A journey through two-dimensional subduction modelling
Summary
In this research we aim to reconstruct a fully functioning numerical model that can represent the evolution of the South Sandwich Trench from its initiation 80 Myr to its present day position. To do so we constructed a two-dimensional numerical model in the finite element code ELEFANT. We constructed a visco-plastic subduction model constraining a
viscosity field based on the flow laws of the dominant deformation mechanisms in the lithosphere and mantle: diffusion creep, dislocation creep and Peierls creep. To represent brittle or plastic yielding we implemented Byrelees law. This research is divided into two sections. In section 1 an intra-oceanic subduction zone is constructed, tuning the viscosity parameters and implementing a plate velocity. Rollback and subduction was observed by the formation of a ridge in both the subducting plate (SP) and overriding plate (OP). In section 2 a setup similar to the South-Sandwich Trench was implemented. As the OP is continental, the plates had to decouple for subduction to evolve as no ridge could form in the OP. When decoupled, the SP could freely subduct and roll back. The limited decoupling between the SP and the surface however prevented the slab from rolling back further than maximal 250 km. Further subduction was only possible through the input of a plate velocity. However, together with the fact that the upper-lower mantle (ULM) boundary was deformable, the obstruction of trench motion caused cessation of the subduction and instabilities at the surface. To further confine this model and create a numerical model that can represent the evolution of the South-Sandwich trench, two main problems remain to be fixed: the SP should be able to decouple further at the surface
so the trench is free to move, and the ULM boundary should become underformable and semipermeable, as observed by geophysical observations.