Empirical Models for Effective Amplitude and Fourier Phase Spectra of Ground Motions Recorded in the Chilean Subduction Zone

Abstract

Chile, situated along the active Nazca–South America subduction zone, experiences frequent moderate-to-large earthquakes that pose a significant risk to its infrastructure and communities. Accurate characterization of ground motion is therefore essential for seismic hazard assessment and earthquake-resistant design. In this work, we develop empirical models for the effective amplitude spectra (EAS) and group delay time, which is a frequency-dependent, phase-related descriptor defined as the negative derivative of the Fourier phase spectra (FPS). Both models are developed using a comprehensive dataset of recordings from the Chilean subduction zone, encompassing events with magnitudes ranging from M 4.6 to 7 and distinguishing between interface and inslab sources. We introduce functional forms for both spectral models inspired by recent advances in ground-motion modeling and apply Bayesian regression to quantify the dependence of spectral amplitudes on magnitude, hypocentral distance, source type, and site conditions. In addition, the FPS is analyzed to identify systematic phase variations associated with group delay, providing insights into wave propagation characteristics. The resulting models provide a spectrum-based alternative to traditional ground-motion models, enabling realistic time-history simulations. These developments contribute to a more refined understanding of frequency-domain properties of subduction earthquakes in Chile, enabling the prediction of ground motion and assessment of structural resistance to seismic effects.