We carried out gravity surveys along four routes across the Peruvian Andes, one in northern Peru, one in central Peru, and two in southern Peru. Two of the routes extend from the coast to the flat lands of the Amazon Basin across the high Andes, while the other two routes extend only from the coast to the central high plateau or the Altiplano. One route through Nazca contains 250 data points over a length of 800 km and thus offers one of the best gravity profiles across the Andes. This profile shows quite asymmetric gravity anomalies associated with the Western Cordillera and the Eastern Cordillera, in marked contrast with the symmetric pattern of topographic profile. Using these profiles, we constructed crustal models varying layer thicknesses rather than layer densities. The crustal thickness has a maximum beneath the Western Cordillera, although the maximum thickness varies from 45 km in northern Peru to 55 km in central Peru and 65 km in southern Peru in contrast to the assumed thickness of 35 km beneath the stable Brazilian shield.

Possible effects of lateral variations of crustal and mantle densities tend to reduce the above estimates of maximum thickness. The very shallow Moho beneath the Peruvian coast steeply deepens eastward in correspondence to the steep western slope of the Andes. The crust is thicker in the Western Cordillera than in the Eastern Cordillera. The Moho beneath the Eastern Cordillera tends to be relatively flat and shallows abruptly eastward across the sub-Andes. The Wadati-Benioff zone lies, in general, well below the model South American continental crust, leaving a mantle wedge in between. A notable exception is the Nazca profile, where the base of the model South American continental crust is in direct contact with the top of the Wadati-Benioff zone. This provides a direct evidence for uplifting of the buoyant slab with an aseismic ridge (the Nazca ridge) at its top, which tends to attach itself to the bottom of the continental crust. We examined whether or not the load due to mountain topography is balanced with the buoyancy load due to excess mass of the crust in the Andes. The Western Cordillera is found to be approximately in isostatic equilibrium, but the Eastern Cordillera is not. This contrast in mechanical state and the difference in recent geology suggest that the Western Cordillera and the Eastern Cordillera have not uplifted through entirely the same tectonic process. ¿ American Geophysical Union 1989