The influence of second-scale convection on the thickness of continental lithosphere and crust

Abstract

The effect of sublithospheric convection on the thickness and structure of the continental lithosphere is studied in numerical models assuming different rheologies (Newtonian, non-Newtonian, and temperature and pressure dependent), heat fluxes, and heating modes (bottom versus internal heating). The stagnant regions near the top of the models are identified with the lithosphere. We distinguish a seismic and thermal lithosphere (controlled by temperature), an elastic lithosphere (controlled by viscosity), and a mechanical lithosphere (controlled by strain rate). Strong lateral thickness undulations with a thick lithosphere above the downwelling regions and a thin lithosphere above the upwelling regions develop. The thickness of the elastic lithosphere is about two-thirds of the mechanical lithosphere, while the ratio of the elastic to the thermal lithosphere varies between 0.4 and 0.8. The time dependence of some models (with internal heating and a surface heat flow of 20 mW/m 2) is characterized by long periods (of the order of 1 Ga) close to steady state and short periods (100 Ma) of changing cell patterns and lithospheric thicknesses. Models showing thick lithospheric roots suggest that the mantle beneath some old shields may be associated with cold, slowly downwelling convective flow rather than being firmly attached to the lithosphere. The missing gravity and topography signals from such regions can be explained by elastic bending of the lithosphere and by dynamic adjustment of the Moho. Observed lithospheric thickness variations in Europe and their minimum to maximum spacing agree well with the models, suggesting sublithospheric convection to be the cause. Compared to other observed features, thickness undulations of the lithosphere seem to provide the strongest indications for sublithospherical convection. Relatively low viscosities in the lower crust may lead to a long-term decoupling of surface topography from Moho topography and mantle dynamics. The Moho depth is capable of adjusting to mantle stresses within relaxation times of the order of 100 Ma, while surface topography may flatten out. Such a mechanism may be important in regions with a deep Moho such as Fennoscandia or with a shallow Moho such as the Pannonian Basin, neither region showing a pronounced surface topography.