 MORPHOLOGICAL AND CHEMICAL CHARACTERISTICS OF SEDIMENT IN THE ROCKNEST EOLIAN SAND SHADOW, GALE CRATER, MARS.  W. Goetz 1, M. B. Madsen 2, K. S. Edgett 3, P.-Y. Meslin 4, D. L. Blaney 5, N. Bridges 6, B. Clark 7, M. Fisk 8, S. F. Hviid 9, G. Kocurek 10 , J. Lasue 4, S. Maurice 4, H. Newsom 11 , N. Renno 12 , D. Rubin 13 , R. Sullivan 14 , R. C. Wiens 15 , and the MSL Science Team, 1 Max-Planck-Institut für Sonnensystemforschung, Katlenburg-Lindau, Germany (goetz@mps.mpg.de), 2 Niels Bohr Institute, University of Copenhagen, Denmark, 3 Malin Space Science Systems, San Diego, CA, USA, 4 IRAP, CNRS/UPS, Toulouse, France, 5 JPL, Pasadena, CA, USA, 6 Applied Physics Laboratory, Laurel, MD, USA, 7 SSI, Boulder, CO, USA, 8 Univ. of Oregon, Corvallis, OR, USA, 9 DLR, Berlin, Germany, 10 Univ. of Texas, Austin, TX, USA, 11 Univ. of New Mexico, Albuquerque, NM, USA, 12 Univ. of Michigan, Ann Arbor, MI, USA, 13 USGS, Flagstaff, Arizona, USA, 14 Cornell Univ., Ithaca, NY, USA, 15 Los Alamos National Laboratories, NM, USA.  Introduction:  Curiosity spent over 40 sols at an eolian deposit informally named 'Rocknest'. The deposit is ~50 cm wide, 15-20 cm high (at crest), and extends over several meters in a north-south direction. Material scooped from 5 troughs, respectively, on sols 61, 66, 69, 74, and 93 (each trough ~4 cm wide and deep, Fig.1) was used to clean the sample acquisition and portioning system (called CHIMRA). Material from the later troughs was also delivered to the chemical laboratories (CheMin, SAM) in the rover body. Scooping exposed some interior banding of the deposit that could be imaged at high resolution by the Mars Hand Lens Imager (MAHLI) [1] and characterized chemically by Laser Induced Breakdown Spectroscopy (LIBS) using the ChemCam instrument onboard [2]. Observations: The prime motivation of this work is to characterize the bright banding observed in scoop troughs #1-4 (Fig. 1). Here we compare camera data (Figs. 2-3, see also [3]) to ChemCam data (Figs. 4-5). Interpretation and conclusions: The Rocknest sand shadow is armored by a thin layer of ~1 mm large subrounded sand grains. Its bulk is made up of fine sand and likely silt and reddish dust (particles < 100 µm). Frequent (comparatively bluish) sand particles (~200 µm) as well as whitish (sulfate rich?) inclusions are scattered throughout the bulk of the deposit (Fig. 2b). The armoring of the deposit by mm-sized particles suggests that this deposit has been immobile for a significant amount of time, since the Martian atmosphere has become too thin to transport large sand grains in the eolian regime. This allowed for slow diagenesis and crust formation in particular. Indeed the floors of all scoop troughs are littered by large angular crust fragments (Figs. 2-3). Scoop troughs #1-4 show internal banding. It is unclear why scoop trough #5 is almost devoid of internal banding. The head walls of some trenches dug at high northern latitudes by the Phoenix Mars lander (May- Fig. 2. Scoop trough #1 before it was filled by material from trough #2 (see Fig. 1). MAHLI ID: 0066MH0075001000E2.  Note cemented clods detached from the far trough wall (green arrows point to sharp corners, reentrants and fractures attesting to crust formation). Blue arrows highlight the bright band that continues on the near trough wall (inset (a)). Furthermore, inset (a) highlights another crust fragment (blue dotted arrows). Inset (b) shows a large clod on the trough floor that contains numerous bright (sulfate rich?) inclusions of submm size.  Fig. 1. Mastcam-34 image of scoop troughs (labeled #1-5) in Rocknest sand shadow deposit (ID: 093ML0544000000E1). Each trough measures ~4 x 15 cm and is ~4 cm deep. Troughs #1-#4 show a bright band at a depth of ~2 cm, #5 does not show a clear banding. The inset shows the shot locations of several ChemCam experiments (from l. to r.: Kenyon-1x10, Epworth2-1x5 [virtually superposed], Epworth31x15). All ChemCam line scans were run from bottom to top. Yellow arrows point to the location of the last shot. The present abstract only discusses Kenyon data. Trough #1 was partly filled up by material from #2, so banding is no longer visible in this image. Also the far wall of trough #2 collapsed on sol 84. Oct. 2008) do show weak banding interpreted as being caused by the scoop. However, in the case of the Rocknest deposit in Gale, the bright band is not only seen at the scoop entrance, but also on the near (and even far) side of the scoop trough suggesting that it is part of the internal structure of the deposit rather than being caused by the compressive action of the scoop (Fig. 2a). The bright band is found at a depth of ~2 cm that is similar to the Martian diurnal skin depth for sandy soil [4]. ChemCam data (Figs. 3-5) does not provide a simple chemical explanation of that band.   The chemical variation shows anomalies for spot locations #5 and #8 (Fig. 4). ChemCam data alone do not allow to decide whether these anomalies are due to enrichment of chemical elements at some depth in the subsurface, or if a specific mineral grain was hit by the laser. Given the firm evidence for crust formation in the near subsurface (Figs. 2-3), the mobilization of some elements (alkali ions in particular) is conceivable. Evidence for this process has been seen frequently by the Spirit rover [5], and should be a widespread process on the surface of modern Mars.   References: [1] Edgett, K. S. et al. (2012) SSR 170, 259-317. [2] Wiens, R. C. et al. (2012) SSR 170, 167-227. [3] Edgett, K. S. et al. (2013) LPS XLIV, this volume. [4] Chamberlain, M. A. and Boynton, W. V. (2005) LPS XXXVI, 1566. [5] Arvidson, R. E. et al. (2011) J. Geophys. Res. 115, E00F03.   Fig. 3. Mastcam-100 image of scoop trough #2, where the ChemCam run "Kenyon" was performed (see Fig. 1). This run consisted of 10 laser spot locations: Yellow crosses mark laser spot locations that are directly visible in RMI images (Remote Microscopic Imager), whereas red crosses mark inferred locations. Blue arrows pont to the bright band (as in Fig. 2). Note the existence of an additional (much narrower) bright band very close to the topmost surface (marked by dotted arrows). The top band may be caused by the scoop whereas the broad band below should be indegenous to the drift deposit. Image ID: 0066MR0369001000E1.  Fig. 4.  Elemental ratios (in wt%) as inferred from multivariate PLS (Partial Least Squares) analysis of the Kenyon data (Fig. 3). Most spot locations are consistent with a basaltic composition (cluster to the upper right) with two prominent outliers: #5 and #8 with, respectively, a more mafic and a higher-silica composition. Fig. 5. (a) Abundance of major oxides as inferred from multivariate PLS (Partial Least Squares) analysis of the Kenyon data (Fig. 3) and (b) relative abundances of minor elements (as inferred from peak analysis). The alkali ions are maximum either at spot #5 (Li) or at spot #8 (Na, K). Ti and Mn are high at spot #5 and low at spot #8. (a) (b) 
