A long-standing challenge in seismology has been reconciling two views of earthquake nucleation: one is based on rock strength, the other on fracture energy. This work shows how recent laboratory experiments and theoretical advances unify these perspectives into a single, scale-dependent framework. The central concept
is that the nucleation stress depends on asperity size relative to the fault’s characteristic dimension, which is typically the down-dip width 𝑊 of the seismogenic zone. Three different regimes emerge: in thin brittle crustal zones, such as those in some volcanic regions, rupture begins only when differential stress exceeds rock strength; for asperities of intermediate size, nucleation is governed by a combination of both strength and energy criteria; while large seismogenetic structures can fail under extremely low differential stress. By connecting nucleation physics to statistical seismology and tectonics, this framework is a powerful tool for interpreting seismic observations. Among its important predictions, it explains systematic variations in earthquake frequency–magnitude distributions in different tectonic settings and the occurrence of clusters of large earthquakes in stable continental regions. Specifically, in the brittle crust, ruptures can efficiently propagate and potentially cascade into large events, producing fewer small earthquakes relative to large ones (low 𝑏-value). In volcanic areas, strongly segmented fault networks or in the lower crust, the same physical processes that produce high stress drops also inhibit rupture growth, resulting in a high 𝑏-value. This framework provides testable predictions that can be evaluated through analysis of seismic catalogs, offering new constraints for physics-based seismic hazard models.
