Modeling lunar magmas in the Artemis Era

As the planetary science community sets its sights on the Moon, the existence of an open-source, up-to-date, and user-friendly modeling tool for lunar rocks is critical to maximizing the scientific return of ongoing and upcoming lunar missions (e.g., Artemis, PRISM, CLPS). While the creation of new code from the ground up is an important aspect of modern computational petrology, here we advocate for the modernization of legacy code, the results of which dominate the literature and shape our current understanding of geologic processes across multiple scales. Modernization of legacy code is critical as it enables the community to put new model results in the context of modern consensus gentium. Here we discuss how modern best practices for code development, publishing, and maintenance should be applied to upgrading legacy code, using lunar petrologic models as an example. We highlight critical gaps in our ability to model lunar processes that could be filled simply with updated modeling tools (i.e., where underlying experimental and analytical data already exist but are not incorporated into existing modeling tools).

Current modeling tools developed specifically for lunar compositions are sparse and can contain outdated parameterizations. One critical knowledge gap is our ability to model silicic lunar magmas, which are evidenced in nature by felsic fragments in returned Apollo samples and silica-rich volcanic domes identified on the borders of lunar mare by remote sensing. The most popular tool for modeling lunar magmas is MAGPOX, born from a series of FORTRAN scripts and ported to MATLAB, which is underpinned by an exclusively basaltic database. MAGPOX requires the crystallization of olivine on the liquidus, and thus has limited application to the full compositional diversity of lunar magmas. rhyolite-MELTS has been used to model silicic magmatism on the Moon, but its use typically requires additional experimental work given that the MELTS database is biased towards terrestrial rocks with lower FeO, higher alkalis, and higher fO₂ than lunar rocks. Notably, the MELTS database includes the same published basaltic lunar rocks used in MAGPOX regressions and so should be trustworthy for silica-poor lunar magmas. Still, the adoption of MELTS by the lunar community requires extensive testing against MAGPOX, PERPLE_X, and large lunar experimental databases.

Before we can even begin to update the parameterizations underpinning MAGPOX or efficiently compare MAGPOX to other models, a modern code library is required to perform adequate testing and benchmarking. In this talk we will explore the state of lunar petrologic modeling, what can be done now, and how we can best bring it into the 21^st century.