Toward Developing a Garnet Lherzolite Saturation Model for Lunar Low-ti, Ultramafic Green Glass Compositions

Introduction: The petrogenesis of the low-Ti lunar ultramafic glasses is critical to the understanding of volcanic processes on the Moon. Popular petrologic models developed to explain the formation of these glasses involve high pressure melting of a variety of sources, including primordial, garnet bearing, lunar mantle and depleted lunar magma ocean (LMO) cumu-lates [1,2,3,4]. Following the high pressure melting some of models involve continued, polybaric, melting and pooling of the melts at shallow depths [4,5] and others involve deep melts ascending to shallow depths where they interact with overturned LMO cumulates [1,2,3]. Members of this later class of models vary in their interpretation of the formation of the original, deep melts; from remelting of depleted magma ocean cumulates, to melting of primordial, garnet bearing, lunar mantle, previously unaffected by LMO processes. To fully evaluate these different scenarios, pet-rologic melting models must be developed for all possible lunar mantle materials, including garnet bearing lithologies. Motivation: Geochemical evidence indicates melts of a garnet-bearing, primordial lunar mantle have contributed to the formation of some lunar ul-tramafic glasses [3], however experimental studies of the lunar ultramafic green glasses indicate that the compositions are not saturated with garnet bearing lithology over expected pressure ranges [1,6,7]. These experimental studies indicate that ultramafic green glasses were last in equilibrium with an olivine + or-thopyroxene assemblage at ~2.4 to ~1.3 GPa [1,2,7]. This range in pressures has been explained by allowing deep, hot melts to assimilate LMO cumulates over a range of pressures, giving rise to the chemical trends found in the ultramafic glass compositions [8]. Because of the interaction with LMO cumulates, determining the origin of the original parental melts cannot be done using traditional inverse experimental petro-logic techniques. Estimating garnet saturated melts of a lunar mantle: A recent study of terrestrial mantle melting has produced a model to estimate garnet saturated per-idotite melts [9]. This model was used to investigate the possible relationship of melts in equilibrium with a garnet and the green glass compositions of Apollo 15 Group C [7]. The results of this study presented evidence that the predicted garnet saturated melt compositions could be related to the Apollo 15 green glasses of Group C, as defined by [8]. However, they also found