Multivariate statistical analysis of site effect indicators for the Montereale and Capitignano area (AQ) following the seismic sequence in central Italy (2016-2017)

This study investigates site effects for developing predictive models of site response, for both research purposes and future geognostic applications, by using a statistical analysis method to define possible correlations between different kinds of seismological and geological data. The test area is the intermontane Montereale basin (Abruzzo region, Italy), which was investigated for site effects studies and microzonation purposes because it has been affected by the 2016-2017 Central Italy seismic sequence. The available geological and geophysical information include geological profiles, thickness and geometry of the sedimentary infilling, shear-wave velocity profiles, downhole and borehole data. Seismological analysis (such as earthquake and noise spectral ratios, duration lengthening of seismic signals) has been performed to retrieve parameters useful for seismic response. A GIS database has been created to better manage all this information, by georeferencing all the measurement points and results. Finally, Factor Analysis multivariate technique was used to evaluate the most important site effect indicators, in terms of correlation between the involved parameters, and to deduce how the geological context influences their behavior. The performed analysis highlighted the statistical variability of the strong motion, that depends on the position inside the basin, over the Holocene deposits, and on the edges, at the limit between different lithologies. Moreover, some of the correlations between geological and geophysical indicators are strongly coherent with the basin geological features.