Here is the abstract of the article that is published in AGU Monograph Plate Boundary Zones, eds. S. Stein & J.T. Freymueller, 2002.

The Global Moment Rate Distribution Within Plate Boundary Zones

Corné Kreemer, , William E. Holt , and A. John Haines

We present tectonic moment rates for most plate boundary zones inferred from a global strain rate model. The strain rate model reflects the accommodation of ongoing lithospheric motions within the seismogenic portion of the lithosphere, determined from geodetic velocities and Quaternary fault slip rates. We find a minimum compatible global tectonic moment rate within the plate boundary zones of 7.0 x 10^21 N m/yr, of which 63% is accommodated along subduction zones, 19% along oceanic ridges and transforms, 14% in continental areas, and 4% within zones of diffuse oceanic deformation. Due to the shortness of the earthquake catalog, no direct inferences on seismic coupling can be made from seismically released strain, except for ridge-transform systems where the seismic moment rate is not dominated by rare large events. The seismic coupling for ridge-transform systems with low spreading rates is 5-20%, whereas for fast spreading rates it is 1-3%. Unlike observed seismic moment rates, predicted moment rates, based on the seismicity rate of shallow events above a certain cut-off magnitude, can be meaningfully correlated with tectonic moment rates from the global strain rate model. We find that the correlation is particularly strong for subduction zones and to a lesser extent for continental areas. For ridge-transform systems we find that a spreading rate of about 55-60 mm/yr marks the transition from slow spreading ridges with relatively high seismicity rates to faster spreading rates with relatively low seismicity rates. The global correlation that we find between seismicity rates and tectonic moment rates for subduction zones and regions of (diffuse) continental deformation implies that seismicity rate can be used as a measure for tectonic activity. If the tectonic moment rate is known everywhere, regional deviations in seismicity rate from expected rates may reflect true variations in either the seismic coupling, maximum magnitude, or seismogenic thickness. We infer that the Chilean, Sunda, and Mariana-Izu-Bonin subduction zones are characterized by either low coupling or a relatively large maximum moment. A comparatively high seismicity rate along the New Hebrides may reflect a regional maximum magnitude that is smaller compared to other subduction zones.