Multi-Timescale Fault Interactions and Earthquakes: A Review

Abstract Earthquake models are based on elastic rebound, which predicts cyclic stress accumulation and release on fault planes. Consequently, attention has been focused on the rate of strain accumulation on individual faults to infer their stress evolution, which, together with earthquake history, forms the basis for earthquake forecasting. Although much progress has been made, unanticipated devastating earthquakes have repeatedly reminded us of the need to improve current earthquake models. One such improvement may come from an improved understanding of fault interactions, which cause stress transfer between faults. This article synthesizes the author’s numerical studies of fault interactions across different timescales, alongside a review of recent progress in understanding these processes. Over geological timescales, old faults could be abandoned, and new faults develop to optimally accommodate relative plate motions or changes of the regional stress field. Current seismicity and geodetic strain rates represent snapshots of the ongoing evolution of fault systems; their spatial and temporal variations become clearer when viewed through the lens of long-term fault dynamics. Over multiple earthquake cycles, fault interaction causes static and dynamic stress changes and variable partitioning of tectonic loading. These stress perturbations cause earthquake clustering in time and roaming in space. Fault interaction also influences the initiation, propagation, and stopping of seismic ruptures, especially along stepovers and bends. This review presents examples of numerical modeling of fault interactions at different timescales. The nonlinear interactions between faults make them complex dynamic systems, where individual fault behavior is difficult to predict, but studying fault interactions can enhance our understanding of seismicity patterns of the fault systems.