Are initial phases of seismic swarms driven by a cascade of events or precursory slow slip?

Resolving fine temporal stressing rate changes can provide crucial information on the driving mechanisms leading to observed seismicity rate change and surface deformation signals and favours the distinguish of various earthquake nucleation hypotheses, e.g., the preslip model and the cascade model. The initial phase of seismic swarms might be an interesting candidate, because (1) the long-duration seismicity during seismic swarms provide us with better chances to reveal any stress change evolution, and (2) significant seismic slip is expected to occur caused by the most energetic event, which could bias our analysis of slow slip existence, while the initial phase contains much less seismic slip.

In this research, we revisit the 2011 Hawthorne seismic swarm (Nevada, USA) with well-recorded seismicity and abundant geodetic data, and test whether the derived observables can distinguish between two distinct slip nucleation hypotheses (cascade and preslip models). Firstly, to support the cascade model, we calculate the Coulomb stress change from the geodetic-estimated fault slip models, which allows us to analyse the spatio-temporal distribution of seismicity. Secondly, to test the preslip model, a modified rate-and-state model is proposed to connect the seismicity rate to the shear stressing rate, which is derived from a new slip history function - a logistic function. We apply this new method to the 2011 Hawthorne seismic swarm, and estimate the shear stressing rate history. The results show that: (1) A slow slip event is required to explain the observed deformation and seismicity in the initial phase of the swarm. Although the seismicity can be triggered by preceding nearby earthquakes, the cascade model alone cannot explain the observed surface deformation signals. (2) Slow slip is accelerating during the initial phase, and this pattern is consistent with the acceleration of slip during the nucleation of ruptures observed in laboratory experiments and numerical simulations. (3) The most energetic event (M4.6) could have been triggered by a slow slip event, nearby preceding seismicity, or both of them. 

The study of the initial phase during the 2011 Hawthorne seismic swarm allows us to explore the driving mechanism leading to the spatio-temporal evolution of seismicity. We conclude that the slow slip is required to interpret the surface deformation and recorded seismicity, and the triggering of the observed earthquakes in a cascade model cannot be ruled out. This study contributes to providing a new method to model the shear stressing history, which helps to illuminate the physics of the nucleation of earthquakes and the role of slow fault slip in the future.