Thermochemical models of outer core convection with heterogeneous core-mantle boundary heat flux
Souvik Naskar, Jonathan E. Mound, Christopher J. Davies, Andrew T. Clarke
TL;DR
This paper demonstrates that thermochemical convection in a rotating spherical shell can generate both chemically and thermally stabilized regions beneath the CMB, including regional inversion lenses (RILs) when lateral CMB heat flux variations are imposed. Using 92 numerical models with $E=10^{-5}$, $ ilde{Ra}_T$ up to 4000 and $ ilde{Ra}_\xi$ up to 10^5, with $Pr_T=1$ and $Pr_\xi=10$, the authors show LEA below the CMB drives chemically stable regions, while heterogeneity in outer-boundary heat flux drives RILs that are typically ~100 km thick and stronger with greater thermal forcing. The key contributions are (i) establishing the coexistence and competition between LEA-driven chemical stability and thermally driven RILs, (ii) quantifying how RIL thickness and strength scale with $ ilde{Ra}_T$ and boundary heterogeneity $q^*$, and (iii) connecting these dynamics to seismic and geomagnetic observables, suggesting a geophysically plausible top-core stratification compatible with both observations and dynamo behavior. The findings imply a more nuanced upper-core structure than a global stable layer, with potential implications for interpreting seismic tomography and geomagnetic secular variation.
Abstract
Thermochemical convection in Earth's outer core is driven by the crystallisation of the inner core that releases latent heat and light elements. A key question in core dynamics is whether a stable layer exists just below the core-mantle boundary. Recent core convection simulations, accounting for CMB heterogeneities, propose locally stable regions (or regional inversion lenses, RILs) rather than a global layer, allowing both stable and unstable regions to coexist. In this study, we consider a suite of numerical simulations of thermal, chemical, and thermochemical convection models focussed on Ekman number ($E=10^{-5}$) with thermal and chemical flux Rayleigh numbers $\widetilde{Ra}_T=30-4000$ and $\widetilde{Ra}_C=30-100000$, and thermal and chemical Prandtl numbers $Pr_T=1$ and $Pr_ξ=10$. Analysis of purely chemical models reveals light element accumulation (LEA) below the CMB, resulting in either locally stable regions near the poles or global layers, depending on the strength of chemical forcing. These chemically stratified regions persist in our thermochemical models even if the thermal field is fully destabilising. The addition of a heterogeneous CMB heat flux leads to the formation of RILs driven by thermal stratification. Stable regions in these thermochemical models have varying locations, properties, and morphologies depending on whether thermal or chemical convection dominates. In the investigated parameter range, these RILs are O(100 km) thick, and their strength and thickness generally increase with the strength of thermal driving; they are comparatively less sensitive to the strength of chemical driving. Our simulations reveal a diverse range of possible stable regions and/or a global layer at the top of Earth's core, with a seismically plausible range of thickness and strength, which may also have a signature in geomagnetic observations.
