The Near-Infrared Spectral Characteristics of Water Ice, Epsomite, and Halite Mixtures Relevant to Europa

TL;DR

This work provides the first near-infrared spectra of grain-scale mixtures of $\mathrm{H_2O}$ ice, epsomite, and halite at cryogenic temperatures and assesses how well linearly and intimately mixed Hapke models reproduce the measurements. By systematically varying grain sizes and using representative $<D>=\frac{2}{3}$ of the aliquot average, the study shows water ice features dominate mixed spectra and that smaller ice grains strongly influence the $2.0\,\mu$m feature in epsomite, while larger grains saturate absorption features, complicating grain-size discernment. The results indicate that both mixing schemes fit water ice spectra within ~10%, but epsomite modeling is less reliable—potentially due to incorrect cryogenic epsomite optical constants—highlighting the need for updated constants for accurate abundance retrievals on Europa. The findings underscore the importance of quantitative spectral mixture analyses for interpreting Europa Clipper/MAJS data and point to the necessity of revised epsomite constants to improve fidelity in endmember unmixing and grain-size determinations on icy surfaces.

Abstract

We publish the first NIR spectra of grain particulate mixtures of water ice, epsomite, and halite at cryogenic temperatures. Furthermore, we perform a quantitative assessment of the ability of both intimately- and linearly-mixed models to reproduce laboratory data of different grain mixtures of water ice, as well as water ice mixed with epsomite. We find that smaller grains of water ice impart a stronger influence than larger grains of water ice on the 2.0 μm spectral feature in epsomite, and grain size signatures for both halite and epsomite are challenging to discern for larger grain sizes as a result of the saturated absorption features. These findings may indicate that an observation bias toward smaller grain sizes of ice could exist, and that quantitative assessments provided by spectral mixture analyses will be the most reliable method for determining compositions and abundances of materials. We also find that the linearly-mixed and intimately-mixed models of water ice appear to match the laboratory spectra as expected, though still display some inconsistencies, often either in the continuum or the absorption features. When modeling pure water ice and water ice mixed with epsomite, no discernible difference is observed between the fits of the linearly- and intimately-mixed models. Future spectral mixture analyses that use epsomite should be aware of a potential error in the published epsomite optical constant data, in which the cryogenic data appear to be taken at ambient conditions.

Figures (11)

• Figure 1: Glove box setup for the experiments conducted in this study. The LN$_2$ dewar and gold standard are mounted on a linear stage that allow for the gold and sample to move into and out of the field-of-view of the spectrometer fiber-optic for measurements. Bottom left inset: photograph of 50% $<63$$\mu$m + $50\%$$250-500$$\mu$m water ice.
• Figure 2: Block diagram of the CRONOS laboratory setup; dimensions are not necessarily to scale. All materials and equipment were given at least 30 minutes in the antechamber before moving into the main chamber, to ensure no atmosphere was exchanged between the two areas.
• Figure 3: Spectra of water ice of the four grain size bins used in this study. Top: solid lines represent pure grain size bins. Bottom: dashed lines represent $50\%$/$50\%$ by volume mixtures of several combinations of water ice grain sizes. Notice that there is minimal discrepancy between the grain size mixtures, and may therefore be challenging to discriminate visually without spectral mixture analysis. Measured temperatures for each sample are provided in the figure legend as well as in Table \ref{['tab:samples']}.
• Figure 4: Spectra of epsomite of the four grain size bins used in this study, acquired both under ambient (dashed lines) as well as cryogenic (solid lines) conditions. Fine structure in the spectral absorption features in the cryogenic spectra are expected, and are characteristic of colder temperatures. Measured temperatures for each sample are provided in Table \ref{['tab:samples']}.
• Figure 5: Spectra of halite of the four grain size bins used in this study, acquired both under ambient (dashed lines) as well as cryogenic (solid lines) conditions. Fine structure in the spectral absorption features in the cryogenic spectra are expected, and are characteristic of colder temperatures. Measured temperatures for each sample are provided in Table \ref{['tab:samples']}.