Characterizing uncertainty and variability in shear wave velocity profiles from the Italian seismic microzonation studies

This study investigates the variability and uncertainty of shear wave velocity (Vs) with depth, focusing on the standard deviation of the natural logarithm of Vs (σlnVs) using a dataset of nearly 15,000 profiles from the Italian seismic microzonation studies. Seismic microzone clusters (SM), defined by geological and geophysical homogeneity, and geographical clusters (GC), based on survey density, were compared to evaluate their effectiveness in characterizing σlnVs variability. Spatial correlation analyses were performed to define high-quality SM clusters, ensuring strong internal geological and geophysical consistency with a maximum pairwise distance of 4.5 km between Vs profiles. Results demonstrate that SM clusters reduce σlnVs uncertainty by 14 % within the first 30 m, 9 % from 30 to 50 m, and 4 % from 50 to 80 m compared to GC clusters, highlighting the value of geological and geophysical refinement. These results can support a more accurate randomization of Vs profiles with depth in local seismic response analyses using 1D simulation codes, improving the reliability of site-specific seismic hazard assessments. The findings are validated against literature uncertainty thresholds, confirming the robustness of the SM approach. By analyzing 1120 SM clusters, this study offers a comprehensive framework for propagating uncertainties in seismic response simulations and surpasses the limitations of localized case studies. The large dataset of Vs profiles, associated with SM clusters, is publicly available at https://doi.org/10.5 281/zenodo.11263471 (Mori et al., 2024).