Nature of seismic coupling along simple plate boundaries of the subduction type

Abstract
The downdip width of the seismogenic zone is defined for 19 subduction zones. This width is measured from the base of the accretionary prism to the maximum depth of nucleation of thrust events along the plate boundary. Those two points are taken to define the upper and lower depth transitions from stable to unstable frictional sliding. The lower depth transition is found to be between 35 and 70 km. The dip angle of the thrust zone is also reevaluated. We find a linear increase in the dip angle as a function of depth, the slope of which varies between 0.2° and 0.6° km−1. The downdip width obtained, which is generally narrower than previously determined by most other authors, varies from about 50 to 150 km. We also determine the ratio of the rate of slip that occurs in earthquakes to the rate of relative plate motion. This ratio is defined as the seismic coupling coefficient (a). We obtain two different estimates of the seismic coupling coefficient: an average value from 90 years of seismicity and a value obtained using the slip‐predictable recurrence model for large earthquakes. We find a large variation in the computed values of a along and among subduction zones. For most of the subduction zones a is much less than 1.0; for several it is less than a few percent. Worldwide, we find no significant correlation between either the seismic coupling coefficient or the width of the seismogenic zone and subduction parameters such as the age of the oceanic lithosphere that is being subducted, plate convergence rates or absolute velocity of the upper plate in the hot spot reference frame. Such correlation exists only for a few individual subduction zones where other parameters do not vary as much. The observed variations in seismic coupling could be explained as differences in the frictional behavior of materials at the plate interface. Some of these differences may be attributed to the subduction of large bathymetric features, the roughness of topography, the presence of unstable triple junctions and active‐spreading ridges, and sediment composition.