Effects of rheology and mantle temperature structure on edge-driven convection: Implications for partial melting and dynamic topography

Abstract

Edge-driven convection, which affects partial melting, intraplate volcanism, and dynamic topography, is small-scale convection that occurs along a lithospheric keel with a sharp contrast in lithospheric thickness. Various factors, including Rayleigh number, lateral mantle temperature heterogeneity, and geometry of the keel, influence the edge-driven convection, and the correlation between edge-driven convection and surface expressions (dynamic topography and volcanism) is complicated. We performed a finite element study to quantify the effects of these factors on dynamic topography and partial melting. We found that the dynamic topography is more prominent when a strong edge-driven convection cell develops, which corresponds to homogeneous mantle temperatures and the absence of mantle wind. In contrast, the development of edge-driven convection cells and dynamic topography near the lithospheric keel are hindered when the mantle temperature is strongly heterogeneous (laterally varying ~280 K). This indicates that a large lateral contrast in mantle temperature results in a strong mantle wind that may prevent the development of edge-driven convection cells. An increase in the Rayleigh number results in more vigorous convection and enhances partial melting. Our study shows that the location of volcanic activity at craton edges and passive margins can be reproduced in models with weakly heterogeneous mantle temperature for given mantle viscosity. The existence of a strong mantle wind (e.g., related to subducting slabs or mantle plumes) may inhibit the formation of an edge-driven convection cell and its related partial melt near a lithospheric keel. However, mantle conditions with weak temperature heterogeneity (<~14 K) or high mantle viscosity (>17×1019 Pa∙s), which corresponds to the Rayleigh number of 1.8×106, do not induce partial melts despite the development of edge-driven convection cells. Our model parametrized the condition and location of edge-driven convection cells and partial melts, which can contribute to understanding anomalous intraplate volcanisms, such as in Jeju Island south of the Korean Peninsula and the Tanzania Craton near the East African Rift.