How does 'interseismic coupling' relate to Megathrust earthquakes?

A number of large earthquakes have recently occurred at subduction zones monitored from geodetic networks. The geodetic data combined with seismological data, allow determining the spatio-temporal evolution of fault slip during the phase of seismic wave radiation as well as during aseismic slip preceding or following these seismic ruptures. Such studies place constraints on the spatial distribution and long term budget of seismic and aseismic slip, which determine the location and frequency of large earthquakes. They can also help assess the physical properties (e.g., lithology, temperature, fluid pressure) that determine the mode of fault slip. In this presentation we will review the results obtained from studying diverse recent examples, including the Tohoku Oki earthquake, and discuss key general questions: is interseismic coupling stationary? how does it relate to the location and frequency of large megathrust earthquakes? are seismic ruptures the result of a deterministic or fundamentally random process? While addressing these questions from a purely observational approach is extremely challenging, resorting to numerical simulations of the 'seismic cycle', calibrated through comparison with observations, offers the best prospect for progress.