Lithosphere Removal in the Central Andes: Reconciling Seismic Images and Elevation History

Abstract:

Shortening of the upper crust should be accompanied by thickening of the deeper lithosphere. However, for many orogens, including the central Andes, geophysical and geological observations indicate that mantle lithosphere is spatially heterogeneous, being anomalously thin or absent. Seismic studies of the central Andes suggest that mantle lithosphere is locally removed. Also, present day elevations of the central Andean Plateau have been explained by rapid removal of mantle lithosphere over the last 10 Ma. Yet, the geological record is innately incomplete, and seismic tomography and receiver functions can offer only a present day snapshot of the subsurface. None of these techniques provides concrete insight into the physical processes responsible for current Andean elevations (3-5 km). A 2D plane-strain thermo-mechanical code, SOPALE, is used to examine the deep lithospheric dynamics connected to mantle lithosphere removal within a subduction zone setting, such as the Andes. Three models have been tested: removal by viscous dripping, by delamination, and a model with no removal. The removal models contain a high density eclogite root, creating a contrast between mantle lithosphere and mantle material. For the viscous drip models, mantle lithosphere is removed within 2.5-5 Myrs, descending subvertically through the mantle, causing subsequent surface rebound. Prior to this rebound, surface topography subsides locally over the dense root. This subsidence is influenced by crustal rheology, where weaker crustal rheologies produce deep, narrower basins (25-75 km wide, ~1 km deep), and stronger crustal rheologies produce shallow, broader basins (300-400 km wide, ~0.5 km deep). Delamination, which involves the coherent removal of mantle lithosphere along the Moho, affects a larger region, and is reflected in broader basins that extend into the back-arc. In all models, the deep lithosphere dynamics have an appreciable effect on surface topography, therefore, removal events should be recorded in the elevation history of an orogen. Past studies that have examined mantle lithosphere removal have assumed simple and idealized systems. We are currently working on more complex model configurations in order to reflect the detailed geometry and compositional and rheological heterogeneities within the central Andes.