51st Lunar and Planetary Science Conference (2020) 1641.pdf

VEINS IN GLEN TORRIDON, GALE CRATER, MARS: EXPLORING THE POTENTIAL TRANSITION INTO THE SULFATE-BEARING UNIT. Patrick J. Gasda1, D. Das2, M. Nellessen3, E. Dehouck4, W. Rapin5, P.- Y. Meslin6, H. Newsom3, A. Baker3, M. Hoffman3, G. Ganter3, D. Fey7, R. Kronyak8, J. Frydenvang9, R. C. Wiens1, S. Clegg1, S. Maurice6, O. Gasnault6. 1LANL, 2McGill University, 3UNM, 4U. Lyon, 5Caltech, 6IRAP/CNES, 7MSSS, 8U. Tennessee, Knoxville, 9Copenhagen Museum of Natural History.

Introduction: Calcium sulfate veins are remnants the Harlaw Rise and Glen Etive sites, and strata that of groundwater that once was present in the subsurface skirts Central and Western Buttes [13]. of Mars. Hence these veins, and their major, minor, Methods: We use a combination of ChemCam trace chemistry, hydration, and mineralogy are win- chemistry and remote micro imager (RMI) data and dows into the past subsurface aqueous processes of imaging data from Mastcam and Mars Hand Lens Im- ancient Mars. The conditions inferred from the study of ager (MAHLI) from the NASA Curiosity mission. veins are key to understanding the long-term habitabil- Results: Figure 2 summarizes the distribution of ity of the martian subsurface. Veins are ubiquitous in the veins along the traverse through GT (sols 2300– the Murray formation, phyllosilicate rich lacustrine 2600). Figure 2 shows where veins were observed in mudstone deposits in Gale crater. Vein morphology RMI data, with large circles indicating the position of and abundance, hydration, minor chemistry, and min- ChemCam observation points within a raster that indi- eralogy has been well-documented [1–9]. cate a vein target (bottom row), indicate a vein with >20 wt% CaO (2nd row), contained nodules (3rd row), or contained likely cements (top row). Figure 2 shows that veins are typically seen in the coherent endmember targets (Jura strata), and more rarely in rubbly endmember targets (KfH strata). As the rover ap- proached the buttes, the number of veins seen and sampled by ChemCam dramatically increased. Cements and nodules were also observed much more frequently near the buttes. Veins, nodules, and cements were also observed at Harlaw Rise, part of KfH strata. Veins are not observed in pebbles. Veins or patches surrounded by sand are labeled “other” in Figure 2. Discussion: Generally, veins are less abundant in Figure 1: Traverse map through Glen Torridon. GT compared to lower Murray strata [cf. 7]. In GT, As of sol 2600 (Dec 2019), Curiosity has traversed veins that cross-cut each other and the bedrock strata through most of the clay-bearing Glen Torridon (GT) were observed most frequently in the coherent area to Central and Western Buttes (Fig 1). These endmember targets, part of the Jura member. In addi- buttes are interpreted as the edges of the Greenheugh tion to the cross-cutting veins that occur in rubbly pediment. The strata skirting the pediment and the endmember targets, light-toned banding occurs at the buttes and strata to the east of the pediment likely rep- base of Central Butte in the target Sourhope (Fig 3A). resent a transition zone into the sulfate unit identified Light-toned banding follows bedrock bedding and may from orbit [10]. As the rover approaches the sulfate represent primary evaporite deposits or veins that fol- unit of Gale crater, it is important to track the changes low bedding due to coarser grain sizes or weaknesses in the veins. The sulfate unit may represent a transition, along bedding planes. Similar rock textures with light- or the onset of the transition, from a relatively “warm toned banding and nodular gypsum have been studied and wet” Mars to dryer conditions [10,11]. on Earth, including in southern New Mexico, where At least two distinct chemical endmembers and large inland seas disappeared during the Jurassic peri- three distinct facies have been observed in GT. [12] od, forming large deposits of gypsum [14]. discusses the bedrock chemical endmembers observed Nodules have been observed in a few locations by ChemCam: a ~54 wt% SiO2 and 6–10 wt% MgO [15], primarily in KfH rubbly endmember bedrock, and “coherent” endmember, and a ~56 wt% SiO2 and 4–6 have compositions rich in Fe, Mn, or P, reminiscent of wt% MgO “rubbly” endmember. The pebbles observed diagenetic features in the Sutton Island member (strata throughout Glen Torridon match the rubbly endmem- 100–200m below GT) [16]. Embedded Fe-rich dark- ber composition. Different facies include Jura member toned surface features have been observed in targets bedrock, pebble strewn areas, capping cross-bedded near the buttes in association with the Ca-sulfate rich sandstone Knockfarril Hill (KfH) member that include light-toned veins and surface features (Fig 3B). Veins 51st Lunar and Planetary Science Conference (2020) 1641.pdf

sometimes cross-cut the nodules. Pebbles are frequent- consistent with GT strata as being clay-rich bedrock ly pitted; the pits may be remnants after weathering or deposited in a wetter environment. dissolution of the nodules. Implications: As the rover continues to ascend Mt Cements are inferred from chemical composition Sharp, this ongoing work will shed light on the transi- trends and have been directly observed in the target tion from clay-rich facies to sulfate-rich facies, which Gleneagles (Fig 3C). In targets without veins or nod- potentially represents the overall drying of the martian ules in images, but have >4 wt% CaO, we can infer the climate. Hence, understanding the transitional strata presence of sulfate cements. Three points in Gleneagles has important implications for understanding Mars that are enriched in CaO also show light-toned materi- geologic and climatic history. als within the ChemCam laser pits (Fig 3C) providing Acknowledgments: NASA Mars Exploration program; further evidence of a cement. Ca-sulfate cements likely CNES, France; Carlsberg Foundation. formed when Ca sulfate rich fluids permeated the bed- References: [1] Nachon et al (2014) JGR:P, 119 (9), rock while it was still porous, before later fracturing 1991–2016. [2] Nachon et al (2016) Icarus, 281, 121–36. [3] Rapin et al (2016) EPSL, 452, 197–205. [4] Schwenzer et al that produced the more typical cross-cutting veins. (2016) MAPS, 51(11), 2175–2202. [5] Gasda et al. (2017) The bedrock near the buttes tend to have a chemical GRL, 44(17), 8739–48. [6] L’Haridon et al. (2018) Icarus, composition intermediate between the rubbly and co- 311, 69–86. [7] Nachon et al. (2020) Sedimentology, in herent endmembers. Some KfH strata (e.g., Harlaw press. [8] Das et al. (2020) JGR:P in review. [9] L’Haridon Rise) is rich in nodules, similar to the butte skirting et al. (2020) JGR:P in review. [10] Milliken et al. (2010) units, which may be explained by coarser grain size of GRL, 37(4). [11] Fraeman et al. (2016) JGR:P, 121(9), the unit. It is unclear why the nodules only occur at one 1713–36. [12] Dehouck et al. (2020) this meeting. [13] Fedo et al. (2020) this meeting [14] Kirkland et al. (1995) New location along the GT traverse. Mexico Bureau of Geology and Mineral Resources Bulletin ChemCam does not observe large changes in vein 147. [15] Minitti et al. (2020) this meeting [16] Meslin et al. major oxide composition as compared to previous (2018) LPSC 1447. Murray veins. Veins in GT tend to have very low (<10 ppm) Li and almost no B has been detected, which is

Figure 2: Summary of ChemCam vein, nodule, and cement observations in GT (sols 2300–2600) for coherent (lavender), rubbly (gold), and other targets (grey). If the RMI observation includes a vein, there is a point in “total veins” row. The dot’s size repre- sents the number of ChemCam LIBS observations (up to 7) that indicate a vein. Dots with black outlines indicate veins with >20 wt% CaO.

Figure 3: MAHLI images of A) Sourhope; B) Everbay; C) Gleneagles. Labels: 1) crosscutting veins; 2) light-toned banding; 3) nodules; 4) light-toned surface features; 5) dark-toned surface features. Circled ChemCam cyan points have elevated CaO.