Testing whole-plate motion steadiness over the seismic cycle

Margins between tectonic plates host most large earthquakes recorded in the lithosphere. Over periods of tens to hundreds of years, relative plate motions along portions of crustal seismogenic faults promote the slow accrual of stress (i.e., the inter-seismic stress gain) that is later suddenly released via earthquakes (i.e., the co-seismic stress drop) – a process generally referred to as seismic cycle. Virtually all models of seismic hazard assessment assume that whole-plate motions (i.e., motions that are adequately described via Euler vectors) remain steady over the seismic cycle, and that the impact of inter- and co-seismic stress variations is solely crustal deformation in the vicinity of seismogenic faults. From the standpoint of plate dynamics, however, plate-margin stress variations during the seismic cycle generate torques that may be comparable in magnitude to those associated with viscous stresses at the lithosphere/asthenosphere interface, which resist plate motions. On this basis, it is plausible to hypothesize that whole-plate motions may be susceptible to temporal variations over the seismic cycle. The availability of progressively longer and denser GNSS position time series measured at sites located inside several tectonic plates indeed favor testing such a hypothesis. Here I will show results from recent studies that analyze publicly available GNSS data and infer temporal variations of the motions of several tectonic plates. These changes appear consistent with the torque variations associated with inter- or co-seismic phases of large earthquakes occurred along their margins. I will speculate on whether the link between whole-plate motions and the seismic cycle is robust enough to draw any additional information in the context of models of seismic hazard assessment.