Surface expressions of 3D lithosphere drips: Insights from numerical modelling

Lithosphere dripping is a mechanism in which a perturbation in the deep mantle lithosphere grows downward and sinks due to its negative buoyancy until it is completely removed. Such drips may be common because the lower mantle lithosphere is generally cooler and therefore denser than the underlying mantle. Hence, this process has been proposed to affect multiple regions including the western United States (e.g., Wallowa mountains, Oregon and Sierra Nevada, California), and central and northern South America (e.g., Puna plateau and Sierra Nevada de Santa Marta, in Colombia). Previous modelling studies have shown that lithospheric drips produce a set of characteristic surface expressions that can be used to identify them such as, localized symmetric topographic deflections, surface heat flow increases, and contraction/extension deformation patterns in the crust. However, most previous studies use 2D modelling geometries whereas lithospheric drips are expected to be cylindrical and should have radially symmetric surface expressions. In this study, we investigate the fundamental differences in the dynamics and surface observables of lithosphere drips in 2D and 3D. Models use the finite element code ASPECT to investigate drips triggered by an imposed sinusoidal perturbation in the continental mantle lithosphere. Initial models consider a three-layer structure consisting of a crust, mantle lithosphere and sub-lithospheric mantle, with constant viscosities and densities in each layer. In the reference model, the mantle lithosphere has a viscosity of 1.0x1021 Pa·s and is 200 kg/m3 denser than the sub-lithospheric mantle. Preliminary results show that the growth of 3D perturbations is significantly faster, reaching a depth of 660 km approximately 1 Ma earlier than a 2D drip. Also, 3D drips induce 1-2 km more subsidence and produce greater stresses in the continental crust, which reach maximum values of ±1.5 GPa during drip detachment. Future work will investigate more complex rheologies (lithospheric elasticity, plastic deformation, non-Newtonian viscosities) to quantify the differences between 2D and 3D drip models, and to provide insights into the implications of the use of 2D geometries to model 3D settings.