Temporal evolution of the stress state in a supercontinent during mantle reorganization

Abstract

SUMMARY Previous numerical simulation models of isoviscous mantle convection in both 2-D and 3-D geometries have revealed that the presence of a supercontinental lid produces large-scale horizontal mantle flow, thereby reorganizing the thermal structure of the mantle interior. Large-scale upwelling plumes arising from the core–mantle boundary (CMB) beneath the supercontinent are observed in some models. These upwelling plumes have been hypothesized to produce tensional stresses in the supercontinent and to be responsible for the subsequent continental lifting and breakup. In this study, numerical simulations using a 3-D spherical shell model are conducted in order to elucidate temporal changes in stress state in an idealized supercontinent during mantle reorganization. The model supercontinent is assumed to be an undeformable, highly viscous supercontinental lid (HVSL) with respect to the mantle. The mantle is heated from the base and from within. The HVSL is imposed abruptly on well-developed convection systems and remains fixed during the simulations. I first consider the scenario in which the HVSL is imposed on isoviscous or weakly temperature-dependent viscosity convection exhibiting short-wavelength thermal structures. The simulation results reveal that a tensional stress regime rapidly predominates throughout the HVSL (within 50–100 Myr in model time) in response to the horizontal mantle flow beneath the HVSL. The subsequent large-scale plumes upwelling from the CMB also induce tensional stresses in the supercontinent, but the effects of these stresses on the overall stress state appear to be secondary. For most of the models investigated herein, the maximum deviatoric tensional stress generated in the HVSL once convection is re-established is 30–90 MPa, which is probably comparable to the minimum stress magnitude required to break up the supercontinent. To investigate the effects of the HVSL on other convection regimes, I next consider the scenario in which an HVSL having a viscosity contrast of 10 is imposed on much longer-wavelength (i.e. degree-one- or degree-two-dominant) convection systems with moderately temperaturedependent viscosity. The patterns of such low-degree-dominant convection are affected little by the imposition of the HVSL, and notable mantle reorganization does not occur. On the other hand, when a reasonable yielding stress is imposed at the continental margins, newly formed, weak upwelling plumes develop due to enhanced downwelling at the margins, rather than large-scale mantle reorganization. However, the newly upwelling plumes do not seem to produce a visible tensional stress regime in the HVSL interior, and the spatial distributions of the stress regime and stress magnitude in the HVSL depend on the spatial relationship between the HVSL and the convection pattern. Since the Series L models of the present study do not include the yield stress in the oceanic part of the lithosphere, plate-like behaviour and subducting plates are not actualized. The results for Series S models, in which the oceanic lithosphere freely sinks and the convection pattern is easily organized, imply that the broadly tensional regime prevailing in the supercontinent may support the penetration of mantle plumes into the supercontinent, the emergence of hotspots and subsequent continental uplift and break up.