Such data are being used to define the kinematics of the boundary zones, and (in conjunction with other geological and geophysical data) provide constraints that can be used to develop and test models of their mechanics (17). For example, GPS data show the full variation of the motion across the Andes, from the interior of the oceanic Nazca plate to the interior of stable South American continent, and provide a detailed look at the process of ocean-continent convergence and continental mountain building (18). Site positions and velocities are available on the WWW. GPS programs are providing similar data at other ocean-continent and ocean-ocean convergent boundaries, and so are significantly improving our knowledge of the complex processes there (7,9,19).

Figures 5 & 6 - GPS data and an interpretation across the Nazca-South America plate boundary zone in Peru and Bolivia. About 30-40 mm/yr of slip, roughly half of the overall convergence rate, is accumulating on the locked plate boundary thrust fault and should be released in future great earthquakes. This estimate avoids some of the difficulties inherent in previous aseismic slip estimates based on the earthquake history. About 10-15 mm/yr of crustal shortening occurs inland at the sub-Andean foreland fold and thrust belt, indicating that the Andes are continuing to build. This shortening rate is significantly greater than inferred from seismic moments, suggesting that the shortening is largely aseismic. Little along-trench motion of coastal forearc slivers is observed, despite the oblique convergence geometry (18). Pink dots show shallow seismicity (<60 km depth).