Modeling the Thermochemical Evolution of the Lunar Magma Ocean using Igneous Crystallization Programs

Since the properties of the lunar magma ocean (LMO), including its composition and dimensions, are largely unknown, the study of its thermochemical evolution requires igneous crystallization models capable of exploring a large range of possible LMO properties. However, the capability of such igneous crystallization programs to accurately model LMO solidification could only recently be sufficiently evaluated, since LMO crystallization has been simulated in experimental studies that explicitly consider the progressive changes of pressure, temperature and composition during magma ocean differentiation [1, 2, 3]. Using the results of these experiments, we tested the ability of the igneous crystallization programs FXMOTR [4] and alphaMELTS [5] to reproduce experimental mineralogies and crystallization sequences as well as the thermal and compositional evolution of the liquid phase. We found that neither program succeeded in reproducing the experimental results due to their specific limitations. However, using a combined model using FXMOTR for early and alphaMELTS for late crystallization stages, we can reproduce the crystallization sequence, the temperatures of phase saturation, the mineral modal abundances, as well as the temperature change with the degree of solidification with sufficient accuracy. This combined modeling approach can be applied to systematically study the effects of varying initial LMO composition and depth on the thermochemical evolution of the LMO, providing a base for subsequent modeling of lunar mantle evolution, including cumulate overturn, mantle melting and mare basalt formation. To constrain realistic magma ocean depths for different LMO compositions, we both match the Moon's moment of inertia considering the density profile of the cumulate and fit the thickness of the anorthositic crust. However, we note that the crust thickness can only provide a lower limit for the magma ocean depth, since the efficiency of plagioclase floatation has not been sufficiently quantified yet. [1] Rapp and Draper, Meteoritics & Planetary Science (2018). [2] Charlier et al., Geochimica et Cosmochimica Acta 234 (2018). [3] Lin, et al., Nature Geoscience 10.1 (2017). [4] Davenport Planet. Sci Res. Disc. Report 1 (2013). [5] Smith and Asimow, Geochemistry, Geophysics, Geosystems 6.2 (2005).