Iron abundances on the lunar surface as measured by the Lunar Prospector gamma-ray and neutron spectrometers

Global measurements of iron abundances on the lunar surface are presented using data from the Lunar Prospector (LP) Gamma-Ray Spectrometer (GRS) and Neutron Spectrometer (NS). In this study, we derive relative iron abundances from the low-altitude, high spatial resolution (similar to(45 km)(2)) LP data using the 7.6 MeV neutron capture gamma-ray doublet. As part of the LP-GRS analysis, we demonstrate the importance of accounting for variations in neutron number density across the lunar surface by measuring neutron fluxes using LP-NS data. In a first step of comparing the LP-GRS data with previously published iron abundances inferred from Clementine Spectral Reflectance (CSR) data, we show that the existing CSR FeO data are nonlinear with respect to the LP relative iron abundances. We use the LP data to linearize the relationship between the CSR and the relative iron values then recalibrate the CSR data to iron abundance using returned soil abundances. We then correlate the CSR data, except for major anomalies, with the LP relative iron measurements to convert the LP data to absolute iron abundances. When we compare the LP-GRS and revised CSR data sets, we find a very good correspondence. There are two locations (Mare Tranquillitatis and South Pole-Aitken (SPA) basin) that show major discrepancies, suggesting that the CSR data are locally overestimating iron abundances. In both these regions, the discrepancies identified by the LP-GRS/CSR comparison are possibly explained by mineralogical differences that are not accounted for in the CSR to FeO calibration. In regards to our understanding of the Moon, the LP data have found the following: (1) There exist large expanses of mare basalt in the western mare regions that have very high iron abundances (22-23 wt.% FeO) that are underrepresented but not absent from the returned sample collection and are highly unusual for mare soils, which typically contain a significant amount of highlands contamination. (2) The low iron abundances in the lunar highlands (similar to5 FeO wt.%) are consistent with a previous analysis using thermal and epithermal neutrons and with the idea that the lunar crust formed by a relatively simple magma ocean process. (3) The comparison of LP and CSR derived iron abundances suggests that the material within SPA basin is similar to a norite-type rock without an enriched mantle FeO signature. (4) A comparison of LP and CSR data at Tycho Crater shows a large discrepancy such that the CSR data show moderate iron abundances of 89 wt.% FeO while the LP data show very low iron abundances of 3-4 wt.% FeO. This discrepancy cannot yet be easily explained by any known process.