Mineral Distributions In Lunar Crust: Results From Radiative Transfer Modeling Of Central Peaks

The high abundance of pure anorthosite (>90%) from the returned Apollo lunar samples indicate the preexistence of a global magma ocean on the Moon, which hypothesis is called the Lunar Magma Ocean model (LMO). The lunar crust is formed by the crystallization and floatation of plagioclase at the late stage of LMO due to its low density and buoyant, thus it is abundant of anorthosite based on LMO. However, the collection of lunar samples are limited to the near side, and are suggested to not contain pure highlands sample. Remote sensing study could earn a global view of the compositional and mineralogical distribution on the Moon. Especially the lunar central peaks, which are believed to have formed by a giant impact and can represent the composition of deep crust. Based on the crater size-depth relation, we can learn both the lateral and vertical composition of the crust by studying different sized impact craters on a global scale. Then we can validate or improve the former sense of LMO. We build a radiative transfer model, and firstly, the imaginary part (k) of the optical constant for olivine using is calculated from laboratory spectra with radiative transfer model, then correlate k with Fo number (molar Mg/(Mg+Fe) in olivine) to obtain their correlation parameters. Secondly, we consider the influence of space weathering by adding SMFe (Submicroscopic Fe) into the radiative transfer model, and we add lunar sample porosity and backscattering effect into the radiative transfer model. Last but not least, olivine, orthopyroxene, clinopyroxene and plagioclase are added as the endmembers and mixed from 0% to 100% at 5% intervals to build the spectral library. For an unknown observed spectrum extracted from central peaks, its mineral mode can be derived by finding the best match from the spectral library. After surveyed the spectra of 101 central peaks with their dimeters ranged from 30-200 km all over the Moon, we derive the mineral abundances of 86 central peaks having recognizable mineral absorption features with our radiative transfer model. Our research shows that, the most abundant mineral in the lunar crust is plagioclase, and the most abundant mafic mineral in the lunar crust is orthopyroxene, followed by olivine and clinopyroxene, which is consistent with the lunar highlands sample studies. The plagioclase in FHT (Feldspathic Highlands Terrane) is more abundant than PKT (Procelarum KREEP Terrane) and SPAT (South Pole-Aitken Terrane), considering the crust in FHT is thicker than PKT and SPAT, which may imply mafic minerals tend to be more concentrated in thin luanr crust. The primary rock types of the central peaks for three terranes are:FHT, Anorthosite, noritic anorthosite, anorthositic norite, and minor anorthositic troctolite, norite to olivine norite; PKT, anorthosite, noritic/gabbroic anorthosite, anorthositic troctolite and nortie; SPAT, gabbroic norite, norite, anorthositic norite and anorthosite. We choose 6 typical central peaks from FHT, PKT and SPAT, then use the Modified Gaussain Modle (MGM) to deconvlove their spectra and analyse the compositional variations in three terranes. The absorption feature of olivine shows that Mg’(molar Mg/(Mg+Fe)) is high in FHT (>90), medium in PKT (25-57). The Mg’in PKT central peaks are lower than the Magnesium suite (Mg’>80)。The SPAT is abundant of orthopyroxene (> 30%), and its major rock type is norite. We correlate plagioclase (mafic minerals) content in central peaks with their proximity to lunar crust-mantle interface (P) for the whole Moon, FHT, PKT and SPAT respectively. It turns out that no obvious correlation is observed between the variation of plagioclase (mafic minerals) content and P. We then examined the P of central peaks having pure anorthosite exposures, and find their range of variation is large, from 25 km to 60 km. This observation may indicate the absence of a pure anorthosite layer in the lunar crust, or this layer have already been modified by later magma intrusions or some gigantic basin impacts, and generating a heterogeneous distribution of lunar crustal compostion both lateraly and verticaly. The compositional difference in three terranes may be caused by different composition and evolve degree of the parent magma.