Nontronite Detections in Nili Fossae Based on an Impact

46th Lunar and Planetary Science Conference (2015) 2852.pdf

NONTRONITE DETECTIONS IN NILI FOSSAE BASED ON AN IMPACT-ALTERED NATURAL NONTRONITE SAMPLE RESEMBLE REGIONAL-SCALE SPECTRAL VARIABILITY PREVIOUSLY ASSOCIATED WITH PHYLLOSILICATE DIAGENESIS. L. R. Friedlander1 and T. D. Glotch1, 1Geosciences Department, Stony Brook University (255 Earth and Space Sciences Building, Stony Brook, NY 11794-2100, lo- [email protected]).

Introduction: There is evidence for extensive clay CRISM Image Processing and Analysis. We proc- mineral deposits on Mars from visible near-infrared essed CRISM image FRT000097E2 (Figure 1) using (VNIR) remote sensing techniques [1-8]. Most of these the latest version of the CRISM Analysis Toolkit identifications have been made in the older, heavily (CAT: version 7.2.1) for ENVI/IDL Classic (version bombarded southern highlands of Mars [9]. Several are 5.1) image processing software. FRT000097E2 con- associated directly with impact craters, or with depos- tains image data in the most up-to-date format avail- its of impact ejecta. For example, in [6] several occur- able from the CRISM archive (TRR3) and was radi- rences of characteristic spectral features indicative of ometrically calibrated as well as converted to I/F from specific clay mineral species were described within raw CRISM experiment data record (EDR) prior to craters surrounding the Nili Fossae. Phyllosilicate- download. We performed atmospheric correction on bearing outcrops have been identified on the rim of this image using the latest version of the volcano-scan Endeavor Crater using orbital remote sensing data method contained within CAT and converted CRISM [10], and saponite has been identified in the center of a I/F to apparent surface reflectance using the systematic crater within the Mawrth Vallis region [11]. Indeed, processing tools embedded in CAT [21]. Regions of craters have been specifically targeted for investigation interest were selected using the spectral angle mapping as self-contained locations where the martian subsur- (SAM) tool in ENVI/IDL to locate areas within the face can be directly probed [12,13]. CRISM image with spectral features resembling those The use of craters as “windows [12]” into the mar- of target laboratory spectra. Individual spectra were tian subsurface may be problematic for remote sensing. then extracted from these regions, ratioed and com- The impacts that form craters also expose the minerals pared to laboratory spectra directly. present at the time of impact to thermal and shock alteration. Both of these processes have been previously shown to alter mineral structures and spectral signa- tures [14-17]. This abstract explores the possibility that impact-induced spectral changes may produce some of the regional-scale spectral variability observed by remote sensing on Mars. Interestingly, the impact- induced changes observed for a natural nontronite sample occassionaly overlap and may be conflated with spectral change that has previously been associ- Figure 1. Local context of CRISM image FRT000097E2 from the Nili Fossae region. ated with clay mineral diagenesis on Mars [6, 18-20]. Results: Several regions of FRT000097E2 were Materials and Methods: The Clay Minerals Soci- expected to contain phyllosilicates (Figure 2a), but we ety source clay NAu-1 (nontronite) was ground and were especially interested in two small craters within separated to its < 2 µm size-fraction and sent to the the image (Figure 2b) for what they might reveal about Flat Plate Accelerator at NASA’s Johnson Space Cen- the relationship between regional spectral variability ter where it was exposed to a serious of experimental and impact processes. The regions in and around these impacts at six controlled peak impact pressures: 10.2, craters were detected by the SAM algorithm as con- 19.7, 25.2, 30.6, 34.6, and 39.1 GPa. VNIR reflectance taining spectral features consistent with both impact spectra (0.25 Ð 2.5 µm) of the returned (~0.15 g) im- altered and diagenetic phyllosilicates (Figure 3). Spec- pact-altered samples were collected using an ASD In- tra extracted from these regions were then compared to struments (now PANalytical) Field Spec 3 Max Spec- laboratory spectra of impact-altered nontronite and troradiometer fitted with an 8-degree field of view several diagenetic clay minerals. This comparison foreoptic. These spectral data were then used in the showed that some regional variability previously at- analysis of a Compact Reconnaissance Imaging Spec- tributed to diagenesis may be partially explained by trometer for Mars (CRISM) image from the Nili Fos- impact alteration, especially in heavily bombarded sae region. regions (Figure 4). 46th Lunar and Planetary Science Conference (2015) 2852.pdf

Figure 3. SAM pixel classifications overlain on a single- band 1.3 µm image of FRT000097E2. End-member spectral Figure 2. False-color CRISM image showing the distribution comparisons: impact-altered nontronite after impacts at 39.1 of phyllosilicates (red), olivine (green) and low-Ca pyroxene GPa (a), CRISM library nontronite NDJB26 (b), CRISM (blue) within FRT000097E2 overlain on HiRISE image library smectite BKR1JB006 (c), CRISM library illite PSP006778_1995 from the same geographic region (a). The LAIL01 (d), CRISM library chlorite LACL14 (e), and illite overlay highlights the relationship between clay mineral with a relaxed spectral angle differential of 0.100 radians (f). deposits and two small craters in the southwest corner (b). Arrows indicate locations of extracted spectra. Discussion: NAu-1 nontronite contains and Al-rich contaminant, most likely kaolinite. The persistence of these Al bands in our VNIR reflectance results likely contributes to the observed spectral changes. However, pure minerals are rare on planetary surfaces and so the effect of impacts on the spectroscopic signature of this natural mineral mixture and its similarity to “diagenetic” trends warrants further detailed investigation. References: [1] Bibring J.-P. et al. (2006) Science, 312, 400-404. [2] Poulet F. et al. (2007) JGR, 112, E08S02. [3] Loizeau D. et al. (2007) JGR, 112, E08S08. [4] Mangold N. et al. (2007) JGR, 112, E08S04. [5] Mustard J. F. et al. (2009) JGR, 114, E00D12. [6] Ehlmann B. L. et al. (2009) GRL, 37, L06201. [7] Ehlmann B. L. et al. (2013) Space Sci. Rev., 174, 329-364. [8] Michalski J. R. et al. (2010) Icarus, 206, 269-289. [9] Carter J. et al. (2013) JGR: Planets, 118, 831-858. [10] Wray J. J. et al. (2009) GRL, 36, L21201. [11] Mckeown N. K. et al. (2009) JGR, 114, E00D10. [12] Schwenzer S. P. et al. (2009) MEPAG white paper. [13] Schwenzer S. P. et al. (2010) LPS XLI, Abstract #1589. [14] Friedlander L. R. et al. (2012) LPS XLIII, Abstract #2520. [15] Friedlander L. R. et al. (in review) JGR: Planets. [16] Che C. et al. (2011) JGR, 116, E05007. [17] Gavin P. et al. (2013) JGR: Planets, 118, 1-16. [18] Milliken R. E. (2014) Proc. 8th Intl. Conf. Mars, Abstract #1253. [19] Bishop J. L. Figure 4. Extracted spectra from FRT000097E2 showing et al. (2013) Plan Space Sci., 86, 130-149. [20] Viviano C. E. regional variability (a) compared to spectral variability be- et al. (2013) JGR: Panets, 118, 1858-1872. [21] http://pds- tween primary phyllosilicates and diagenetic products (b) geosciences. wustl.edu/missions/mro/crism.htm. and spectral change produced by impact alteration (c).