Simulating Fractional Melting of the Martian Mantle Using the pMELTS Algorithm

Abstract

Despite numerous investigations, the conditions and source materials that created the shergottite meteorites have not been fully constrained. Previous models commonly fit some parameters but struggle to fully explain the observed compositions and invoke processes of crustal assimilation without a full assessment. In this work, I apply the program pMELTS to investigate how variations in melting conditions propagate to magma chemistry. I use fractional melting to simulate melt production in upwelling mantle, and test whether changes in mantle temperature, lithospheric thickness, or water content create better fits to shergottites. For both changes in mantle temperature and lithospheric thickness, the observed melt compositions never approach the shergottites in parameters such as Mg #, CaO, and Al₂O₃ contents. However, elevated mantle water leads to increasing CaO/Al₂O₃ ratios, which approach the shergottite range. Calculated water contents >1.5 wt. % are required to fully match the shergottite range, which is higher than that proposed in previous works. Instead, I propose that a combination of partial source depletion and elevated water contents can fit the observed shergottite compositions with reasonable water contents (∼0.5 wt. % and lower). I also demonstrate that assimilation of a depleted harzburgitic mantle at crust-mantle boundary pressures (∼1 GPa) can elevate magmatic SiO₂ contents to levels seen in the shergottites. Melting of a variably hydrated partially depleted mantle, followed by reaction with a similarly depleted mantle at the crust-mantle boundary is thus demonstrated to be a mechanism that fully fits the chemistry of the shergottites.