Water in Lunar Pyroclastic Deposits : Linking Orbital Observations To

Introduction: Water (OH and/or H2O) on the lunar surface and preserved in lunar materials can record information about solar wind, cosmic ray, impact, and lunar interior processes. In particular, understanding if any lunar surface hydration signatures can be linked to endogenous (e.g., magmatic) sources has the potential to inform us of the nature and distribution of volatiles in the lunar interior. Water content of Apollo samples have been studied via pyrolysis [1] and other more direct [2] approaches. Over one hundred measurements of major constituents of Apollo samples show that the water content in bulk samples is on the order of 44.3 ± 83.4 (2σ) ppm and has an upper limit 63.0 ± 96.7 (2σ) ppm (Figure 1). Though such measurements may suffer from terrestrial contamination, they still provide a useful constraint for comparison against estimates based on remote sensing data (e.g., M reflectance spectra). In addition, measurements of melt inclusions in Apollo 17 orange glass reach ~1400 ppm [3], whereas diffusion modeling of water in Apollo 15 green glass indicates pre-eruptive contents of 745 pprm [2]. Partial melting models for water in lunar apatite suggest sources with at least 64 ppb – 5 ppm water [4]. Questions also remain as to how much water is consistent with current petrologic and magma ocean models, which suggest low values [5]. Orbitally-acquired reflectance spectra provide an additional method of assessing lunar hydration, though it must first be determined what the sources are of any such signatures. Volcanic glasses in pyroclastic deposits represent materials sourced from deep within the lunar interior (~400 – 500 km) that formed as explosive fire fountains enriched in volatiles [6]. Thus orbital mapping of water in lunar pyroclastic deposits can provide large-scale information on the distribution of water in the lunar interior and at sites not sampled during the Apollo missions. Here we examine lunar surface hydration mapped with new thermally-corrected M data (~280 m/pixel) [11] to assess which, if any, signatures may be related to water from the lunar interior. Lunar surface hydration mapped at latitudes between ±30° is expected to be low and within the range measured for Apollo bulk soil samples. In contrast, water in lunar pyroclastic deposits may deviate from this trend depending on volatile content of magma source regions, concentration mechanisms, degassing history, and post-emplacement modification. Methods: In our previous work we demonstrated that Hapke’s Effective Single Particle Absorption Thickness (ESPAT), calculated at ~2.9 μm from single scattering albedo spectra, can be used as a linear proxy for water content [11]. Laboratory experiments and numerical simulations were used to determine specific H2O% ESPAT trends in order to estimate water content directly from ESPAT values. Reflectance spectra were measured for a series of synthetic hydrated basaltic glasses and terrestrial anorthosite samples of different particle sizes and water contents to determine relevant H2O%-ESPAT trends and to examine how such trends are affected by composition and/or particle size. The H2O%-ESPAT trend can also be numerically simulated if the absorption coefficient for a material (e.g., hydrated glass) is known. A relationship between water concentration (c) and the absorbance (A) at ~2.8 μm was derived for silicate glasses by Stolper [12]: c = (18.02·A) / (d· ρ· ε) (1) where d, ρ, and ε are the thickness, density, and extinction coefficient of silicate glasses, respectively; in this work we adopted the average value of ε = 67 L/mol·cm reported by [12] for a variety of glass compositions. Using radiative transfer theory [13], we can determine the relationship between A and ESPAT to convert the above equation to a function relating H2O% and ESPAT. An H2O% ESPAT trend can then be simulated for different particle sizes once the value of ε is known or assumed.