Abstract

Bivergent wedges are common tectonic features but remain poorly understood. Much of our current understanding is derived from analogue models of wedges developed assuming specific conditions, and several aspects of these models, such as their coupling with exogenic processes, are complex. At a tectonic scale, wedge growth depends on the extraction and adhesion of material accreted into it. One of the main processes, tectonic erosion, is usually attributed to sources of roughness such as seamounts, ridges, horsts and grabens. However, subduction erosion also plays a key role in both the fracturing of the upper crust and the development of faults at high basal friction interfaces. In this sense, basal accretion of a bivergent wedge includes sediments and crustal rocks and is directly tied to subduction erosion and underplating. The material accreted at the base of the upper plate would cause uplift accommodated by normal faults and subsidence at the wedge tip. Based on the bivergent wedge model with high basal friction, sporadic events of erosion can result in stacking of upper plate slices that pile up at the S-point, which in turn define the pro- and retro-sides of the wedge at the surface. Recent investigations of the Chilean margin using local earthquake tomography have identified zones of seismic anisotropy within the continental lithosphere that can be associated with basally accreted upper plate slices. These anomalies are adjacent to zones with high Vp/Vs that can be due to a serpentinized lithospheric mantle. They are also located near the frictional limit of the subduction earthquakes. As Northern and Central Chile margin has been erosive since at least late Oligocene times, we propose a crustal extraction mechanism in which subduction erosion fractures the crust and creates splay faults that extract slices of the upper plate that originally moved downward with the slab as far as the S-point. These slices push the overlying serpentinized mantle up over the S-point. We also propose that the subduction wedge model should allow for crustal rocks to advance beyond the continental crust lower limit, coexisting with mantle material. As a result, the S-point would be located within the mantle lithospheric wedge and not be controlled by the Moho as has been proposed for other subduction zones.