis: An example from the Po Plain (northern Italy)

This study deals with the impact of deep-soil discontinuities on groundmotion amplification in the Po alluvial basin (northern Italy), and consequently, on long-period probabilistic seismic-hazard analysis estimates. The focus is on Castelleone, where a seismic station of the Italian National Seismic Network has been deployed since 2009. To define a detailed shear-wave velocity (VS) profile, extensive active and passive geophysical surveys were carried out. The active measurements included seismic profiles examined through standard refraction techniques and multichannel analysis of surface waves. The passive methods included ambient vibrationmeasurements in both single and array configurations. In particular, three microtremor arrays with increasing apertures were performed to capture the experimental dispersion curve down to 0.2 Hz. The horizontal- to-vertical spectral ratio (HVSR) showed two main peaks, at 0.17 and 0.70 Hz. A joint inversion of the experimental phase-velocity dispersion and HVSR curves was performed to obtain the VS profile. The data show two main discontinuities at ~160 and 1350 m in depth. According to the Italian and European seismic codes, the shallow VS discontinuity can be ascribed to the seismic bedrock (i.e., VS > 800 m=s), whereas the deeper one is observed at the transition between Pliocene-Quaternary deposits and Miocene marls. Preliminary site-specific seismic-hazard analysis in terms of displacement response spectra for periods up to 10 s shows that neglecting the effects of the deeper discontinuity implies underestimation in hazard estimates of up to about 49% for mean return period (MRP) of 475 years and about 57% for MRP of 2475 years, with possible consequences regarding the design of very tall buildings and large bridges.