Anatomy of Synthetic Ground-Motion Variability in Central Italy: Insights from Kinematic Rupture Scenarios

ABSTRACT This study investigates how kinematic rupture parameterization controls ground-motion variability in simulations. We analyze a broadband synthetic dataset generated by Čejka et al. (2024) using the hybrid integral composite technique for an Mw 6.2 normal-faulting event in Central Italy. Rupture scenarios share the same fault geometry and seismic moment but vary in their kinematic source parameters. Ground motions are analyzed using a linear mixed-effects regression, and variability is investigated by decomposing residuals into between-scenario and within-scenario components. The between-scenario term is further resolved into contributions associated with the kinematic parameterization, whereas the within-scenario term is separated into a repeatable receiver-dependent component and a remaining residual, interpreted as the apparent aleatory part. Results show a pronounced frequency dependence. The between-scenario variability increases toward higher frequencies, indicating a stronger control of kinematic rupture parameterization. The within-scenario variability exhibits systematic spatial features, whereas the remaining residual component is largest at low frequency, where selected scenarios reveal coherent azimuthal patterns related to rupture directivity. Overall, the study transfers a residual-decomposition approach commonly applied to empirical data to a synthetic broadband dataset, providing a consistent framework for interpreting how kinematic rupture complexity shapes ground-motion variability and supporting future integration of simulations in ground-motion modeling.