Self-similarity of continental-scale Twitter communication networks promotes synchronous outbreaks of information spreading

Understanding how information spreads across regions to trigger large-scale outbreaks remains a key challenge, particularly regarding the influence of regional network structures on phase transitions. In this study, we analyze global Twitter communication networks and reveal clear statistical self-similarity across continents. We show that such self-similar structures facilitate synchronous outbreaks characterized by a single transition, whereas double continuous transitions arise only under extremely sparse intercontinental connectivity. To reproduce these patterns, we introduce a coupled hyperbolic geometric model, S1–S1, that captures both the self-similarity and spreading dynamics observed in empirical data. We further demonstrate that breaking self-similarity in the degree distribution promotes the emergence of double transitions. An edge-based compartmental theory accurately predicts all simulation results. These findings suggest a unified framework linking network geometry, self-similarity, and the mechanisms underlying secondary outbreaks in real-world information diffusion.