The Influence of Fault Geometrical Complexity on Surface Rupture Length

Propagating earthquakes must overcome geometrical complexity on fault networks to grow into large, surface rupturing events. We map step‐overs, bends, gaps, splays, and strands of length scales ∼100–500 m from the surface ruptures of 31 strike‐slip earthquakes, recording whether ruptures propagated past the feature. We find that step‐overs and bends can arrest rupture and develop a statistical model for passing probability as a function of geometry for each group. Step‐overs wider than 1.2 km, single bends larger than 32°, and double bends larger than 38° are breached by rupture half of the time. ∼20% of the ruptures terminate on straight segments. We examine how the distribution of geometrical complexity influences surface rupture length, inferring an exponential relationship between rupture length and event probability. Our findings support that geometrical complexity helps limit the size of large events and provide insights into the competition between energy supply and dissipation during rupture propagation. Plain Language Summary Zones of geometrical complexity along faults can behave as barriers that halt propagating earthquakes. We map five types of geometrical complexities from historical surface rupture maps and regional fault maps: step‐overs, bends, gaps, splays, and strands at 1:50,000 scale, corresponding to features >100–500 m in length. This is a finer scale than previous studies, which focused on kilometer‐scale zones of geometrical complexity. We classify each mapped zone of geometrical complexity as breached (earthquake propagated past) or unbreached (earthquake halted) and measure the width of step‐overs and strands, the length of gaps, and the angle of splays and bends. Based on these measurements, we model the probability that each feature will be breached given its geometry. Step‐overs wider than 1.2 km, single bends larger than 32°, and double bends larger than 38° are breached by rupture half of the time. ∼20% of the ruptures terminate on straight segments. Using our probabilities, we show that the presence and geometry of features in the 100–500 m length scale plays a first‐order control on the low likelihood of large surface rupturing earthquakes. Key Points We map step‐overs, bends, gaps, splays, and strands from surface ruptures at 1:50,000 scale and assess their potential as rupture barriers Step‐overs wider than ∼1.2 km and bends >30° consistently halt propagating ruptures Our findings support that geometrical complexity on th