THE LUNAR MAGMA OCEAN AS A PARADIGM FOR ALL TERRESTRIAL PLANETS?

The concept of a Moon-wide magma ocean was first proposed in the earliest days of the Apollo program and is the most widely accepted explanation for the evolution of the lunar crust and upper mantle. Magma ocean hypotheses have also gained appeal for explaining isotopic heterogeneities in the Earth and Mars because of the requirement of rapid differentiation of their silicate mantles in the first few million years after planetary accretion. Because the Moon is the ‘type locality' for the magma ocean hypothesis, it is important to consider the evidence for a magma ocean, and the isotopic variations in basaltic magmas generated from lunar magma ocean (LMO) cumulates. The LMO hypothesis has been built chiefly on the observation that the Moon's anorthositic upper crust has a large positive Eu anomaly, which is complemented by negative Eu anomalies in the mare-basalt source. This relationship is the foundation to notions that the crust may be the result of plagioclase floatation from a large magma chamber (LMO). Numerous models of the differentiation of the Moon involve an initial state in which the outer few hundred kilometers were totally molten. These models can adequately describe the anorthositic crust and the constant enriched incompatible-element signature of KREEP, the residual liquid from the LMO differentiation event. Likely source regions for mare basalt magmatism are in the uppermost few hundred kilometers of the Moon. Initial isotopic compositions (O, Sr, Nd, Hf, Os) of these source regions can be calculated from derived partial melts. Such isotopic information approximately reflects the anticipated variations generated from crystal-liquid partitioning during LMO differentiation. The observed isotopic and geochemical variations in mare basalts may reveal the degree of isotopic heterogeneity expected in the earliest history of more active and continuously mixing mantles of large planets such as Earth.