Planetary and Space Science 70 (2012) 84–95

Contents lists available at SciVerse ScienceDirect

Planetary and Space Science

journal homepage: www.elsevier.com/locate/pss

Gale Crater: Formation and post-impact hydrous environments

Schwenzer S.P.a,h,n, Abramov O.a,i, Allen C.C.b, Bridges J.C.c, Clifford S.M.a,1, Filiberto J.d,2, Kring D.A.a,3, Lasue J.a,e,j, McGovern P.J.a,4, Newsom H.E.f, Treiman A.H.a,5, Vaniman D.T.g, Wiens R.C.e, Wittmann A.a,k,6 a Lunar and Planetary Institute, USRA, 3600 Bay Area Blvd., Houston TX 77058, USA b ARES, NASA JSC, Mail code: KA, 2101 NASA Road One, Houston, TX, 77058, USA c Space Research Centre, Dept. of Physics & Astronomy, University of Leicester, LE1 7RH, UK d Southern Illinois University, Geology Department—MC 4234, 1259 Lincoln Dr, Carbondale, IL 62901, USA e Los Alamos National Laboratory, Space Remote Sensing, ISR-2, Mail Stop D-466, Los Alamos, NM 87545, USA f Institute of Meteoritics and Dept. of Earth and Planetary Sciences MSC03-2050, University of New Mexico, Albuquerque, NM 87131, USA g Planetary Science Institute, 1700 East Fort Lowell Rd., Tucson, AZ 85719, USA h The Open University, Earth and Environmental Sciences, Walton Hall, Milton Keynes, MK7 &AA, United Kingdom i U.S. Geological Survey, Astrogeology Research Program, 2255N. Gemini Dr., Flagstaff, AZ 86001, USA j Universite´ de Toulouse; UPS-OMP; IRAP; Toulouse, France k Department of Earth and Planetary Sciences, Washington University St. Louis, Campus Box 1169, 1 Brookings Dr., St. Louis, MO 63130-4899, USA article info abstract

Article history: Gale Crater, the landing site of the 2011 Mars Science Laboratory mission, formed in the Late Noachian. Received 1 March 2012 It is a 150 km diameter complex impact structure with a central mound (Mount Sharp), the original Received in revised form features of which may be transitional between a central peak and peak ring impact structure. The 1 May 2012 impact might have melted portions of the substrate to a maximum depth of 17 km and produced a Accepted 2 May 2012 minimum of 3600 km3 of impact melt, half of which likely remained within the crater. The bulk of this Available online 6 June 2012 impact melt would have pooled in an annular depression surrounding the central uplift, creating an Keywords: impact melt pool as thick as 0.5–1 km. The ejecta blanket surrounding Gale may have been as thick as Gale Crater 600 m, which has implications for the amount of erosion that has occurred since Gale Crater formed. Impact-processes After the impact, a hydrothermal system may have been active for several hundred thousand years and Hydrothermal a crater lake with associated sediments is likely to have formed. The hydrothermal system, and Phyllosilicates Astrobiology associated lakes and springs, likely caused mineral alteration and precipitation. In the presence of S-rich host rocks, the alteration phases are modelled to contain sheet silicates, quartz, sulphates, and sulphides. Modelled alteration assemblages may be more complex if groundwater interaction persisted after initial alteration. The warm-water environment might have provided conditions supportive of life. Deep fractures would have allowed for hydraulic connectivity into the deep subsurface, where biotic chemistry (and possibly other evidence of life) may be preserved. & 2012 Elsevier Ltd. All rights reserved.

1. Introduction n Corresponding author at: Lunar and Planetary Institute, USRA, 3600 Bay Area Blvd., Houston, TX 77058, USA. Tel.: þ1 44 1908 659987; Gale Crater, target of the Mars Science Laboratory (MSL) fax: þ1 44 1908 655151. mission, has significant potential for elucidating impact processes E-mail addresses: [email protected], [email protected] (S.P. Schwenzer), carlton.c.allen@nasa.gov (C.C. Allen), and products on Mars, including hydrothermal systems that [email protected] (J.C. Bridges), [email protected] (S.M. Clifford), might have supported life. Gale is a 150-km diameter impact [email protected] (J. Filiberto), [email protected] (D.A. Kring), structure, formed during the late Noachian, at the border of the [email protected] (J. Lasue), [email protected] (P.J. McGovern), southern highlands and Elysium Planitia (4.491S, 137.421E). Some [email protected] (H.E. Newsom), [email protected] (A.H. Treiman), [email protected] (D.T. Vaniman), [email protected] (R.C. Wiens), aspects of Gale’s geology have been studied extensively, but not [email protected] (A. Wittmann). its impact formation and the era shortly after its formation, which 1 Tel.: þ1 281 486 2146. we will investigate here. 2 Tel.: þ1 618 453 4849. From morphology and geology, Gale is known to include four 3 Tel.: þ1 281 486 2119. major geologic units: the ejecta blanket, a raised rim, a flat floor, 4 Tel.: þ1 281 486 2187. 5 Tel.: þ1 281 486 2117; fax: þ1 281 486 2162. and a central mound, which consists of sediments and potentially 6 Tel.: þ314 9356151. an underlying central peak (Greeley and Guest, 1987; Pelkey and

0032-0633/$ - see front matter & 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.pss.2012.05.014 S.P. Schwenzer et al. / Planetary and Space Science 70 (2012) 84–95 85

Jakosky, 2002; Cabrol et al., 1999; Malin and Edgett, 2000; Edgett Gale’s central mound, recently named Mount Sharp (NASA, and Malin, 2001; Anderson and Bell, 2010; Milliken et al., 2010; 2012), is a particular target of interest—a 6 km-thick pile of Thomson et al., 2011; Le Deit et al., 2011). These units have sediments of various morphologies and compositions. The mound suffered erosion and are locally buried in aeolian ‘dust’. However, sediments have been interpreted as late Noachian to early the erosion has not been uniform across the crater; the northern Hesperian in age, and are divided by an angular unconformity rim (adjacent to Elysium) is both lower and less steep than the into two groups (Thomson et al., 2011). The upper group forms rest of the crater rim (Fig. 1). Some parts of the ejecta blanket rounded-hillslopes and its IR reflectance spectra (from CRISM, on have been preserved, especially in the highlands to the south the MRO spacecraft) are dominated by sulphate signatures (Fig. 1, Greeley and Guest, 1987). (Thomson et al., 2008a; Milliken et al., 2010). Channels, which The MSL landing site is located inside the crater (Golombek et al., appear to be sourced from the upper section of the layered mound 2011), on the NNW quadrangle part of the flat floor between the material, cut into the underlying sediments and form depositional central mound and the rim. The floor is covered by relatively little fans on lower slopes (Thomson et al., 2008b). The lower group is dust, and can be divided into three distinct units based on thermal strongly layered, and many layers have IR reflectance spectra inertia, which are interpreted as impact melt, volcanic rocks, and characteristic of hydrous phyllosilicates (Milliken et al., 2010). indurated sediment (Pelkey and Jakosky, 2002; Pelkey et al., 2004). Therefore, the lower phyllosilicate bearing strata appear to be Numerous small craters ranging from 0.4orlessto3.5kmin formed by a different process than the overlying sediments diameter have excavated the local subsurface (Pelkey et al., 2004). (Milliken et al., 2010; Thomson et al., 2011).

Fig. 1. Maps of Gale Crater. (a) topography, highest point of the central mound (triangle on mound) at 788 m; lowest point on the crater floor (circle to the NW of mound) at -4649 m, but probably located in a small crater; highest point on the crater rim at 1413 m (star to S of central mound); lowest point on the crater rim at 3300 m (square to NW of mound) (b) slope steepness (c) map of the directly impact-related geologic units (blue¼country rock and ejecta, orange¼crater wall and slump deposits, grey¼moat zone deposits, yellow¼central mound; not shown are any younger sediments), and (d) small craters and their potential target lithologies. Black, thick outline in panel c is 150 km diameter Gale Crater (thick black line is the crater rim). Smaller craters above 3/4 km diameter are mapped on the basis of visible imagery. Dark green line is outline of the central mound. Light green cross is the location and approximate size of the proposed MSL landing ellipse as per March 2011 (Caltech 2011). The small craters are classified for the rocks they potentially expose using the map by Greeley and Guest (1987) and the investigations by Pelkey and Jakosky (2002) and Pelkey et al. (2004). Colour legend: black thin lines: unclassified. blue: ejecta blanket, red: rim area, yellow: sediments of the crater floor, grey: impact-melt sheet, violet: crater mound. Note the variety of rocks that might be found in their walls and ejecta. 86 S.P. Schwenzer et al. / Planetary and Space Science 70 (2012) 84–95

The existence of mid-mound channels and fan deposits on the to the currently favoured view of sedimentary deposition of these lower slopes has been interpreted as evidence for a lake in the minerals as result of the overall Martian climatic environment youngest preserved history of Gale (Cabrol et al., 1999). Therefore (Milliken et al., 2010; Le Deit et al., 2011; Ehlmann et al., 2011). the early sedimentary environment may have been followed by later episodes of inundation related to Late Amazonian flooding of the Elysium basin (e.g., Cabrol et al., 1999; Clifford and Parker, 3. Topography and impact formation of Gale Crater 2001)). Thus, Gale Crater may have episodically hosted standing bodies of water over a period of almost two billion years (Cabrol 3.1. General topography and target description et al., 1999), resulting in the deposition of sediments concurrent with the presence of potentially habitable environments during Gale is a 150-km diameter impact structure, at 4.491S, the crater’s multi-stage geologic history. 137.421E(Fig. 1a), on the slope of the highland-lowland dichot- Although this sedimentary history has been investigated inten- omy boundary. Our topographical map shows that regional slopes sely, Gale’s impact and cratering history (original and subsequent) in this area are shallow (11, Fig. 1b), and steeper slopes are is poorly known. Here, we explore the impact history of Gale found mostly on crater rims and on Gale’s central mound. Gale Crater in terms of its formation, its post-impact hydrothermal Crater itself is complex, with a generally steep rim and the likely system, and its post-impact habitability. First, we calculate Gale’s remnant of a central peak within the central mound (Fig. 2). This impact-generated form and materials (e.g., height of central uplift, remnant is close to the crater’s geographic centre, and so probably volume of impact melt). Then we describe important character- represents a true central peak rather than a fragment of a peak istics relevant to the presence and abundance of aqueous and ring. The northern third of the crater interior is occupied by a hydrothermal environments due to the impact-deposited heat. mound of sedimentary strata, which has been intensely investi- These data lead to a consideration of the hydrothermal aftermath gated (Fig. 1c; e.g., Milliken et al., 2010; Thomson et al., 2011). of the Gale impact and its likely mineralogical record. The mound is clearly an erosional remnant of a more extensive sediment pile; its original extent is not clear, but may have entirely filled Gale. The lowermost strata of the mound are 2. Methods morphologically and mineralogically distinct from the thick sequence that makes up most of the visible slopes. To describe the crater formation and hydrothermal and A crater’s depth-diameter ratio is an important constraint on hydrous aftermath of the energy released at the Gale impact site, its formation and subsequent history, but this parameter is we utilise a variety of methods. To arrive at the impact-generated ambiguous for Gale. First, between Gale’s rim and its central factors (e.g., height of central uplift, volume of impact melt) we mound is a ‘moat’ much deeper in the north than the south. The use a set of standard equations from the literature, which we cite northern part of the moat is partially filled with sediments, as we work through the individual parts of the crater (Section 3: including layered deposits beneath the central mound material; Topography and impact formation of Gale cater). We next apply debris flows and aqueous deposits off of the crater rim, and the model for hydrothermal circulation of Abramov and Kring deposits and possibly a delta shed off the central mound (Cabrol (2005) to Gale Crater and compare it to ice-bearing cases et al., 1999; Pelkey and Jakosky, 2002; Pelkey at al. 2004). The (Barnhart et al., 2010, Ivanov and Pierazzo, 2011) to understand south and southeast sides of the moat are partially filled by the plumbing of a likely hydrothermal system underneath the massive lobes of landslide material, shed off the nearby crater newly formed, late Noachian Gale Crater in light of potential rim. Second, Gale was emplaced onto a regional slope, so that the climate scenarios (Section 4: Post impact hydrology). southern rim is higher than the northern. Third, Gale’s ejecta is While the above steps are applications of literature data and preserved only on its southern rim, suggesting that its northern models, we next evaluate the mineralogical aftermath of impact- rim and adjacent flank of the crater has experienced some generated hydrothermal activity and the more general post- erosion. Taking all this into account, the maximum topographic impact hydrologic system. To assess the mineralogical effects of relief we find at Gale is6 km, but the maximum depth from a hydrothermal circulation in Martian target rocks, we have pre- point on its rim to the nearest moat bottom is only4.5 km, from viously used CHILLER (Reed and Spycher, 2006) to calculate which we calculate a depth/diameter of 0.03. This value is mineral formation in lherzolitic (Schwenzer and Kring, 2009) consistent with the general morphology of Martian craters and basaltic (Schwenzer and Kring, 2009a) target rocks, and in (Garvin et al., 2000), and suggests that Gale’s Crater rim has not the presence of CO2 (Schwenzer and Kring, 2009b). We have been lowered much by erosion, and that Gale’s moat is underlain further modelled two special cases: Gusev basalts at the Spirit by only a small amount of post-cratering infill (perhapso1 km?). landing site (Filiberto and Schwenzer 2010) and the nakhlite fluid Gale Crater formed in the late Noachian period, as implied by the (Bridges and Schwenzer, 2012). Those previous models, however, abundance of later craters within the crater and on its ejecta were not especially concerned with S-species. Because Gale con- (Thomson et al., 2011). On Fig. 1d, those superposed craters are tains sulphates, we present more S-rich models in this paper. For colour-coded according to the geological units of Fig. 1candthe Gale’s central mound Thomson et al. (2011) describe a succession expected material they would excavate to the surface (Table 1). Note, of sulphate-bearing rocks (with possibly some phyllosilicates) at however, that we did not attempt an exact crater count but illustrate the very bottom of the sequence, followed by a ‘‘a thin but distinct the general crater features and potential access to impact-generated smectite-bearing unit’’, then mixed smectite with olivine, then material. In the crater walls, strata of the target rock are exposed and mixed sulphate and smectite and finally sulphate-bearing layers, give potential access to pre-impact material and to impact-generated based on analysis of the CRISM images of Gale. With S-bearing rocks and alteration minerals. Rocks of different target depths could models we show how phyllosilicates and sulphates can result be observed in the crater wall and the sedimentary slump and slide from hydrothermal alteration. Details of the modelling approach deposits adjacent to them (orange area in Fig. 1c). Recent HiRISE are incorporated into Section 4. We also discuss experimental data images (ESP_023034_1755, ESP_022467_1760) of the crater wall bearing on reactions between sulphate brine and phyllosilicates closest to the landing site ellipse may show evidence of megabreccia involving cation exchange and physical as well as mineralogical deposits, which might contain evidence for impact-generated altera- transformations. With this we describe a formation mechanism for tion. Through erosional transport processes they likely have con- the hydrous minerals found in Gale Crater that is complementary tributed to the materials in the ellipse. S.P. Schwenzer et al. / Planetary and Space Science 70 (2012) 84–95 87

Fig. 2. HiRISE and CRISM imagery taken in the southern part of the central mound. (a) topographic overview (FRT00011F99_07_IF164L_TOPO; CRISM, 2010a); red box indicates the position of the HiRISE image. (b) image FRT00011F99_07_IF164L_IRA1 (CRISM, 2010b) that shows the blocky nature of the terrain, which could be megabreccia on the south flank of the central uplift. (c) accompanying CRISM observation for the phyllosilicate bands (FRT00011F99_07_IF164L_PHY1; CRISM, 2010b). Note that phyllosilicate signatures occur where the slopes are steepest, i.e., at the edges of large blocks. Those edges are prominent, because the impact-generated fractures might have been eroded more intensely. Those fractures are also the sites where hydrothermal water flow would have been most intense and therefore where erosion is most likely to expose altered target rock.

Table 1 ellipse (Fig. 1). They may excavate sediments, melt breccias and/or Small craters above 3/4 km diameter puncturing Gale Crater classified by their melt sheet deposits. To the WSW of the landing ellipse, an potential target lithologies. See also Fig. 1. impactor hit a collapse zone in the inner crater wall. The terrace zones of large craters are characterised by deeply-penetrating Target lithology Number of craters normal faults that accommodate the inward and downward Ejecta blanket 67 collapse of the transient crater and mark the transition of the Ejecta blanket and rim strata 2 crater environment to undisturbed crust beyond the final crater Rim strata 16 rim (e.g., Melosh, 1989; Morgan and Warner, 1999). In large Crater floor infill and rim strata 1 impact craters, such as Gale, these fracture zones may serve as Crater floor infill and impact-melt 10 Crater mound material 2 conduits for impact-generated hydrothermal fluids that would Unclassified crater in the frame 20 cause localised areas of mineral alteration (e.g., Osinski et al., Total 118 2001; Osinski et al., in review). For this reason, there will likely be some evidence for hydrothermal activity at Gale. The extent of the alteration is dependent on the amount of energy deposited as heat The crater moat (grey area in Fig. 1c) should contain impact in the target - and the amount of water present. The amount of melt and melt breccia, which were covered by later sedimentary heat energy generated by the original impact can be estimated by deposits. Small craters excavating this area could potentially calculating the transient crater, melt sheet thickness and amount provide insights into the lake and sedimentary history; some of central uplift from the observed diameter of Gale Crater. might have reached the underlying impact strata. These smaller craters excavate the top of the central mound, floor material, rim 3.2. The initial crater: transient crater dimension and depth of and inner crater wall (terrace zone), and the ejecta blanket excavation (Table 1), giving ample opportunity for exposures into those materials. Given its large diameter, Gale Crater is expected to have In the landing ellipse and its immediate surroundings, three formed as a complex crater, possibly with a peak ring. Here we small impact craters are located within a 20 km radius outside the estimate its initial features, i.e., diameter of the transient cavity, 88 S.P. Schwenzer et al. / Planetary and Space Science 70 (2012) 84–95 ejecta thickness, height of the central uplift, and the amount of which case the amount of impact melt may be smaller (Melosh, melt formed. We begin with the semi-empirical model of Croft 1989; Pierazzo and Melosh, 2000, Abramov et al., 2012).

(1985) to calculate the transient cavity diameter, DTC, for an This melt sheet would not have a uniform thickness across the observed crater with a final rim diameter D: crater interior, because rebound of the crater floor and emergence

7 of the central uplift will lead to only a thin veneer of impact melt D ¼ðDnÞ0:15 7 0:04 D0:85 0:04 ð1Þ TC (if any) in the centre of the crater and the pooling of the where all variables are in kilometres, and Dn is the transition remaining impact melt in an annular moat surrounding the 3 diameter between simple and complex craters. On Earth, this central uplift. Nonetheless, the minimum estimate of 1800 km transition diameter varies from 2 km for impacts into sedimen- impact melt that remained in Gale Crater is similar to estimates tary lithologies to 4 km for impacts into crystalline lithologies for the 100 km diameter Popigai crater in Siberia, which retained 3 (Grieve et al., 1977; Croft, 1985). An average value of 3.1 km 1750 km of impact melt in deposits 4600 m thick (Masaitis (Pike, 1988) is often assumed. For other terrestrial planets Dn et al., 1999). appears to vary as the inverse of the surface gravity. Craters on n þ 3:1 Mars have a transition diameter of D ¼ 5:11:9 km (Pike 1988). 3.4. Depth of melting and stratigraphic uplift Substituting this value in Eq. (1), along with a final crater diameter D¼150 km for Gale Crater, we calculate a corresponding Another important characteristic of an impact is the maximum þ 37 transient cavity diameter DTC ¼ 9025 km. The upper limit for the depth of melting (dm) in response to the passage of the shock n Martian value of D is 8.2 km, which is similar to that derived wave. For Gale Crater we estimate dm to be between 17 km as for from Pi-scaling (8.4 km, Holsapple and Schmidt, 1982), the an equivalent lunar crater (depth of melting/depth of transient latter of which implies DTC 97 km. cavity ratio of 0.55 for a transient crater diameter of 90 km, Kring (1995) proposes a relationship between the radius of a Cintala and Grieve 1998) and 30 km for an equivalent terrestrial complex crater, RC, and the radius of the transient cavity, RTC: crater (depth of melting/ depth of transient cavity ratio of 1 for a transient crater diameter of 90 km, Grieve and Cintala, 1992), R ¼ 0:86 R1:07 ð2Þ C TC consistent with case studies at Chicxulub Crater, where the where all units are in metres. Based on this relationship, the transient cavity was comparable in size (Pierazzo et al., 1998; transient crater diameter for a 150 km diameter complex crater is Kring, 2005). estimated to be 83 km. All these estimates agree within their The scaling relationship of Grieve et al. (1981) can be used to respective error limits. So, we have adopted an intermediate value estimate the amount of stratigraphic uplift: of 90 km as transient crater diameter for Gale. For comparison, on 1:1 hsu ¼ 0:06 D ð4Þ Earth the transient crater diameter of the 180 km diameter Chicxulub Crater was 100 km (Kring, 1995; Morgan et al., which is 15 km for Gale Crater consistent with estimates made 1997). using other approaches (e.g., see Grieve, 2006, p. 6). Interestingly, Assuming that the depth of the transient cavity scales to 1/3 of comparison of the estimates of Gale’s depth of melting with the its diameter (Cintala and Grieve, 1998) and the excavation depth amount of stratigraphic uplift of its central uplift indicates that a is one-third to one-half of this transient depth (Melosh, 1989), significant portion of the central uplift consisted of impact melt material from a depth of 10 to 15 km was excavated and and could have collapsed to form a peak ring. However, Gale, and deposited as ejecta during the formation of Gale. For a complex a number of similarly sized craters on Mars about 100 km in crater, the maximum thickness of the ejecta deposits can be diameter, have obvious central mounds or uplifts. estimated according to a scaling relationship defined by McGetchin et al. (1973); see also Kring (1995): 4. Post-impact hydrology 0:74 3:0 7 0:5 de ¼ 0:14 R ðr=RÞ , ð3Þ 4.1. Water flow and temperature evolution in impact-generated where d is the ejecta thickness, R is the radius of a complex e hydrothermal systems crater, r is the distance from the centre of the crater, and all variables are in metres. According to this relationship, the max- Impact-deposited heat at Gale, concentrated in the melt-sheet imum thickness of ejecta on the rim of Gale Crater was 570 m. and central uplift, would have been capable of generating intense hydrothermal activity in the target site. Hydrothermal systems 3.3. Volume and distribution of impact melt were generated at many large impact craters and basins on Earth (e.g., Boltysh, Chicxulub, Manicouagan, Rochechouart, and Sud- Utilising the average impact velocity on Mars of 10 km/s bury; Ames et al., 2006; Hecht et al., 2004; Lambert, 2010; (Hartmann, 1977) and using scaling relationships for Earth Newsom et al., 2010; Spray et al., 2010; Watson et al., 2010; (Grieve and Cintala, 1992), and the Moon (Cintala and Grieve, Zurcher¨ and Kring, 2004), and there is no reason to expect a 1998), the amount of melt produced in Gale can be constrained. different outcome for Mars, provided water was present in the Taking the terrestrial case as upper limit and the lunar case as subsurface. Similarly, numerical simulations show that hydro- lower limit, between 3600 and 21000 km3 of melt were produced. thermal convection systems would have been generated around This range agrees well with the 5300 km3 estimated for the Gale Martian impact craters in clement and frozen conditions impact by the melt scaling expression of Abramov et al. (2012). (Newsom, 1980; Rathbun and Squyres, 2002; Abramov and Between 35 and 60% of this melt would have been ejected from Kring, 2005; Barnhart et al., 2010; Ivanov and Pierazzo, 2011). the transient cavity (Cintala and Grieve, 1998); assuming a typical As an example, consider an impact into Mars that forms a average of 50% ejected impact melt, 1800 to 10500 km3 of impact central-peak crater of 100 km diameter. Rocks in and near the melt remained inside Gale. Integrating this volume across the central peak would be heated to 1170 K, with temperature area of the transient cavity by assuming all impact melt produced declining away from the centre (Abramov and Kring, 2005). This but not ejected remains in the horizontal limits of the transient heat will produce a hydrothermal convection system (provided cavity, an average melt sheet thickness of 0.3 to 1.7 km results. water is present), in which groundwater would be drawn toward Our calculations are not applicable for highly oblique impacts, in the central peak, channelled by fractures; and discharged as hot S.P. Schwenzer et al. / Planetary and Space Science 70 (2012) 84–95 89 water or steam. The hot melt sheet might not be permeable, so and volcanic-hydrothermal processes (e.g., Murchie et al., that discharge would be concentrated in the central peak and the 2009a,b), but also impact-generated hydrothermal alteration crater rim area. Much of this discharge would end up in the (e.g., Marzo et al., 2010). Alteration mechanisms include dissolu- crater’s moat, forming a crater lake, before evaporating or tion or modification of the magmatic minerals, precipitation of infiltrating back into the ground. With time, this convective new mineral assemblages, and transport of some of the most system would cool the rocks of the crater rim and central uplift, soluble chemical species away from the system. causing a diversity of water flow regimes – such as development Detailed analysis of Martian impact craters has demonstrated of multiple convective cells between the innermost and outer- that post-impact hydrothermal activity can cause venting and most parts of the crater. Further, the solidifying melt sheet would hydrous alteration in central uplift settings (e.g., Marzo et al., crack and become permeable. Such a hydrothermal system (in a 2010; Faire´n et al., 2010). Changela and Bridges (2010) described 100 km diameter crater) would be active for 300,000 years the alteration mineralogy of the nakhlite Martian meteorites and (Abramov and Kring, 2005), and its total water discharge would showed that it consisted largely of a siderite-smectite-amorphous be 3 103 km3 (mostly recirculated). The above values are silicate gel assemblage within fractures. They suggested that this calculated for non-freezing conditions. Lower Martian surface hydrothermal assemblage formed though impact-induced fluid temperatures and an ice instead of water-bearing target will activity. This hydrothermal activity in Martian craters and nakh- shorten the lifetime of the system and might concentrate water lite meteorites are associated with fluids that carried a variety of flow towards the central uplift (Barnhart et al., 2010; Ivanov and ions and so were capable of precipitating a range of minerals, Pierazzo, 2011). including silica and salts. (Schwenzer and Kring, 2010; Bridges and Schwenzer, 2012). Multiple origins for alteration minerals 4.2. Impact-generated hydrothermal mineralogy may be common in diverse crater settings (Osinski et al., in review). Knowledge of mineral assemblages, rather than just Hydrous minerals (phyllosilicates and sulphates, e.g., Thomson single-mineral occurrences at the scale of instrument footprints et al., 2011) have been detected in the lower strata of the Gale from orbit, will constrain the formation conditions of alteration. Crater mound. Interpreting the origin and nature of these miner- At the time of writing (January 2012), the MER rover Opportunity als on the basis of the available orbiter data can be supported by has reached Endeavour crater. Just before reaching the winter thermochemical modelling. For this there is a need to understand parking position it – for the first time ever – carried rover-based the geologic context and the pre-hydrous mineralogy first, but exploration into an area where orbiter-based observation has many important details will only become accessible from on-site detected phyllosilicate signatures. But Curiosity remains the first rover based investigation. We therefore take a more general rover for which orbiter-observed phyllosilicates have been part of approach based on the potential host lithology and its alteration. the mission planning from a pre-launch stage. New models were carried out on sulphur-rich soils measured by In Gale Crater, spectral evidence for alteration phases is most the MER rover Spirit. With this we lay the foundation for prominent in the lowest part of the central mound. They are interpreting MSL results from the first measurements of host interpreted as part of the stratigraphic sequence of Gale Crater rock and alteration phases and following along as the mission and as important provenance and climate indicators. Stratified progresses. deposits rich in phyllosilicates and sulphates are common on the lower northern flanks of the mound. These are traceable over long distances and have been interpreted to be sedimentary in origin 4.2.1. Composition of the Martian crust (Milliken et al., 2010) with post-impact paleoclimatic significance. The need to constrain geologic context is especially evident However, some phyllosilicates occur in the southern flank of the when trying to combine all of the available data on the Martian central mound (Fig. 2). The hydrous alteration phases occur in the crust. The igneous chemistry and mineralogy obtained from rover lowest stratigraphic units and the exact timing of sequences is observations, orbital spectroscopy, and Martian meteorites have currently unknown (Thomson et al., 2011), but we propose that they significant differences (Filiberto et al., 2006; McSween et al., could be potentially impact-related. This could occur in two ways: 2009; Schmidt and McCoy, 2010; Filiberto, 2011). In fact, the the alteration could be located in uplifted target material that might MER missions have demonstrated that within one large crater a form the lowest strata or they could be formed by alteration of early variety of different rocks can be sampled (e.g., Gellert et al., 2006; sediments. In terrestrial examples (Chicxulub, Boltysh; Zurcher¨ and Ming et al., 2008; Schmidt and McCoy, 2010). This implies that Kring 2004, Watson et al., 2010) post-impact hydrothermal systems the composition of the crust is laterally and stratigraphically have affected overlying, post-impact sedimentary deposits. It is, heterogeneous and may reflect igneous systems, i.e., volcanic and therefore, important to consider hydrothermal alteration of target plutonic rocks including cumulate rocks (Treiman, 2005; Filiberto rocks, impact-lithologies and early post-impact sediments as a et al., 2006; McSween et al., 2009), that changed in space and possible formation process for the hydrous alteration phases. time, but unless detailed information is available from Gale itself the entire variety has to be considered as potential target rock chemistry. We therefore choose rock compositions from the MER 4.2.3. Modelling alteration mineralogy from the parameters mission as proxy for our models (see modelling section in Section available - and outlook to new data 4.2.3). The minerals formed by impact-generated alteration are diverse, mostly of lower temperature (up to 150 1C) origin and 4.2.2. Compositional constraints for alteration assemblages can be further investigated with thermochemical modelling. The The original igneous diversity of Martian rocks is overprinted hydrothermal and aqueous conditions of an impact site evolve by hydrothermal and other alteration (e.g, Bibring et al., 2005; from initially high temperatures in the crater centre with steep Ehlmann et al., 2011; Mustard et al., 2008). While hydrous sheet decline outwards to lower and more uniform temperatures later in silicates such as chlorite and nontronite are detected in several the lifetime of the system (Abramov and Kring, 2005). The chan- Noachian terrains, they are scarce or absent in younger terrains ging temperatures – combined with the declining vigour of water (Bibring et al., 2005, 2006; Mustard et al., 2008). Their ubiquity in circulation – change the thermochemical environment. For host the Noachian suggests that they may have been formed by a rocks of lherzolite composition (Martian meteorite LEW88516), variety of water-driven geologic processes such as weathering Schwenzer and Kring (2009) showed that at 150 1Cserpentineand 90 S.P. Schwenzer et al. / Planetary and Space Science 70 (2012) 84–95 chlorite will be most abundant where the local permeability is low, and that nontronite will be dominant in the more permeable host rocks. If the host rock is more feldspar rich, e.g., the basaltic shergottite Dhofar 387, the resulting mineral assemblage will contain secondary feldspar as well as zeolites (Schwenzer and Kring, 2009c). If the protolith is a dunite, a serpentine-magnetite assemblage results

(Schwenzer and Kring, 2009c). High amounts of CO2 lead to carbonate formation at the expense of hydrous silicates (Schwenzer and Kring, 2009b; Bridges and Schwenzer, 2012), and cooling and discharge of the alteration brine can lead to silica deposits (Filiberto and Schwenzer 2009, Bridges and Schwenzer, 2012). In Gale Crater, sulphates have been observed, for whose formation we provide a first set of model data by using two soils measured by the MER Spirit at Gusev (Gellert et al., 2006). As described in the methods section, we use CHILLER (Reed and Spycher, 2006) to model alteration mineral assemblages resulting from the alteration of the soil compositions Peace and Paso Roble in contact with a dilute brine at 150 1C. The starting rock and water compositions are summarised in Table 2. ThecalculationsindicatethatadditionofSdoesnothavea major influence on the alteration assemblages at very high water to rock ratios (W/R) compared to the S-poor system, here represented by our calculations on LEW88516 (Schwenzer and Kring, 2009). In the very diluted system two factors dominate the alteration assemblage: the composition of the initial fluid and the fact that Al and Fe are the ions forming the least soluble salts. Both S-rich compositions form an initial alteration assemblage with 480% haematite accompanied by diaspore and pyrite (Fig. 3 aandb). At intermediate to low W/R the dissolved rock composition dominates the nature of the alteration assemblage. The most significant differences from the S-poor system are the formation of pyrite at the expense of haematite and iron bearing hydrous silicates and the formation of anhydrite instead of Ca-bearing silicates in the S-rich system (Fig. 3 a and b). This, in turn, reduces the amount of SiO2 bound in the silicates, causing quartz or Fig. 3. Model results of alteration mineral assemblages in dependence of water to amorphous SiO2 to precipitate. Moreover, the total amount of rock ratio (W/R). (a) host rock ‘Peace’, and (b) host rock ‘Paso Roble’. Peace host alteration products from the S-rich systems is less than from the rock chemistry froms up to 10% sulphates, while the more sulphur rich S-poor systems. Therefore, more ions remain in solution: composition Paso Roble forms up to 25% sulphates. Both compositions as measured by the MER Spirit (Gellert et al., 2006), models for 110 bar and 150 1C. 0.04 mol/kg for LEW88516 at W/R¼1 and 0.07 and 1.62 mol/kg for Peace and Paso Roble, respectively. As a consequence, a more concentrated brine is available for later reaction or deposition. evolution of such systems it is expected that those mineral The calculations were done for 150 1C, where the stable phases assemblages are subject to changes by the cooling fluid during are anhydrite and pyrite. From the temperature and water flow the waning stages of the hydrothermal activity. For example, anhydrite is expected to become hydrated with temperature decrease (e.g., Christensen et al., 2008) and pyrite might be Table 2 oxidised if conditions change. Starting conditions for CHILLER models of Peace and Paso Roble compositions as The above results demonstrate how sulphates and hydrous sheet measured by MER Spirit (Gellert et al., 2006). Choice was for soils with high SO2- silicates can be formed directly by hydrothermal alteration. The content, especially Paso Roble, which has the highest SO2 content measured. Water composition deduced from Deccan trap fluids (Minissale et al., 2000) and accuracy of the models is dependent on the knowledge of the altered adjusted to the Fe/Ca and Fe/Mg ratio of Martian rocks. host rock and the alteration mineral assemblage. If that is known in greater detail, e.g., in our study of the nakhlite alteration (Bridges and Rock Water Schwenzer, 2012), a more precise adjustment of modelling input Peace Paso Roble data will allow us to iteratively bracket the formation conditions of the observed assemblage. With the MSL instrument suite it will be Component Concentration concentration Ion Concentration possible to obtain unaltered host rock chemistry of alteration- [wt-%] [wt-%] [moles] precursor rocks, to determine alteration assemblages and succes- 3þ 3 sions, and to develop an understanding of the geologic context of SiO2 37.30 24.90 Fe 9.2 10 2þ 3 Al2O3 2.24 6.27 Mg 20.5 10 their occurrence. In order to do so, alteration of the initial set of in- FeO 20.40 16.10 Ca2þ 2.5 103 situ or deposited hydrous phases has to be considered. 6 MgO 21.53 5.19 HCO3 16.8 10 2 3 CaO 4.90 6.94 SO4 2.9 10 3 Na2O n.d. 1.40 Cl 58.7 10

K2O n.d. 0.40 4.2.4. Phyllosilicates in contact with S-bearing solutions - TiO2 0.45 0.88 experimental data MnO 0.47 0.03 Experiments with silicate phases susceptible to cation SO 8.48 25.29 2 exchange show that interaction with groundwater after their P2O5 0.49 0.05 initial formation can alter both their composition and that of S.P. Schwenzer et al. / Planetary and Space Science 70 (2012) 84–95 91 the surrounding solution. In Vaniman et al. (2004) and Vaniman Boltysh and Chicxulub Craters (Jolley et al., 2010; Watson et al., and Chipera (2006) it was found that Ca-bearing smectites, 2010,Zurcher¨ and Kring 2004). It is also possible that hydro- palagonite (JSC Mars-1, Allen et al., 1998), and zeolites (clinopti- thermal springs at the lake bottom could generate their own lolite or chabazite) exposed to Mg-sulphate brine all exchange Ca chemical sediments, like the ‘black smoker’ and ‘white smoker’ for Mg in solution to form a more Mg-rich silicate phase and hot springs in the Earth’s oceans. produce a mixed-cation brine that precipitates Ca-sulphate as well as Mg-sulphate salts. Although the smectites used in these 4.4. Effects of impact-generated hydrothermal activity on experiments contained less Ca than either the palagonite or the habitability zeolites, the extent of cation exchange was greater in smectite- brine association than in either the palagonite or the zeolites. A hydrothermal system of such long duration and large Later experiments with nontronite and a range of Mg-sulphate volume of water, e.g., for a 100 km diameter crater 300,000 brine concentrations (Vaniman et al., 2011) indicate that at 293 K years and total water discharge of 3 103 km3 (see 4.1), would and a brine molality of 0.5, about 50% of the interlayer Ca in the be attractive for colonisation by thermophilic and/or hyperther- nontronite is exchanged for Mg from solution. Even at much mophilic microorganisms (e.g., Abramov and Kring, 2005; Osinski lower brine concentration (0.05 mol MgSO4), the exchange is still et al., in review). On Mars, however, the question is if there was substantial (25%). any biologic activity. And if there was, the abundance of organic The implication of these experiments is that reactive transport or biological ‘‘contaminated’’ locations might be much rarer and should be considered in evaluating deposits that contain phyllo- transfer between them difficult (Cockell et al., 2012). But if there silicates (especially smectites), zeolites, or incipiently altered ever was thermophilic activity in the post-impact hydrothermal volcanic glass along with salt hydrates. Where such hydrous system its unique signatures may be more readily defined than on silicates and salt hydrates occur together, the chemical composi- the biologically more active Earth. tion and mineral products, if a closed system with a single cycle of Mars’ climate changed to cold and dry conditions around the evaporation, will not necessarily be the same as those where time Gale Crater was formed. This makes impact sites important initial formation was followed by exposure to groundwater. This places for survival, if life had existed at that time. Impact- adds some complication in the interpretation of such assem- generated hydrothermal systems are likely to have provided blages, but also can be used to advantage in interpreting the refugia of liquid water in a progressively icy environment, with history of water-rock interaction at a site. Multiple chemical and a broad range of long-lived aqueous and thermal environments mineral analyses within a deposit may provide a map of chemical that were potentially linked by deep aquifers and evolved slowly and mineral variation that bounds the extent of groundwater with time (Newsom, 1980; Newsom et al., 1996; Lindsay and interaction and can in some cases define flowpaths. In massive Braiser, 2006; Cockell et al., 2005). The mineral alteration assem- deposits such interpretations may be poorly constrained, but blages produced in these environments included sheet silicates stratified deposits (e.g., Vaniman et al., 2001) can provide a (see Section 4.2.3) whose surfaces may have helped catalyse the framework for such interpretations. For this type of analysis the first steps towards life (Brack, 2006). Mineral reactions and the stratified central mound at Gale Crater is a promising target. hydrothermal circulation of groundwater may have also delivered nutrients and energy to support growth of an initial biomass 4.3. Impact crater lake formation and evolution (Varnes et al., 2003). Moreover, craters that contained impact-driven hydrothermal Besides the underground activity, impact craters create topo- systems offered the widest range of habitable conditions in any graphic lows in a planet’s surface, and so readily develop crater single early Martian environment. In addition, sediments, espe- lakes by flow of surface and/or groundwater (Newsom et al., cially those containing clay minerals produced by the hydrother- 1996; Cabrol and Grin, 1999; Cabrol et al., 2001; Newsom, 2010). mal alteration, are promising targets that may preserve evidence In a crater like Gale, the water of an early crater lake would be of past habitability and even of life itself. The sedimentary heated significantly by the impact melt sheet and the shock- geology of Gale Crater has been well studied from orbit and was heated and uplifted basement (McKay and Davis, 1991; Newsom an important consideration in landing site selection (e.g., Grant et al., 1996). Surrounded by so much hot rock, and especially if et al., in; Marsoweb, 2010; Edgett, 2010; Bell et al., 2010). capped by a self-sealing ice cover, lakes in large impact craters Hydrothermal alteration seen in ALH84001 (Kring et al., 1998) can remain liquid for thousands to hundreds of thousands of and the nakhlite Martian meteorites (Bridges et al., 2001; years, maintaining connections to deep aquifers (if present Treiman, 2005; Changela and Bridges (2010)) shows a subset of beneath the Noachian cryosphere). the widespread secondary mineralogy, i.e., smectites and carbo- A lake acts as a sink for sediments, and the Gale Crater Lake nates, observed from orbit and analysed by rovers. Orbital data would not have been an exception. Given the presence of water have also demonstrated that the Martian crust is mineralogically and heat, a crater lake like this should facilitate extensive heterogeneous both laterally and at depth. Curiosity’s investiga- alteration of its rocks and sediments. These sediments would tions will allow determination of the alteration conditions, the have included material transported by both fluvial and aeolian nature of the fluids and the resulting habitability. processes; in addition, material from the inner crater wall and central peak slopes would have produced debris flow and possibly turbidite-type deposits on the crater floor. If the crater lake was 5. Gale’s central mound—an additional perspective on the covered by ice, the sediment supply could have been limited due formation of hydrous phases to reduced water flow, so the lake sediments may have been thin deposits of fine-grained material more easily subject to later Current orbiter-observation based research interprets the erosion than coarser grained material. The lowest sedimentary history of Gale’s interior in terms of sedimentary processes strata in Gale could therefore be post-impact lake strata. This (Greeley and Guest, 1987; Cabrol et al., 1999; Malin and Edgett, sediment may have been extensively altered by post-impact 2000; Edgett and Malin, 2001; Pelkey and Jakosky, 2002; Bell hydrothermal processes, including extensive cementation and et al., 2006; Anderson and Bell, 2010; Milliken et al., 2010; diagenesis. Terrestrial examples for early post-impact sediments Thomson et al., 2011; Le Deit et al., 2011). However, both the being altered by the impact-generated hydrothermal system are crater rim and the extent of the original ejecta blanket around 92 S.P. Schwenzer et al. / Planetary and Space Science 70 (2012) 84–95

Gale are still visible, which indicates that rocks excavated by the to 1/3-1/2 of its diameter (Melosh, 1989), its ejecta may contain impact should still be present in the ejecta blanket, the crater rock from 41 km depth, possibly including sediments from a wall, and the central uplift. Moreover, the fact that Gale’s Crater crater lake or impact melt rock from Gale’s formation. This ejecta rim and remnants of the ejecta blanket are still evident today should be roughly sorted, with samples expected to be sorted by suggests that either Gale was never substantially buried by such depth and with those originating from the deepest excavation depositional periods and/or that the rock sequences that predate depths being closest to the crater rim (Moore, 1977). A traverse to the sedimentation are more resistant to aeolian scouring and the rim of this crater could provide a rough stratigraphic section removal than the younger sediments that overlie them and thus of rock beneath the landing ellipse. If Gale’s impact strata could was excavated from underneath a potential sedimentary cover. be identified in this crater’s ejecta, the thickness of moat-zone The distinction between those options or any combination of sedimentary cover at the time of the 3.5-km-diameter crater’s them should become possible with ground-based investigations impact would be known. of sediments, especially in crater-rim derived fan deposits. Those Further, the landing ellipse includes several small craters that deposits might have sampled and preserved overlying strata and are a few hundred metres in diameter. Some are on, or near, the they will have collected material from all layers excavated in the ’’straw-man traverse’’ of Edgett (2010). These craters should current crater rim, providing an archive of the history of their provide access to materials just beneath the fan surface, both source region even within the landing ellipse of Curiosity. The via analyses of ejecta and via stand-off’’ analyses of outcrops on rocks of the central uplift are most likely exposed at the south- the craters’ walls. Searching for samples with the same features as western edge of the mound within the ridges and outcropping observed in the crater wall will allow correlation of wall and blocks (Fig. 2), out of Curiosity’s range unless the mission is ejecta observations before Curiosity starts its detailed chemical greatly extended. The phyllosilicate signatures in these outcrops and mineralogical analysis with instruments such as the Chem- in the VNIR spectra shown in Fig. 2 indicate hydrous alteration of Cam, ChemMin and APXS. To date, a variety of interesting targets what we infer to be remnants of a central peak. has been highlighted (see also Caltech, 2011), which include the Immediately west of the landing ellipse, a 3.5 km diameter investigation of at least one small impact crater (Target 3, Caltech, crater provides an opportunity to investigate rock beneath the 2011). This diversity of rocks has the potential to provide an landing site (Fig. 4). This crater shows a complex geology, with interpretable sequence of chemistry and mineralogy, providing scalloped walls suggestive of erosion or slump collapses, layered information critical to understand whether the alteration miner- sediments on its floor, some superposed craters, and a moderate alogy in the central mound is indigenous to Gale or an external cover of dust. Its ejecta is exposed, and excavated from beneath product deposited after transport from distant sources, as well as the dust. If this crater excavated material from a depth equivalent whether these deposits have been altered by processes such as reactive transport and cation exchange after initial formation. The Curiosity rover is intended to leave the landing ellipse, and move toward and onto the central mound of Gale Crater (Edgett, 2010; Anderson and Bell, 2010). The rationale for the proposed traverse is to sample the diverse sedimentary strata and hydrous minerals observed from orbit (Edgett, 2010). We want to add that the eroded and blocky nature of the lower part of the central mound opens the possibility of finding very early sediments, for example from a post-impact crater lake, or even impact-related strata and central-uplift rocks, potentially including pre-altered clay-bearing target rocks, and that the hydrous silicates and sulphates could also be the product of impact-generated hydro- thermal activity and possible late-stage groundwater interaction. This should be considered in addition to arguments in favour of sedimentary and climate driven origin of the observed phases. The central mound covers the area in which a central uplift formed during the crater formation. The intense deformation of the target that accommodated the rebound of the crater floor must have produced extensive fracturing and therefore pathways for hydrothermal activity. It therefore is possible that early sediments in the lowest strata of Gale were affected by an impact-generated hydrothermal system. In addition, blocky and potentially impact-brecciated material with hydrous mineral signatures has been observed in one location of Gale’s central mound, but they are on the southern mound flank outside the rover’s expected range (Fig. 2). Even though this distant site is not likely to be visited by Curiosity there remains the possibility that clasts from a hydrothermally altered central peak may be incor- porated as detritus in the mound sediments that have been mapped from orbit along Curiosity’s likely traverse from the north of the crater, across the plain and onto the sedimentary mound. We summarise the materials Curiosity might encounter Fig. 4. Part of HiRISE image ESP_011984_1755 (http://www.uahirise.org/ on its traverse in Table 3. ESP_011984_1755) showing an approximately 3.5 km diameter impact crater at The hydrological history of Gale Crater may have been diverse, the western edge of the proposed landing ellipse. Black line indicates the edge of including impact-generated water flow, surface precipitation and the landing ellipse as suggested by Edgett (2010). White arrow points towards the traverse proposed by K. Edgett, which would be in the crater’s annular moat, at a runoff as part of Gale’s early history, and formation of a crater distance of about 8.4 km inwards from the crater rim. lake. In particular the water supply from an impact-generated S.P. Schwenzer et al. / Planetary and Space Science 70 (2012) 84–95 93

Table 3 Summary of the three most important geologic steps in Gale’s history, the rocks and minerals those produce and an account, where Curiosity might find them. Note that we are excluding the crater rim and the top of the central mount, named Mount Sharp, because current plans do not include visiting those locations.

Geologic process Rocks and minerals produced Locations within current mission plans Aeolian Dust and sediments Overall dust cover and in dunes in the crater moat, pieces and particles fallen down from the top of the central mount Lake(s) Sediments and alteration/ Excavated by small craters puncturing the crater moat’s dust cover, mixed into the evaporation minerals fan deposit in the landing ellipse, lowest strata in the central mount Impact-generated hydrous Alteration minerals, early lake Transported from the crater walls and mixed into the fan deposit; excavated by activity sediments small craters from under the crater moat’s dust cover; in the lowest strata of the central mount Impact and crater formation Ejecta, impact-melts and breccias, Transported from the top of the crater rim and mixed into the fan deposit, excavated uplifted basement by small craters puncturing through the sedimentary cover in the crater moat, lowest strata of the central mount

hydrothermal system and the resulting alteration (Fig. 3) would On the basis of the current observations at least five potential have delivered hot water with dissolved ions to the surface geologic processes could have led to the detectable presence of (Schwenzer and Kring, 2010; Bridges and Schwenzer, 2012). alteration mineral assemblages: (1) excavation of pre-existing While this hydrothermal system was still active a crater lake alteration phases from the Martian crust during the formation of may have resulted that was fed by water derived from the Gale and older geologic processes, including impact craters; (2) melting of nearby ground ice or discharged by a sub-permafrost shock metamorphosed minerals and glassy impactites that are aquifer. The discovery of hydrous silicates and sulphates inside susceptible to alteration; (3) formation of a catchment area by the Gale supports past water activity (Milliken et al., 2010), which crater for sedimentation of hydrous-mineral-bearing strata sup- might be related to the alteration of basaltic rocks (Fig. 3; see also plied by fluvial or aeolian processes; (4) weathering of target rock McAdam et al., 2008, Schwenzer and Kring, 2010, Filiberto and and sediments in Gale under non-freezing conditions; and (5) Schwenzer, 2011, Schwenzer and Bridges 2012). Moreover, the generation of impact-generated hydrothermal systems that could channels that emanate from the central mound (Milliken et al., have formed hydrous phases, including phyllosilicates. Key obser- 2010) and crater rim provide persuasive geomorphic evidence vations to distinguish which of the above processes (or which that a volume of water significant enough to cause erosion flowed combination of the above processes) led to the observed altera- within the crater. All these factors might have contributed to the tion assemblages are alteration mineral assemblages and textural resulting mineralogical alteration, massive erosion, and sedimen- context, which will then allow deduction of the formation tation that occurred in the crater; they also are likely to have temperature range and other parameters. In addition, unaltered produced potentially habitable conditions, creating habitable target rock compositions can reveal the nature of the alteration environments that could have been inhabited—or uninhabited process, in situ or through element transport. Knowledge of (Cockell et al., 2012). Formation temperatures and geologic composition and mineralogy of the host rock, the alteration context will be two critical factors to distinguish sedimentary phases, and the geologic context will be obtained by detailed from in-situ formed alteration products and decide on the rover investigations. It will then be possible to put constraints on combination of geologic processes that lead to the formation of the formation mechanisms and conditions of hydrous alteration the observed mineralogy. Curiosity will return data to decide phases in Gale Crater—and its habitability. between these models.

Acknowledgements 6. Summary and conclusions This work was supported, in part, by National Aeronautics and Gale’s history from the late Late Noachian to today covers a Space Administration (NASA) Mars Fundamental Research grants wide range of processes, starting with the formation of a 150 km NNX07AK42G (D.A.K. and S.P.S.) and NNX09AL25G (A.H.T. and diameter complex crater with a central mound (recently named J.F.); by NASA MDAP grant number NNX09AI42G (P.J.M.); and by Mount Sharp) and ending with today’s aeolian erosion/sedimen- NASA Planetary Geology and Geophysics grant NNH07DA001N tation. The impact is thought to have melted portions of the (H.E.N.). The NASA Mars Science Laboratory project supported substrate to a maximum depth of 17 km and produced a R.C.W. and D.V., while a NASA Astrobiology Institute fellow- minimum of 3600 km3 of impact melt, half of which likely ship awarded to O.A. through the NASA Postdoctoral programme remained within the crater. The bulk of this impact melt would supported his work. We thank Charles Cockell and Nathalie Cabrol have pooled in an annular depression surrounding the central for discussions that improved earlier versions of the manuscript. uplift, creating an impact melt pool as thick as 0.5–1 km. The This is Lunar and Planetary Institute (LPI) contribution 1626. ejecta blanket surrounding Gale may have been as thick as 600 m, which has implications for the amount of erosion that has occurred since Gale Crater formed. After the impact, a References hydrothermal system might have been active for an extended period of time and a crater lake with associated sediments is Abramov, O., Kring, D.A., 2005. Impact-induced hydrothermal activity on Early likely to have formed. This water activity could have led to the Mars. Journal of Geophysical Research, 110, http://dx.doi.org/10.1029/ formation of phyllosilicates and sulphates. Even later, water- 2005JE002453. Abramov, O., Wong, S.M., Kring, D.A., 2012. Differential melt scaling for oblique borne and possibly airborne sediments were deposited in Gale, impacts on terrestrial planets. Icarus 218, 906–916. producing the central mound (and possibly other deposits). Allen, C.C., Jager, K.M., Morris, R.V., Lindstrom, D.J., Lindstrom, M.M., Lockwood, Gale and its sediments have been eroded significantly, both to J.P., 1998. JSC Mars-1: a Martian soil simulant, Space 98. In: Galloway, R.G., Lokaj, S. (Eds.), American Society of Civil Engineers, pp. 469–476. produce the moat around the central mound and to lower Gale’s Ames, D.E., Jonasson, I.R., Gibson, H.L., Pope, K.O., 2006. Impact-generated Hydro- northern rim. thermal System–Constraints from the large Paleoproterozoic Sudbury Crater. 94 S.P. Schwenzer et al. / Planetary and Space Science 70 (2012) 84–95

In: Cockell, Charles, Gilmour, Iain, Koeberl, Christian (Eds.), Biological Pro- Filiberto, J., 2011. Geochemical differences between surface basalts and Martian cesses Associated with Impact Events, pp. 55–100. meteorites: the need for Martian sample return. The Importance of Solar Anderson, R.B., Bell, J.F., 2010. The Geomorphology of the Proposed MSL field site in System Sample Return Missions to the Future of Planetary Science. March 5–6, Gale Crater. Presentation at the 4th Landing Site Workshop, 27–29 September Woodlands, Texas, US. Abstr. #5004. 2010. /http://marsoweb.nas.nasa.gov/landingsites/msl/workshops/4th_work Filiberto, J., Nekvasil, H., Lindsley, D.H., 2006. The Earth/Mars dichotomy in Mg/Si shop/talks/5_Edgett_Gale_opt.pdfS. and Al/Si ratios: is it real? American Mineralogist 91, 471–474. Barnhart, C.J., Nimmo, F., Travis, B.J., 2010. Geochemical and geomorphological effects Filiberto, J., Schwenzer, S.P., 2011. Hydrothermal alteration mineralogy of Home on post-impact hydrothermal systems incorporating freezing. Icarus 208, 101–117. Plate: thermochemical constraints for their formation conditions. Lunar and Bell III, J.F., Edgett, K.S., Rowland, S., Malin, M.C., 2006. The Gale crater mound: a Planetary Science Conference XLII, #2072. strong candidate landing site for the 2009 Mars Science laboratory. First MSL Garvin, J.B., Sakimoto, S.E.H., Frawley, J.J., Schnetzler, C., 2000. North Polar Region Landing Site Workshop. May 31–June 2, 2006. Pasadena CA. Craterforms on Mars: Geometric Characteristics from the Mars Orbiter Laser Bell III, J.F., Edgett, K.S., Rowland, S., Malin, M.C. 2010. The Gale Crater Mound: a Altimeter. Icarus 144, 329–352. Candidate Landing Site for the 2009 Mars Science Laboratory. Presentation at Gellert, R., Rieder, R., Bruckner,¨ J., Clark, B.C., Dreibus, G., Klingelhofer,¨ G., Lugmair, the 4th Landing Site Workshop. 27–29 September 2010. /http://marsoweb. G., Ming, D.W., Wanke,¨ H., Yen, A., Zipfel, J., Squyres, S.W., 2006. Alpha nas.nasa.gov/landingsites/msl/workshops/1st_workshop/talks/wed-PM/ particle X-ray spectrometer (APXS): Results from Gusev crater and calibration 02-Bell_Gale.ppt#256S.1, (visited March 2011). report. Journal of Geophysical Research, 111, http://dx.doi.org/10.1029/ Bibring, J.-P., Langevin, Y., Gendrin, A., Gondet, B., Poulet, F., Berthe´, M., Soufflot, A., 2005JE002555. Arvidson, R., Mangold, N., Mustard, J.F., Drossart, P., the OMEGA team, 2005. Golombek, M., Grant, J., Vasavada, A.R., Grotzinger, J., Watkins, M., Kipp, D., Noe Dobrea, Mars surface diversity as revealed by the OMEGA/Mars Express Observations. E., Griffes, J., Parker, T., 2011. Final four landing sites for the Mars Science Science 307, 1576–1581. Laboratory. 42th Lunar and Planetary Science Conference. Abstract#1520, Houston. Bibring, J.-P., Langevin, Y., Mustard, J.F., Poulet, F., Arvidson, R., Gendrin, A., Gondet, Grant, J.A., Golombek, M.P., Grotzinger, J.P., Wilson, S.A., Watkins, M.M., Vasavada, B., Mangold, N., Pinet, P., Forget, F., the OMEGA team, 2006. Global miner- A.R., Griffes, J.L., Parker, T.J. The science process for selecting the landing site alogical and aqueous Mars history derived from OMEGA/Mars Express data. for the 2011 Mars Science Laboratory. Planetary and Space Sciences. http://dx. Science 312, 400–404. doi.org/10.1016/j.pss.2010.06.016, in press. Brack, A., 2006. Clay Minerals and the Origin of Life. In: Bergaya, F., Theng, B.K.G., Greeley, R., Guest, J.E., 1987. Geologic Map of the Eastern Equatorial Region of Mars. Lagaly, G. (Eds.), Handbook of Clay Science, Developments in Clay Science, 1; U. S. Geological Survey, Miscellaneous Investigations Series, MAP I-1802-B. 2006, pp. 379–391. Department of the Interior, U. S. Geological Survey, prepared for NASA. Bridges, J.C., Catling, D.C., Saxton, J.M., Swindle, T.D., Lyon, I.C., Grady, M.M., 2001. Grieve, R.A.F., 2006. Impact structures in Canada. Canada (Geological Association Alteration Assemblages in Martian Meteorites: Implications for Near-Surface of Canada), 210 p. Processes. Space Science Reviews 96, 365–392. Grieve, R.A.F., Cintala, M.J., 1992. The analysis of differential impact melt-crater Bridges, J.C., Schwenzer, S.P., 2012. The Nakhlite Hydrothermal Brine. 43rd Lunar scaling and implications for the terrestrial crater record. Meteoritics and and Planetary Science Conference. Abstr. #2328. Planetary Science 27, 526–538. Cabrol, N.A., Grin, E.A., 1999. Disrtibution, Classification, and Ages of Martian Grieve, R.A.F., Dence, M.R., Robertson, P.B. 1977. Cratering processes - As inter- Impact Crater Lakes. Icarus 142, 160–172. preted from the occurrence of impact melts, in: Impact and explosion Cabrol, N.A., Grin, E.A., Newsom, H., Landheim, R., McKay, C.P., 1999. Hydrogeo- cratering: Planetary and terrestrial implications. In: Proceedings of the logic Evolution of Gale Crater and Its Relevance to the Exobiological Evolution Symposium on Planetary Cratering Mechanics, Flagstaff, Ariz. September 13– of Mars. Icarus 139, 235–245. 17, 1976. Pergamon Press, Inc, New York, pp. 791–814. Cabrol, N.A., Wynn-Williams, D.D., Crawford, D.A., Grin, E.A., 2001. Recent aqueous Grieve, R.A.F., Robertson, P.B., Dence, M.R., 1981. Constraints on the formation of environments in Martian impact craters: an astrobiological perspective. Icarus ring impact structures, based on terrestrial data. In: Schultz, P.H., Merrill, R.B. 154, 98–112. (Eds.), Multi-Ring Basins, Proceedings of the Lunar and Planetary Science, 12A; Caltech 2011. /http://msl.gps.caltech.edu/Gale/Targets.htmlS (visited March 2011). 1981, pp. 37–57. Changela, H.G., Bridges, J.C., 2010. Alteration assemblages in the nakhlites: Hartmann, W.K., 1977. Relative Crater Production Rates on Planets. Icarus 31, Variation with depth on Mars. Meteoritics Planet. Sci 45, 1847–1867. 260–276. Christensen, A.N., Olesen, M., Cerenius, Y., Jensen, T.R., 2008. Formation and Hecht, L., Wittmann, A., Schmitt, R.-T., Stoffler,¨ D., 2004. Composition of impact

Transformation of Five Different Phases in the CaSO4 H2O System: Crystal melt particles and the effect of post-impact alteration in suevitic rocks at the Structure of the Subhydrate b-CaSO4 0.5 H2O and Soluble Anhydrite CaSO4. Yaxopoil-1 drill core, Chicxulub crater, Mexico. Meteoritics and Planetary Chemistry of Materials 20, 2124–2132. Science 39, 1169–1186. Cintala, M.J., Grieve, R.A., 1998. Scaling impact melting and crater dimensions: Holsapple, K.A., Schmidt, R.M., 1982. On the scaling of Crater Dimensions 2. Implications for the lunar cratering record. Meteoritics and Planetary Science 33, Journal of Geophysical Research 87, 1849–1870.

889–912. Ivanov, B.A., Pierazzo, E., 2011. Impact cratering in H2O-bearing targets on Mars: Clifford, S.M., Parker, T.J., 2001. The Evolution of the Martian Hydrosphere: Thermal field under craters as starting conditions for hydrothermal activity. Implications for the Fate of a Primordial Ocean and the Current State of the Meteoritics and Planetary Science 46, 601–619. Northern Plains. Icarus 154, 40–79. Jolley, D., Gilmour, I., Gurov, E., Kelley, S., Watson, J., 2010. Two large meteorite Cockell, C.S., Balme, M., Bridges, J.C., Davila, A., Schwenzer, S.P., 2012. Uninhabi- impacts at the Cretaceous–Paleogene boundary. Geology 38, 835–838. tated Habitats on Mars–Icarus 217, 184–193, http://dx.doi.org/10.1016/ Kring, D.A., 1995. The dimensions of the Chicxulub impact crater and impact melt j.icarus.2011.10.025. sheet. Journal of Geophysical Research 100, 16979–16986. Cockell, C.S., Lee, P., Broady, P., Lim, D.S.S., Osinski, G.R., Parnell, J., Koeberl, C., Kring, D.A., 2005. Hypervelocity collisions into continental crust composed of Pesonen, L., Salminen, J., 2005. Effects of asteroid and comet impacts on sediments and underlying crystalline basement: comparing the Ries (24 km) habitats for lithophytic organisms—A synthesis. Meteoritics and Planetary and Chicxulub (180 km) impact craters. Chemie der Erde 65, 1–46. Science 40, 1901–1914. Kring, D.A., Swindle, T.D., Gleason, J.D., Grier, J.A., 1998. Formation and relative CRISM, 2010a. /http://crism.jhuapl.edu/msl_landing_sites/MSLimages4/Gale_Cra ages of maskelynite and carbonate in ALHA84001. Geochimica Cosmochimica ter_Revisited/FRT00011F99_L_TRR3/FRT00011F99_07_IF164L_TOPO.PNGS. Acta 62, 2155–2166. visited October 2010. Lambert, P., 2010. Target and impact deposits at Rochechouart impact structure, CRISM, 2010b. /http://crism-map.jhuapl.edu/summary.php?obs=FRT00011F990 France. In: Gibson, R.L., Reimold, W.U. (Eds.), Large Meteorite Impacts and 7IF164SS. visited October 2010. Planetary Evolution IV: Geological Society of America Special Paper, vol. 465; Croft, S.K., 1985. The Scaling of Complex Craters. In: Proceedings of the Fifteenth 2010http://dx.doi.org/10.1130/2010.2465(25), pp. 509–541. Lunar and Planetary Science Conference. Part 2, Journal of Geophysical Lindsay, J., Braiser, M., 2006. Impact Craters as Biospheric Microenvironments, Research, vol. 90. pp. C828–C842. Lawn Hill Structure, Northern Australia. Astrobiology 6, 348–363. Le Deit, L., Hauber, E., Fueten, F., Pondrelli, M., Zegers, T., Gasselt, S. van, Masse´,M., Malin, M.C., Edgett, K.S., 2000. Sedimentary Rocks of Early Mars. Science 290, Verpoorter, C., and the ISSI ILD team. 2011. Geological comparison of the Gale 1927–1937. crater moundto other equatorial layered deposits (ELDs) on Mars. 42nd Lunar Marsoweb, 2010. /http://marsoweb.nas.nasa.gov/landingsites/index.htmlS. Vis- and Planetary Science Conference. #1857. ited repeatedly, last in September/October 2010. Edgett, K.S., 2010. Curiosity’s candidate field site in Gale Crater, Mars. Presentation Marzo, G.A., Davila, A.F., Tornabene, L.L., Dohm, J.M., Faire´n, A.G., Gross, C., Kneissl, at the 4th Landing Site Workshop, 27–29 September 2010. /http://marsoweb. T., Bishop, J.L., Roush, T.L., McKay, C.P., 2010. Evidence for Hesperian impact- nas.nasa.gov/landingsites/msl/workshops/4th_workshop/talks/5_Edgett_Ga induced hydrothermalism on Mars. Icarus 208, 667–683. le_opt.pdfS. Visited March 2011. Masaitis, V.L., Naumov, M.V., Mashchak, M.S., 1999. Anatomy of the Popigai impact Edgett, K.S., Malin, M.C., 2001. Rock stratigraphy in Gale crater, Mars. Lunar and crater, Russia. In: Dressler, B.O., Sharpton, V.L. (Eds.), Large meteorite Impacts and Planetary Science, XXXII: #1005. Planetary Evolution II; Geological Society of America, Special Paper, 339; 1999, Ehlmann, B., Mustard, J.F., Murchie, S.L., Bibring, J.-P., Meunier, A., Fraeman, A.A., pp. 1–17. Langevin, Y., 2011. Subsurface water and clay mineral formation during the McAdam, A.C., Zolotov, M.Y., Mironenko, M.V., Sharp, G., 2008. Formation of early history of Mars. Nature 479, 53–60. silicaby low-temperature acid alteration of Martian rocks: Physical-chemical Faire´n, A.G., Chevrier, V., Abramov, O., Marzo, G.A., Gavin, P., Davila, A.F., constraints. Journal of Geophysical Research, 113, http://dx.doi.org/10.1029/ Tornabene, L.L., Bishop, J.L., Roush, T.L., Gross, C., Kneissl, T., Uceda, E.R., Dohm, 2007JE003056. J.M., Schulzt-Makuch, D., Rodrı´guez, A.P., Amils, R., McKay, C.P., 2010. Noa- McGetchin, T.R., Settle, M., Head, J.W., 1973. Radial thickness variation in impact chian and more recent phyllosilicates in impact craters on Mars. Proceedings crater ejecta—Implications for lunar basin deposits. Earth and Planetary of the National Academy of Sciences USA 107, 12095–12100. Science Letters 20, 226–236. S.P. Schwenzer et al. / Planetary and Space Science 70 (2012) 84–95 95

McKay, C.P., Davis, W.L., 1991. Duration of liquid water habitats on early Mars. Pierazzo, E., Melosh, H.J., 2000. Melt production in oblique impacts. Icarus 145, Icarus 90, 214–221. 252–261. McSween Jr., H.Y., Taylor, G.J., Wyatt, M.B., 2009. Elemental Composition of the Pike, R.J., 1988. Geomorphology of Impact Craters on Mercury. In: Chapman, C.R., Martian Crust. Science 324, 736–739. Matthews, M.S., Vilas, F. (Eds.), Mercury. University of Arizona Press, pp. 165–273. Melosh, H.J., 1989. Impact Cratering. Oxford University Press, New York. Rathbun, J.A., Squyres, S.W., 2002. Hydrothermal Systems Associated with Martian Milliken, R.E., Grotzinger, J.P., Thomson, B.J., 2010. Paleoclimate of Mars as Impact Craters. Icarus 157, 362–372. captured by the stratigraphic record in Gale Crater. Geophysical Research Reed, M.H., Spycher, N.F., 2006. Users Guide for CHILLER: A Program for Comput- Letters, 37, http://dx.doi.org/10,1029/2009GL041870. ing Water-Rock Reactions, Boiling, Mixing and Other Reaction Processes in Ming, D.W., Gellert, R., Morris, R.V., Arvidson, R.E., Bruckner,¨ J., Clark, B.C., Cohen, Aqueous-Mineral-Gas systems and Minplot Guide. Third Edition. University of B.A., d’Uston, C., Economou, T., Fleischer, I., Klingelhofer,¨ G., McCoy, T.J., Oregon, Oregon 67. Mittlefehldt, D.W., Schmidt, M.E., Schroder,¨ C., Squyres, S.W., Tre´guier, E., Schmidt, M.E., McCoy, T.J., 2010. The evolution of a heterogeneous Martian Yen, A.S., Zipfel, J., 2008. Geochemical properties of rocks and soils in Gusev mantle: Clues from K, P, Ti, Cr, and Ni variations in Gusev basalts and Crater, Mars: Results of the Alpha Particle X-Ray Spectrometer from Cumber- shergottite meteorites. Earth and Planetary Science Letters 296, 67–77. land Ridge to Home Plate. Journal of Geophysical Research, 113, http://dx.doi.o Schwenzer, S.P., Kring., D.A., 2009. Impact-generated hydrothermal systems: rg/10.1029/2008JE003195. capable of forming phyllosilicates on Noachian Mars. Geology 37, 1091–1094. Minissale, A., Vaselli, O., Chandrasekharam, D., Magro, G., Tassi, F., Casiglia, A., Schwenzer, S.P., Kring, D.A., 2009a. Impact-generated hydrothermal systems in 2000. Origin and evolution of ‘intracratonic’ thermal fluids from central- mafic to ultramafic Noachian crust of Mars. Workshop on Modeling Hydrous western peninsular India. Earth and Planetary Science Letters 181, 377–394. Environments on Mars: #4015, 2 Abb.; Lunar and Planetary Institute, Houston. Moore, H.J., 1977. Nevada test site craters used for astronaut training. Journal of Schwenzer, S.P., Kring, D.A., 2009b. Impact-generated hydrothermal alteration on

Research of the U. S. Geological Survey 5 (6), 719–733. Early Mars in presence of CO2. Meteoritics and Planetary Science 44, A188. Morgan, J., Warner, M., 1999. Chicxulub: the third dimension of a multi-ring (abstr. #5262). impact basin. Geology 27, 407–410. Schwenzer, S.P., Kring, D.A., 2009c. Impact-generated hydrothermal alteration on Morgan, J., Warner, M., Brittan, J., Buffler, R., Camargo, A., Christeson, G., Denton, P., Mars: clay minerals, oxides, zeolites, and more. Lunar and Planetary Science Hildebrand, A., Hobbs, R., Macintyre, H., Mackenzie, G., Maguire, P., Marin, L., XL: #1421. Nakamura, Y., Pilkington, M., Sharpton, V., Snyder, D., Suarez, G., Trejo, A., Schwenzer, S.P., Kring, D.A., 2010. Evaluating the effects of sulphur on alteration 1997. Size and morphology of the Chicxulub impact crater. Nature 390, assemblages in impact-cratered terrains on Mars. Lunar and Planetary Science 472–476. Conference, XLI, #1614. Murchie, S.L., Mustard, J.F., Ehlmann, B.L., Milliken, R.E., Bishop, J.L., 12 co-authors, Spray, J.G., Thompson, L.M., Biren, M.B., O’Connell-Cooper, C., 2010. The Mani- 2009a. A synthesis of Martian aqueous mineralogy after 1 Mars year of couagan impact structure as a terrestrial analogue site for lunar and martian observations from the Mars Reconnaissance orbiter. Journal of Geophysical planetary science. Planetary and Space Science 58, 538–551. Research, 114, http://dx.doi.org/10.1029/2009JE003342. Thomson, B.J., Bridges, N.T., Milliken, R., Baldrige, A., Hook, S.J., Crowley, J.K., Murchie, S.L., Seelos, F.P., Hash, C.D., Humm, D.C., Malaret, E., 18 co-authors, Marion, G.M., de Souza Filho, C.R., Brown, A.J., Weitz, C. M. 2011. Constraints 2009b. Compact reconnaissance imaging spectrometer for mars investigation on the origin and evolution of the layered mound in Gale Crater, Mars using and dataset from the mars reconnaissance orbiter’s primary science Mars Reconaissance Orbiter data. Icarus, 10.1016/j.icarus.2011.05.002. phase. Journal of Geophysical Research, 114, http://dx.doi.org/10.1029/ Thomson, B.J., Bridges, N.T., Milliken, R., Hook, S., Baldridge, A., Crowley, J., Marion, 2009JE003344. G., 2008a. Signs of aqueous activity in Gale Crater, Mars, as viewed by Mars Mustard, J.F., Murchie, S., Pelkey, S.M., Ehlmann, B.L., Milliken, R.E., 31 co-authors, Reconnaissance Orbiter. American Geophysical Union, Fall Meeting 2008. 2008. Hydrated silicate minerals on Mars observed by the Mars Reconnais- Abstract #P41B-1365. sance Orbiter CRISM instrument. Nature 454, 305–309. Thomson, B.J., Bridges, N.T., Milliken, R., Bell III, J.F., Calvin, W.C., Weitz, C.M., NASA (2012) /http://www.nasa.gov/mission_pages/msl/multimedia/pia15292. 2008b. New constraints of the origin and evolution of the layered deposits in htmlS. Gale crater, Mars. Lunar and Planetary Science XXXIX. Abstract #1456. Newsom, H.E., 1980. Hydrothermal alteration on impact melt sheets with Treiman, A.H., 2005. The nakhlite meteorites: augite-rich igneous rocks from Mars. implications for Mars. Icarus 44, 207–216. Chemie der Erde 65, 203–270. Newsom, H.E., 2010. Heated lakes on Mars. In: Cabrol, N., Grin, E. (Eds.), Lakes on Vaniman, D.T., Chipera, S.J., Bish, D.L., Carey, J.W., Levy, S.S., 2001. Quantification of Mars. Elsevier, Amsterdam, pp. 91–110. unsaturated-zone alteration and cation exchange in zeolitized tuffs at Yucca Newsom, H.E., Brittelle, G.E., Hibbitts, C.A., Crossey, L.J., Kudo, A.M., 1996. Impact Mountain, Nevada, USA. Geochimica et Cosmochimica Acta 65, 3409–3433. crater lakes on Mars. Journal of Geophysical Research 101, 14,951–14,955. Vaniman, D.T., Chipera, S.J., Bish, D.L., Carey, J.W., Fialips, C.I., Feldman, W.C., 2004. Osinski, G.R., Spray, J.G., Lee, P., 2001. Impact-induced hydrothermal activity in the Sulfate salts, regolith interactions, and water storage in equatorial martian Haughton impact structure, Canada: generation of a transient, warm, wet regolith. Lunar and Planetary Science, XXXV. Abstract #1690. oasis. Meteoritics and Planetary Science 36, 731–745. Vaniman, D.T., Chipera, S.J., 2006. Transformations of Mg- and Ca-sulfate hydrates Osinski, G.R., Tornabene, L., Banerjee, N., Cockell, C., Flemming, R., Izawa, M., in Mars regolith. American Mineralogist 91, 1628–1642. McCutcheon, J., Parnell, J., Preston, L., Pickersgill, A., Pontefract, A., Sapers, H., Vaniman, D.T., Bish, D.L., Chipera, S.J., Rearick, M.S., 2011. Relevance to Mars of Southam, G. (in preparation): Impact-generated hydrothermal systems on cation exchange between nontronite and Mg-sulfate brine. Lunar and Plane- Earth and Mars. Icarus, in review. tary Science XXXXII. Abstract #2276. Pelkey, S.M., Jakosky, S.M., 2002. Surficial Geologic Surveys of Gale Crater and Varnes, E.S., Jakosky, B.M., McCollom, T.M., 2003. Biological potential of martian Melas Chasma, Mars: Integration of Remote-Sensing Data. Icarus 160, hydrothermal Systems. Astrobiology 3, 407–414. 228–257. Watson, J.S.; Gilmour, I.; Jolley, D.W.; Kelley, S.P.; Gilmour, M.A. Gurov, E.P., 2010. Pelkey, S.M., Jakosky, B.M., Christensen, P.R., 2004. Surficial properties in Gale Molecularparameters of post impact cooling in the Boltysh impact structure. Crater, Mars, from Mars Odyssey THEMIS data. Icarus 167, 244–270. 41st Lunar and Planetary Science Conference. Abstract #2296. Pierazzo, E., Kring, D.A., Melosh, H.J., 1998. Hydrocode simulation of the Chicxulub Zurcher,¨ L., Kring, D.A., 2004. Hydrothermal alteration in the core of the Yaxopoil-1 impact event and the production of climatically active gases. Journal of borehole, Chicxulub impact structure, Mexico. Meteoritics and Planetary Geophysical Research 103, 28607–28625. Science 39, 1199–1221.