Broadband Near-Fault Ground-Motion Simulations Using a Modified <i>f</i> - <i>k</i> Method in Central Italy

ABSTRACT In this study, we simulate ground motions for three major earthquakes that occurred in Central Italy: the 2009 Mw 6.3 L’Aquila earthquake and the two largest events of the 2016 Amatrice–Visso–Norcia sequence, the Mw 6.2 Amatrice and Mw 6.5 Norcia earthquakes. These events caused widespread damage, intense ground shaking, and casualties, making them key cases for evaluating and improving ground-motion modeling techniques, especially given the high-quality recordings available from near-source stations. We use the frequency–wavenumber (f-k) method to generate broadband synthetic ground motions ≤7 Hz. This approach balances computational efficiency with physical accuracy, making it well suited for regional-scale applications. Green’s functions are computed from the Central Italian Apennines 1D layered velocity model (Herrmann et al., 2011), and kinematic rupture models are generated using the methodology of Graves and Pitarka (2016; hereafter, GP16), implemented into the f-k framework. Site response is incorporated using frequency-dependent nonlinear amplification factors (Borcherdt, 1994). We evaluate the simulations through a detailed comparison with observed ground motions using RotD50 spectral acceleration and a goodness-of-fit analysis. In addition to simulations at recorded stations, we compute ground motions over a dense grid of virtual sites to assess spatial variability and source-related effects such as rupture directivity and regional amplification. Results are benchmarked against predictions from the empirical ground-motion model ITA18 (Lanzano et al., 2019). Our findings demonstrate that the f-k method can reproduce key features of observed ground motion with good agreement, capturing both near-fault effects and broader regional trends. This study demonstrates that the f-k approach offers a computationally efficient and physically robust alternative for deterministic ground-motion simulation, bridging empirical models and 3D simulations for practical use in hazard scenarios and urgent computing workflows.