HOW TO BUILD A CONTINENTAL PLATEAU – INSIGHTS FROM ANDEAN MOUNTAIN BUILDING

Uplift of the Central Andes since the Eocene without collision is considered a formidable geodynamic paradox. Plateau uplift and the contribution of different processes are a matter of debate. Geophysical data across the Central Andes (ANCORP'96 and associated geophysical studies) indicate the widespread presence of partial melts at mid-crustal level under the plateau between the confining Cordilleras. Isotopic age-dating on syntectonic sediments of the Cenozoic intramontane basins building the Altiplano and seismic sequence-analysis demonstrate that the plateau crust was deformed during partitioning of deformation between various subunits. General acceleration of shortening rate shows no link to plate convergence rate. In contrast, our results show that the differential velocity between upper plate motion and oceanic plate slab-rollback velocity determined amount and rate of shortening as well as their lateral variability. This first order control is tuned by factors affecting the strength balance between the upper plate lithosphere and the plate interface of the Nazca and South American plates. These factors include a stage of reduced slab dip accelerating shortening (33 and 20 Ma) and an earlier phase of higher trench ward sediment flux reducing plate interface coupling with slowed shortening (45 and 33 Ma). Ultimately, the position of the Central Andes in the global southern hemisphere arid belt has a key role in allowing the rise and lateral spread of a high plateau as opposed to the northern and southern Andes. The combination of these parameters (in particular differential trench-upper plate velocity evolution, high plate interface coupling from low trench infill, and the lateral distribution of weak zones in the upper plate leading edge) was highly uncommon during the Phanerozoic leading to very few plateau-style orogens at convergent margins like the Cenozoic Central Andes in South America or the Laramide North American Cordillera.