Rethinking maximum magnitudes: the physics of supercritical ruptures in fault systems

Conventional earthquake hazard models assume that seismic ruptures are constrained by fault geometry and slip rates; anyway, catastrophic events like the 2011 Tōhoku earthquake came as surprises revealing fundamental weaknesses in our understanding of large magnitude seismicity. Here, I demonstrate that supercritical dynamics - a physical regime where stress cascades dynamically link fault failures—enables ruptures far exceeding traditional size limits. Analyzing 20 global regions using historical and instrumental catalogs, I show that this mechanism generates heavy-tailed magnitude distributions, increasing the likelihood of megaquakes by orders of magnitude. These findings reconcile discrepancies between geodetic and paleoseismic evidence, challenging the assumption that fault segmentation inherently limits earthquake size. I propose a unified framework integrating geodetic moment budgets, statistical physics, and paleoseismic extremes. Unlike tapered Gutenberg-Richter models, which underestimate tail risks, a bimodal distribution emerges: moderate events follow fault-system scaling, while extreme ruptures (e.g., Tōhoku 2011, Andaman-Sumatra 2004) arise from system-spanning cascades when faults reach critical stress synchronization. This transition is modeled as a stochastic escape process across stress-modulated potential barriers, explaining why megaquakes tend to cluster despite their rarity. The results imply that hazard assessments must combine physics-based supercriticality with paleoseismic and geodetic data to quantify extreme risks accurately. Bridging these investigations is essential for resilience against tiny-probability, catastrophic-impact events that will occur again sooner or later.