Hybrid Ground-Motion Model for Potential Induced Seismicity at the Prinos CO2 Storage Site (Northern Greece)

ABSTRACT Reports of seismicity linked to human activities have become increasingly common over the past few decades, yet predicting the level of ground motion generated by induced earthquakes remains challenging due to limited observational data and the variable structure of the upper crust. Reliable ground-motion models (GMMs) that reflect region-specific seismological conditions are therefore essential for the proper assessment of induced seismic hazard. Existing GMMs, derived primarily from tectonic events, are not necessarily applicable to shallow, low-to-moderate magnitude-induced earthquakes, for which source, path, and site effects may differ substantially. In this study, we develop the first region-specific hybrid GMM specifically designed for potential injection-induced seismicity at the Prinos CO2 storage site in northern Greece. The modeling framework combines stochastic simulations calibrated against available regional recordings and constrained by an a priori seismological model. GMM parameters are then optimized through iterative calibration and epistemic parametric uncertainty is quantified by retaining suites of acceptable parameter combinations. A synthetic dataset combining region-specific empirical recordings and simulated events covering magnitudes 2≤Mw≤6 and hypocentral distances up to 80 km allows comparison with existing GMMs for induced seismicity. The final hybrid model integrates simulated motions with an expanded empirical dataset that includes machine learning-derived seismic events. The resulting GMM provides a regionally calibrated framework for ground-motion modeling and hazard assessment at the Prinos storage site and contributes to ongoing efforts to develop methodologies for quantifying the ground-motion characteristics of induced earthquakes in Europe.