Abstract. Estimating the recurrence intervals and magnitudes of earthquakes for a given fault is essential for seismic hazard assessment but often challenging due to the long recurrence times of large earthquakes. Fault network geometry (i.e. spatial arrangement between faults) plays a key role in modulating stress interactions and, consequently, earthquake recurrence and magnitude. Here, we investigate these effects of fault network geometry using earthquake cycle models to generate numerous earthquakes on two different networks of normal faults in Italy: the Central Apennines, characterised by a wide network of faults offset across strike, and the Southern Apennines, a narrow fault network where faults are predominantly arranged along strike. For each region, we ran an earthquake cycle simulation on systems of seven normal faults generating approximately 150 earthquakes. In the Central Apennines, co-seismic stress transfer between faults promotes more heterogeneous stress, more partial ruptures, greater Mw variability and less periodic behaviour of large earthquakes (coefficient of variation of recurrence time, CV 0.1–0.9). In contrast, faults in the Southern Apennines experience more homogeneous stress loading, leading to a higher proportion of full-fault ruptures with more regular recurrence intervals (CV 0–0.4). In both fault networks, high long-term slip rate amplifies the effects of fault interactions: faults with higher long-term slip rate are more sensitive to positive stress perturbations from nearby faults compared to slower-moving faults. These results highlight that incorporating stress interactions from fault network geometry into seismic hazard models is particularly important for networks of faults offset across strike, where rupture behaviour is more variable.
Piceda et al. (Wed,) studied this question.