Search

The glass phase in volcanic rocks presents a challenge to obtaining compositional data from visible and short‐wave‐infrared (VSWIR) and mid‐infrared (MIR) spectral data of remote surfaces due to its amorphous structure and variable composition. Nonetheless, glass is a common phase in volcanic materials because it forms via the rapid quench of magma and can constitute up to the entirety of a volcanic deposit. Use of partial least squares regression (PLS) to predict glass contents creates models that are insensitive to viewing geometry and sample conditions such as grain size and spectrally inactive compositional variables, enhancing the ability to detect glasses with remote sensing. PLS models are used here to predict crystallinity and oxide composition of samples from VSWIR and MIR spectral data using training spectra from natural volcanic rocks and geologically relevant synthetic samples. Three spectral resolutions of VSWIR and MIR spectra (1, 10, and 100 nm/band, and 1.9, 19, and 190 cm−1/band, respectively) were tested to assess the effects of collection configuration on different spectrometers. PLS models trained on 1 nm and 1.9 cm−1 data sets have the lowest uncertainties of glass modal abundance for VSWIR and MIR, respectively. MIR models predicting sample wt. % SiO2 and FeO, and VSWIR models of wt. % FeO provide accurate estimates (e.g., RMSE‐P of 3.4 wt. % FeO) at all spectral resolutions. Results are based on training data sets skewed to mafic compositions, which affects model accuracies. Plain Language Summary: Glass is difficult to identify and characterize from remote sensing observations of volcanic rocks. Differences in the number of spectral bands between laboratory and currently orbiting instruments further complicate interpretation of remote observations because glass features are subtle. A machine learning regression technique (partial least squares, or PLS) is used to predict the amount of glass and chemical composition using reflectance and emission spectral data of natural volcanic and synthetic glass samples. Results from three spectral resolutions were compared and affirm that spectral data can quantify phase abundance and composition in natural glassy samples such as volcanic materials. Key Points: PLS models can predict the modal glass abundance of volcanic samples with accuracies of 12%–15% from VSWIR and MIR spectraPLS models can predict bulk sample oxide abundances with accuracies of 1.5–6.9 and 0.9–4.2 wt. % from VSWIR and MIR spectra, respectivelyPLS model accuracy is minimally impacted by the number of available spectral bands [ABSTRACT FROM AUTHOR]

Introduction: Raman spectroscopy is highly sensitive to molecular functional groups including carbon-rich structures such as aromatic moieties in meteorites. Raman spectroscopic investigation of meteoritic carbon is particularly useful as thermal metamorphic grade/history of carbonaceous chondrites can be roughly estimated through the spectral properties of polyaromatic structures in meteorites. For instance, mathematical expressions that allow estimating the peak metamorphic temperature of chondrites (mostly high-temperature chondrites) using the Raman spectral parameters of the first-order carbon peaks were previously reported [1, 2]. Recently, an expression based on the full-width half-maxima of the G band was reported and it seems to be particularly appropriate for lowtemperature carbonaceous chondrites [3]. In this work, we are collecting hyperspectral Raman images and spectra for various low temperature carbonaceous chondrites such as CM2 and C2 ungrouped (C2-ung) chondrites in an effort to estimate and compare the peak thermal metamorphic temperatures experienced by these meteorites. Samples: Our sample set for this investigation contains CM2 chondrites (Aguas Zarcas, Jbilet Winselwan, Murchison, Asuka (A) 12437, Asuka 12236, Asuka 12408, Asuka 12248, Yamato 980039, Northwest Africa 13711) and C2-ung (Tagish Lake, Tarda, and Queen Alexandra Range (QUE) 99038) chondrites. All meteorites were prepared in the form of polished sections with large sample surface areas for measurements. Technique: Raman spectroscopic experiments were conducted using a commercial confocal Raman microspectroscopy system (WiTec alpha300R) equipped with a 600 g/mm grating, 532 nm Nd:YAG laser, and a 50× objective (NA = 0.8). The spectrograph was calibrated using a silicon wafer substrate prior to measurements. Individual Raman spectra were acquired with a 1-s integration time for 60 accumulations within 800-2000 cm-1, while twodimensional maps were acquired with 0.1 s integration time. The power density of the laser beam was ~0.2-0.5 mW/µm2 on the sample surface. The collected spectra were processed using the commercial ProjectPLUSv4 (WiTec GmbH) software package. Artificial fluorescence background was removed from the spectra, which were then fitted with Lorentzian-Breit-Wigner-Fano (LBWF) spectral model to extract the D and G band parameters of aromatic hydrocarbons in the meteorites (e.g., Ferrari and Robertson, 2000; Chan et al. 2020). This method has previously used to successfully obtain the PMT values for different meteorites [e.g., 6]. Results: Raman spectra of all meteorites present well-developed D and G carbon bands at ~1370 and 1581 cm-1. These carbon bands are due to disordered sp3 and graphitic-like sp2 carbon bonds, respectively [7, 8]. The fullwidth half-maxima (FWHM) values for the G band of Aguas Zarcas, Jbilet Winselwan, A 12236, Tarda, Tagish Lake, QUE 99038 were found to be, respectively, 97 cm-1, 93 cm-1, 84 cm-1, 94 cm-1, 90 cm-1, and 60 cm-1. Using the expression by [3], we found that these values roughly correspond to ~71 °C, 79 °C, 101 °C, 77 °C, 86 °C, and 195 °C. For Jbilet Winselwan, we note that we did not include the shock-heated parts into consideration for this work, which would yield much higher temperatures. Based on these results, A 12236 seems to be slightly more heated compared to other currently studied CM2 chondrites. Tarda and Tagish Lake appears to be compatible with being type 2; however, QUE 99038 presents a much higher peak metamorphic temperature. The rest of the meteorites are currently being measured as well, and their results will be presented at the meeting. Acknowledgements: We thank NIPR and NASA-JSC for providing some of the meteorite samples used in this study. MY acknowledges support from TUBITAK (grants #119N207 and #120Y115). [ABSTRACT FROM AUTHOR]

Introduction: Establishing links between the plethora of meteorite samples in collections on Earth and the parent body asteroids they originated from is a valuable outcome of past (NEAR-Shoemaker, Hayabusa, Dawn), present (Hayabusa2, OSIRIS-REx) and future (Lucy, Psyche) space missions [1]. Sample return missions (Hayabusa, Hayabusa2, OSIRIS-REx) provide much needed context for the study of relevant extraterrestrial samples, and by utilizing these specimens of asteroids in the laboratory, we can supply detailed chemical, mineralogical and petrological information for reference libraries of materials relevant to asteroid missions [2,3]. Laboratory-based spectral analysis of well-characterized meteorite samples can be applied to more quantitatively analyze asteroid remote sensing data in conjunction with returned samples. Previous work has evaluated the spectral features of ordinary chondrite material [2], and revealed the individual effects of temperature and grain size on the visible and near-infrared (VNIR) spectra of silicate [4,5] and meteorite [6] powders. In this work, we examine the combined effects of physical variables (temperature, particle size) and chemical variables (petrologic type, metal fraction) on VNIR spectra of ordinary chondrite meteorite powders. Sample Selection: Six ordinary chondrite falls spanning the full range of equilibrated petrologic types were sourced from the American Museum of Natural History (AMNH) meteorite collection: Soko-Banja (LL4), Mangwendi (LL6), Mount Tazerzait (L5), Suizhou (L6), Jilin (H5), and Zhovtnevyi (H6). Methods: Bulk chondrite samples (~2-6 cm3 volumes) were first scanned using AMNH's computed tomography scanner, with resolutions ranging from 6-11 microns/voxel, to characterize their 3D structure and determine the abundance of optically opaque (metal and sulfide) phases. These samples were subsequently mapped for the X-ray intensities of ten major and minor elements at 4 microns/pixel resolution, over a sufficient area (~1 cm2) to characterize mineralogy in a 2D slice [7]. Element maps were linearly combined to determine the mineralogy of each individual pixel in every map, enabling quantitative determination of the relative mineral abundances in each meteorite [8]. Samples were then powdered and sieved to four different size fractions (25-63 µm, 63-90 µm, 90-125 µm, 125-250 µm) to capture the spectral diversity associated with asteroid regoliths dominated by various grain sizes. Each individual size fraction for each meteorite was characterized via X-ray diffraction (XRD) to assess the distribution of metal grains across the size fractions, in case of preferential size sorting of metal grains during sieving. VNIR spectra of the ordinary chondrite material were then measured under simulated asteroid surface conditions (defined as ~10-6 mbar, -100°C chamber temperature, and low intensity illumination) at a series of temperatures chosen to mimic near-Earth asteroid surfaces. These reflectance spectra were collected in increments of 10°C, over the range 10°C to 100°C. Results: The VNIR spectra show minimal variation in both major absorption bands across the simulated near-Earth asteroid temperature regime. Spectral changes due to particle size are consistent across samples, with the smallest and largest grain sizes having the highest reflectance. Previous spectral investigations of ordinary chondrites [3] removed the metal fraction from the powdered meteorite sample. Notably, we retained the metal fraction in our powders instead of analyzing the silicate fraction only. In our measurements, we observe distinct offsets in spectral features when compared to analyses of purely silicate fractions. XRD analysis shows that the largest size fraction of nearly every sample contains relatively more metal, likely due to the retention of metal nuggets in the largest size fraction during sieving. The more petrologically pristine samples (e.g., LL4) from each ordinary chondrite group display relatively shallower band depths than their more petrologically altered counterparts (e.g., LL6). The band depths shift to higher wavelengths as temperature, grain size, and petrologic type increase. Spectral studies of meteorites combined with detailed petrologic analysis of the samples greatly enhance interpretation of current and future planetary remote sensing data sets. Importantly, understanding the spectral contribution of the metal fraction will aid in investigations of metal-rich mission targets such as asteroid 16 Psyche. This work continues an effort to develop a comprehensive spectral library of materials in conditions relevant to airless bodies and contemporaneous asteroid missions. [ABSTRACT FROM AUTHOR]