Characterizing uncertainty in shear wave velocity profiles from the Italian seismic microzonation database

Abstract. This research uses a large dataset from the Italian Seismic Microzonation Database, containing nearly 15,000 measured shear wave velocity (Vs) profiles across Italy, to investigate the uncertainties in seismic risk assessment. This extensive collection allows a detailed study of the seismic properties of soil with unparalleled precision. Our focus is on evaluating Vs variations with depth within uniformly clustered areas, known as seismic microzones. These zones are carefully identified based on their spatial correlation and homogeneity in geological, geophysical, and geotechnical characteristics, which are critical for accurate prediction of seismic response. We contrast these results with clusters formed purely based on geographic survey density (here defined geographic clusters), thereby assessing the depth of our understanding of the subsurface geological and geophysical context. These results were further compared with those reported in the seismic code and literature. This study of depth-dependent Vs variations helps to refine our models of subsurface seismic behaviour. Our main discoveries show that: 1) uncertainties associated with seismic microzones (geological and geophysical clusters) are consistently lower than those identified in geographic clusters, particularly in the first 30 m of depth; 2) Vs profile variations show negligible increases in uncertainty within a certain range of correlation distances (up to about 4,500 m); 3) uncertainties for seismic microzones are lower than those previously reported in seismic codes and in the literature, indicating the effectiveness and precision of our methodological approach. The results of this study significantly improve local seismic response analysis and highlight the critical role of depth and spatial correlation in understanding seismic hazard. The dataset is available at https://doi.org/10.5281/zenodo.10885590 (Mori et al., 2024).