Long-term effect of a basal magma ocean on Martian mantle convection

Abstract

It has been proposed that a basal magma ocean (BMO) may have existed, or even still exists, at the base of the Martian mantle [1]. One formation scenario for such a BMO involves a mantle-scale overturn just after the crystallization of the main magma ocean. In this case, the BMO would be enriched in iron and heat-producing elements (HPE), and hence gravitationally stable at the base of the mantle, with potential effects on the efficiency of mantle convection. The Insight mission has allowed geophysical investigation of the Martian interior and has indeed provided seismic evidence of a basal liquid silicate layer just above the core-mantle boundary. It is thus crucial to understand the effect of such a layer on the long-term evolution of the interior of Mars.Here, we model thermochemical mantle convection and crust production for a Mars-sized planet in a 2D spherical annulus geometry using code StagYY.  Assuming that the top of the BMO is at ~1800 km radius, we parameterize the basal magma ocean as a ‘primordial layer’ with a low viscosity and a high effective thermal conductivity to account for the enhanced effective heat flux in a liquid layer due to turbulent flow. HPE are preferentially partitioned into the silicate liquid layer following a mass balance equation assuming an interstitial porosity.  We systematically vary BMO thickness and interstitial porosity in order to study the outcome of the different HPE distributions.  The liquid density, which is attributed by the different degrees of iron enrichment, is also examined to explore the mechanical stability and entrainment of the BMO.We present results of our models, comparing our present-day temperature profiles with areotherms deduced from seismic observation [2,3].  We find that the interstitial porosity is an important factor that determines the thermal structure of the mantle throughout Martian evolution. A value of ~50% provides the best fit with crustal production history, crustal thickness, HPE enrichment in the crust, as well as the seismically-constrained present-day areotherm. This result suggests that the initial HPE partitioning has not been controlled by end-member fractional crystallization of the main magma ocean (for which interstitial porosity would be close to 0%), and/or that some re-equilibration occurred during subsequent overturn. Meanwhile, the BMO thickness, within the uncertainties from seismic inversion, does not strongly influence Mars thermal evolution. [1] Samuel et al. (2021) doi:10.1029/2020JE006613[2] Khan et al. (2021) doi: 10.1126/science.abf2966[3] Duran et al. (2022) doi: 10.1016/j.pepi.2022.10685