A machine learning-based V-in-olivine oxybarometer for characterizing oxygen fugacity in lunar and terrestrial basalts

Abstract

Oxygen fugacity (fO₂) is essential for understanding the internal structure, evolution, and habitability of terrestrial bodies. However, the mechanisms controlling fO₂ in low-magnesium lunar and terrestrial basalts remain debated due to limitations of existing oxybarometers. Here, we develop a high-precision V-in-olivine oxybarometer using experimental data and machine learning, applicable to both high- and low-magnesium basaltic magmas in the Earth-Moon system. Through this approach, we find that mid-ocean ridge basalts exhibit lower fO₂ than empirical oxybarometers suggest, not supporting the assumption that the asthenospheric mantle is more oxidized than the lithospheric mantle. In arc magmas, fO₂ ranges from ΔFMQ + 0.5 to ΔFMQ + 2.5 and correlates with fluid proxies but not Mg# [molar Mg/(Mg+Fe)], indicating fluid flux as the primary control rather than magma evolution. Lunar basalts show similar fO₂ in high-magnesium types to previous findings but higher fO₂ in low-magnesium varieties, ranging from ΔFMQ - 5.4 to ΔFMQ - 1.9 and correlating with Mg# but age-independent, suggesting control by magma evolution. The differences in fO₂ control between the Earth and Moon during magma evolution arise because magma evolution can alter the ferric-to-total iron ratio, influencing fO₂, with the effect being pronounced under low fO₂ (on the Moon) but minimal under high fO₂ (on Earth). These findings refine the framework for the spatiotemporal evolution of fO₂ in the Earth-Moon system and provide new insights into studying the fO₂ of other celestial bodies.