Abstract
In this study, we propose an approach to generate spatially correlated seismic ground‐motion fields for loss assessment and risk analysis. Differently from the majority of spatial correlation models, usually calibrated on within‐earthquake residuals, we use the sum of the source‐, site‐, and path‐systematic effects (namely corrective terms) of the ground‐motion model (GMM), obtained relaxing the ergodic assumption. In this way, we build a scenario‐related spatial correlation model of the corrective terms by which adjusting the median predictions of ground motion and the associated variability. We show a case study focused on the Po Plain area in northern Italy, presenting a series of peculiar features (i.e., availability of a dense dataset of seismic records with uniform soil classification and very large plain with variable thickness of the sedimentary cover) that make its study particularly suitable for the purpose of developing and validating the proposed approach.
The study exploits the repeatable corrective terms, estimated by Lanzano et al. (2017) in northern Italy, using a local GMM (Lanzano et al., 2016), which predicts the geometric mean of horizontal response spectral accelerations in the 0.01–4 s period range. Our results show that the implementation of a spatially correlated model of the systematic terms provides reliable shaking fields at various periods and spatial patterns compliant with the deepest geomorphology of the area, which is an aspect not accounted by the GMM model. The possibility to define a priori fields of systematic effects depending on local characteristics could be usefully adopted either to simulate future ground‐motion scenarios or to reconstruct past events.
Fluid Flow and Heat Transport Computation for Power-Law Scaling Poroperm Media