Dusty Rocks in Gale Crater: Assessing Areal Coverage and Separating Dust and Rock Contributions in APXS Analyses

Abstract A thin, patchy layer of airfall dust covers rock surfaces examined by the Mars Science Lab rover Curiosity and complicates interpretation of textures in Mars Hand Lens Imager images and compositions determined by Alpha Particle X‐ray Spectrometer (APXS). Using three image processing methods, we estimate dust coverages for Mars Hand Lens Imager images of APXS targets to Sol 1512. Dust coverages of as is rock targets range from 6% to 77% (±5% to 10% estimated error). Targets brushed by the Dust Removal Tool range to lower coverages than as is targets, but quality depends on surface type; brushed mudstones have the narrowest range and lowest coverages (11–25%), while sandstones vary, ranging to higher coverages (12–58%). Groups of rocks with similar compositions (APXS classes) can have strong correlations between dust coverage and SO 3 /Cl (up to r = 0.985). Dust can also strongly affect the lightest elements measured (Na to Ca). By comparing paired as is and Dust Removal Tool analyses, using the determined dust coverages, and finding a best fit dust thickness (generally ~10 μm), we model relative contributions of the dust and bedrock to extrapolate dust‐free compositions for homogeneous APXS classes. The dust is basaltic with high S and Cl. Dust‐free rocks have higher SiO 2 and Na 2 O (up to 6.5 wt% and 0.5 wt% higher, respectively) and lower SO 3 and CaO (up to 5.5 wt% and 1.3 wt% lower, respectively) than dusty equivalents. Dust most influences compositions that are very different from average Mars, including the alkali‐rich, MgO‐poor Jake M class.