Thermo‐Mechanical Numerical Modeling of the South American Subduction Zone: A Multi‐Parametric Investigation

Abstract

AbstractThe South American subduction zone is enigmatic due to its fast subducting plate consumption, strong oscillation in subducting plate velocity, and deep mantle subduction. A key to help understand these subduction zone characteristics is through studying its dynamics with buoyancy‐driven numerical modeling that uses independent variables to best approximate the dynamics of the real subduction system. We conduct a parametric investigation on the effect of upper mantle rheology, subduction interface yield stress and slab thermal weakening. To constrain those model variables, we attempt to find best‐fits by comparing our model outcomes with the present‐day slab geometry and estimates of Cenozoic velocities in the center of the South American subduction zone. Key ingredients that need to be reproduced are low slab dip angles close to the surface, steeper lower mantle dip angles, strong oscillation of the fast Farallon‐Nazca subducting plate velocity and a progressive decrease in trench retreat rate during long‐term subduction. We include these ingredients to define a model fitting score using 10 criteria. Our best fitting models involve a weak subduction zone interface (yield stress of ∼20 MPa) and significant slab thermal weakening to attain the fast Farallon‐Nazca subducting plate velocity and to better reproduce the subduction partitioning since 48 Ma due to reduced shear stresses resisting downdip slab sinking and reduced slab bending resistance. Furthermore, a non‐Newtonian upper mantle promotes slab folding and realistic oscillation of the subducting plate velocity. Whether and how this slab folding process induces temporal variations in Andean deformation remains an open question.