Horowitz & Fischer 2014

8/12/2019 Horowitz & Fischer 2014

1/14

Contrasting styles of waterrock interaction at theMars Exploration Rover landing sites

Joel A. Hurowitz a , , Woodward W. Fischer b

a Department of Geosciences, Stony Brook University, 100 Nicolls Road, Stony Brook, NY 11794-2100, United Statesb Division of Geological and Planetary Sciences, California Institute of Technology, 1200 East California Boulevard, Pasadena, CA 91125,

United States

Received 21 September 2012; accepted in revised form 17 November 2013; available online 24 November 2013

Abstract

The nature of ancient hydrological systems on Mars has been the subject of ongoing controversy, driven largely by a discon-nect between observational evidence for owing water on the Martian surface at multiple scales and the incompatibility of suchobservations with theoretical models that predict a cold early Martian environment in which liquid water is unstable. Here wepresent geochemical data from the Mars Exploration Rovers to evaluate the hydrological conditions under which weatheringrinds, soils, and sedimentary rocks were formed. Our analysis indicates that the chemistry of rinds and soils document awater-limited hydrologic environment where small quantities of S-bearing uids enter the system, interact with and chemicallyalter rock and soil, and precipitate secondary mineral phases at the site of alteration with little to no physical separation of pri-mary and secondary mineral phases. In contrast, results show that the sedimentary rocks of the Burns Formation at MeridianiPlanumhave a chemical composition well-described as a mixture between siliciclastic sediment and sulfate-bearing salts derivedfrom theevaporation of groundwater.We hypothesize that the formermay be derived from therecently investigatedShoemakerFormation, a sequence of impactbreccias that underlie theBurnsFormation. This resulthas important implications forthe styleof chemical weathering and hydrology recorded by these sedimentary materials, revealing long-range transport of ions in solu-tion in an open hydrological system that is consistent only with subsurface or overland ow of liquid water.

2013 Elsevier Ltd. All rights reserved.

1. INTRODUCTION

The Martian geologic record contains a wide variety of geomorphological features consistent with owing andstanding water on the surface (e.g., outow channels, valleynetworks, alluvial fans and deltas, sediment-lled cratersand basins), with the generation of water-related landforms

generally accepted to have been more prevalent in earlyMartian history and waning towards the modern day (forrecent reviews, see Carr, 2012; Grotzinger and Milliken,2012). The discovery of a broad suite of aqueous secondaryminerals produced from alteration of the Martian crust byuids has added further constraints on the nature of

waterrock interactions. Syntheses of these mineralogicobservations from orbital data have driven a hypothesis forfundamental secular changes in secondary mineral formingprocesses over time, from early clay-mineral forming envi-ronments to intermediate sulfate-mineral forming environ-ments to more modern environments in which anhydrousferric iron oxides areproduced ( Bibring et al., 2005; Mustard

et al., 2008; Murchie et al., 2009; Ehlmann et al., 2011 ). Thesegeomorphic and mineralogic observations are broadlythought to reect a change in environmental conditions atthe Martian surface from an early wet (and possibly warm)state to the pervasive dry, cold conditions observed today.

Despite geomorphologic, sedimentological, and mineral-ogical evidence for the presence of liquid water at theMartian surface, at least in its early geologic history,signicant questions remain regarding the state and evolu-tion of the Martian atmosphere (e.g., Greenwood et al.,

0016-7037/$ - see front matter 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.gca.2013.11.021

Corresponding author. Tel.: +1 631 632 6801.E-mail address: [email protected] (J.A.Hurowitz).

www.elsevier.com/locate/gca

Available online at www.sciencedirect.com

ScienceDirect

Geochimica et Cosmochimica Acta 127 (2014) 2538

8/12/2019 Horowitz & Fischer 2014

2/14

2008; Cassata et al., 2012), specically whether it was suffi-ciently thick and had an adequate greenhouse budget for li-quid water to have been stable at the surface of Mars. Anumber of models have been proposed that can providethe warm, clement conditions required for liquid water sta-bility (e.g., Forget and Pierrehumbert, 1997; Sagan andChyba,1997; Halevy et al., 2007), though none of these models havegarnered universal acceptance (for a recent review see Nileset al., 2013). The apparent inconsistency between observa-tions and theory has led to suggestions that Mars may onlyhave experienced transient warm climate episodes drivenby volcanism (e.g., Johnson et al., 2008, 2009 ) or large im-pact events (Segura et al., 2002). Others have questionedthe sedimentary interpretations for liquid water (e.g., Gaidosand Marion, 2003; Niles and Michalski, 2009; Kite et al.,2013), hypothesizing instead that widespread sedimentarydeposits were produced in a cold, polar-type climate formuch of Martian geologic history. In these conceptual mod-els, the previously described geomorphic and mineralogicfeatures are largely generated by processes such as ground-

water escape from beneath a global cryosphere, melting of snowpacks, or glacial melting, as examples.The ancient sedimentary rocks of the Burns Formation

at Meridiani Planum, Mars have recently become a touch-stone in the debate over whether climate conditions wereever appropriate to allow for stable liquid water in surfaceenvironments at the time these deposits were formed (ca.3.53.7 Ga). Based on extensive in situ observations bythe Mars Exploration Rover (MER) Opportunity , the sedi-mentary rocks of the Burns Formation were interpreted tohave formed under dry, desert-like conditions in whichchemically weathered basaltic sediment was transported inan aeolian environment and cemented by evaporating, sul-fate-rich groundwater ( Grotzinger et al., 2005; Squyres and

Knoll, 2005). This interpretation is consistent with modelsof global groundwater transport that indicate that Meridi-ani Planum was a locus for groundwater upwelling andevaporation, requiring a climate system that would havesupported meteoric recharge to drive groundwater trans-port ( Andrews-Hanna et al., 2007, 2010 although seeMichalski et al., 2013 ). If this interpretation is correct, thenthe Burns Formation places an important constraint onMartian climate conditions, and indicates that during lateNoachian to early Hesperian time, the climate was suffi-ciently warm and wet to allow for the existence of at leasta limited hydrologic cycle in which aquifers could be re-charged by rainfall and/or snowmelt.

The Burns Formation is characterized by enigmaticchemistry and mineralogy: its bulk composition has beenbroadly described as a mixture of basalt plus sulfur , andit contains as a ubiquitous mineral component the mineral jarosite K ; Na ; H3O Fe33 SO42OH6, which requiresacidic (pH = ca. 24) conditions to precipitate. This chemis-try is somewhat surprising given theprocess hydrology inter-pretation of the deposit that implies aquifer recharge andsubsequent groundwater upwelling and evaporation. In sucha setting, elemental fractionation along the ow path due todifferential mineral solubility and the generation of alkalinityfrom silicate weathering might reasonably be expected.Highlighting the apparent inconsistency between the charac-

teristics of theBurns Formation andan open hydrologic sys-tem, a number of alternative hypotheses have been advancedthat interpret theBurns Formationas theproduct of sulfuricacid-driven weathering in volcaniclastic ( McCollom andHynek, 2005), impact (Knauth et al., 2005 ), or glacial ice(Niles and Michalski, 2009 ) environments.

These alternative hypotheses, driven by the recognitionof a basalt plus sulfur bulk chemical composition; invokechemical weathering processes described as closed sys-tem , isochemical , or cation-conservative . What ismeant by these terms from a process perspective is: sulfuricacid-charged water is added to basalt, a suite of secondaryminerals is generated, the water leaves the system in the gasphase by vaporization or sublimation (and/or is partiallyretained in hydrated secondary minerals), and left behindis a mixture of newly formed secondary minerals and resi-due that have not been physically separated from eachother. The term cation-conservative (Niles and Michalski,2009) is the most accurate for the hydrochemical conditionsdescribed above, which are neither truly closed nor truly

isochemical

; accordingly, we employ the term

cation-conservative here. Acid fog hypotheses also describe a form of cation-con-

servative chemical weathering that yield basalt plus sulfurchemistries, and are often invoked to explain the chemicalcomposition of soils on Mars. In this model, acidic volcanicaerosols settle out of the atmosphere onto rock, soil, anddust, altering the surfaces they come into contact withthrough low water-to-rock ratio, low pH weathering (e.g.,Settle, 1979; Clark and Van Hart, 1981; Banin et al., 1997;King and McSween, 2005; Hurowitz et al., 2006b; Minget al., 2008; King and McLennan, 2010 ). Similar mechanicscould arise from impact processes in which the decomposi-tion of sulfate minerals generates recycled sulfuric acid

to drive silicate weathering, similar to what is thought tohave occurred in the terrestrial Chixulub impact event(Kring, 2007; Zolotov and Mironenko, 2007; King andMcLennan, 2010 ). The weathering process associated withthe acid fog model is similar to the hypothesis proposed byMcCollomandHynek (2005) , and more recently by Michalskiand Bleacher (2013) , who suggested that the chemistry of theBurns Formation was generated by the co-deposition of acidvolatiles and basalt in volcaniclastic base surge deposits.

Building on analogues in terrestrial polar environments,Niles and Michalski (2009) proposed that Burns Formationsediment was generated during alteration in small pockets of meltwater hosted in massive ice sheets in the equatorial lati-tudes of Mars (including over Meridiani Planum). In thesemelt pockets, low temperature and low water-to-rock ratiochemical weathering occurred between co-deposited basalticdust and volcanic acid aerosols trapped in the ice as it accu-mulated. Later sublimation of the ice left behind altered sed-iments with a basalt plus sulfur bulk composition.

In this report we present evidence from rover geochem-ical data that the sedimentary record of Mars preserves evi-dence for both cation-conservative weathering processesand open system weathering conditions in which there issufficient liquid water present to mobilize and transport cat-ions in solution over more signicant length scales. Thegeochemistry of modern soils and weathered rock surfaces

26 J.A. Hurowitz, W.W. Fischer/ Geochimica et Cosmochimica Acta 127 (2014) 2538

8/12/2019 Horowitz & Fischer 2014

3/14

is well explained by cation-conservative weathering pro-cesses. In stark contrast, the ancient sedimentary rocks of the Burns Formation at Meridiani Planum capture open-system hydrochemical processes that require long range sol-ute transport and associated fractionation of elements. Theexamples we describe from the MER landing sites lay out auseful framework to discover and understand the chemicalfractionations induced by weathering and alteration undera broad spectrum of hydrologic conditions, and will pro-vide examples against which analyses by the Mars ScienceLaboratory Curiosity can be compared as it begins explora-tion of the 5 km-thick sedimentary section at Gale Crater.

2. METHODS

The MERs Spirit and Opportunity collected a wealth of dataon thecomposition, mineralogy, and textural propertiesof soils and rocks on Mars. Spirit operated in Gusev Craterfrom January 2004 through March 2010, when contact withthe rover was lost. Opportunity continues its science mission

at Meridiani Planum, 9+ years since landing in January2004. Over a combined traverse distance of >40 km, thesetwin rovers have collected well in excess of 300 rock and soilchemical analyses using their Alpha-Particle X-ray Spec-trometers (APXS). During analysis, the APXS instrumentsdetermine concentrations of the elements Si, Ti, Al, Fe,Mn, Mg, Ca, Na, K, P, S, Cl, Zn, Ni, and Br. The APXSinstruments are deployed on the end of each rovers armand provide contact analyses of rocks and soils with a eldof view of 3.8 cm and a penetration depth of 120 l mdepending on the material and atomic weight of the elementbeing interrogated ( Rieder et al., 2004; Gellert et al., 2006 ).Increased sampling depth for rocks is achieved through theuse of the Rock Abrasion Tool (RAT), also mounted on

the end of each the rovers arm, which can be employed tobrush rock surfaces clean and abrade rock surfaces to a max-imum depth of 15 mm (Gorevan et al., 2003 ). APXS anal-yses areperformed onnatural soil androck surfaces,brushedrock surfaces, and abraded rock surfaces, and are typicallyreferred to as undisturbed or as is , brushed , and RAT-ed , respectively.

For our analysis of the geochemical data, we employdiagrams that plot molar Al 2O3/(FeO T + MgO + CaO +Na 2O + K 2O) versus SiO2 /SO 3 . This parameter set isparticularly valuable because it captures much of the com-positional variability inherent in unaltered igneous rocksand also enables an assessment of the degree to whichbasaltic rocks and soils have been affected by chemical(e.g., mineral dissolution during chemical weathering) andphysical (e.g., mineral fractionation during physical trans-port) alteration processes subsequent to emplacement.The ordinate axis was primarily chosen to capture the rela-tionship between an oxide that is often described as rela-tively immobile during chemical weathering processes(Al2O3) and oxides that tend to be substantially more mo-bile (FeO, MgO, CaO, Na 2O, K 2O) during those sameweathering processes. Oxidation will signicantly decreaseFe-solubility and under conditions that promote oxidation,Fe and Al mobility can be considered to be quite similar.Ternary diagrams, which plot (FeO + MgO) Al 2O3

(CaO + Na 2O + K 2O) at the apices, also offer a juxtaposi-tion of immobile and mobile element oxides and have beenparticularly useful for understanding chemical weatheringprocesses on Earth ( Nesbitt and Wilson, 1992 ) and Mars(Hurowitz et al., 2006b ). However, because the Martiansoils, igneous rocks, weathering rinds, and sedimentaryrocks under consideration here do not separate from eachother on such ternary diagrams, we have chosen to collapsethese oxide interrelationships onto a single axis and plotthem against the ratio SiO 2 /SO 3. The inability of these ter-nary diagrams to segregate the Burns Formation fromother soils and rocks has led to an interpretation that thesediments that comprise the Burns Formation formed fromlow water-to-rock ratio chemical alteration processes (e.g.,Hurowitz and McLennan, 2007 ); a hypothesis that can betested by casting the data into a different chemical space.

The abscissa (SiO 2 /SO 3) enables an evaluation of themobility of SiO 2 and the degree to which sulfur has inu-enced the geochemical characteristics of a given soil or rocksample. This ratio was employed by McLennan (2003) to

evaluate soil and rock geochemical variations at the Path-nder landing site at Ares Valles, Mars, and Hurowitzet al. (2006a) to evaluate rock geochemistry at Husband Hill,in Gusev Crater, Mars. In fact, SO 3 could be exchanged forchemical species such as Cl or CO 2 in order to evaluate theinuence of hydrochloric acid or carbonic acid on rock andsoil geochemistry. We have deliberately chosen to use SO 3as thedenominator since S is themost abundant anionic con-stituent of the rocks and soils under consideration. In gen-eral, less abundant Cl mimics the behavior of S, shown bythe consistent SO 3:Cl ratio of almost all samples in our data-set (7.5 1.4, n = 50), excluding the Burns Formation andsubsurface soils in trenches,which are discussed more below.APXS does notanalyze C (oranyother elements with masses

below that of Na), and so we cannot evaluate the impact of carbonate-forming processes.

3. RESULTS

The APXS data used in our study can be found in theElectronic Annex as a tabulated *.csv le. The full APXSdataset for both MERs can be downloaded from theMER Analysts Notebook on the Planetary Data SystemGeosciences node with the exception of data for the Shoe-maker Formation at Meridiani Planum, which was re-ported in Squyres et al. (2012).

3.1. Adirondack Class basalts

Adirondack Class basalts are olivine-rich (picritic) bas-alts emplaced as ejecta from impacts into the Hesperian-aged basalt ows that blanket the oor of Gusev Craterat the Spirit landing site. No in-place examples of this rockclass were ever encountered by the Spirit rover, thus theyare termed oat rocks. These basalt lithologies wereencountered during Spirits traverse across the oor of Gusev Crater between sol 0 and 150 of the MER Spirit mis-sion; their characteristics and petrogenesis are fullydescribed in McSween et al. (2006). The RAT-ed surfacesof Adirondack Class basalts are characterized by a uniform

J.A. Hurowitz, W.W. Fischer/ Geochimica et Cosmochimica Acta 127 (2014) 2538 27

8/12/2019 Horowitz & Fischer 2014

4/14

8/12/2019 Horowitz & Fischer 2014

5/14

Fig. 1. Plots of Al2O3/(FeO T + MgO + CaO + Na 2O + K 2O) versus SiO2/SO 3 (A-C, E) and (FeO T + MgO + CaO + Na 2O + K 2O)/SO 3versus SiO2/SO 3 diagrams (D, F), all data plotted as mole percentages of the oxides. Figures A and B demonstrate the effects of cation-conservative weathering (CCW) on relatively unweathered basalt and regolith, respectively. On Figure B, El_D = El Dorado, Fo 50 = olivinewith the composition (Fe 0.5 ,Mg 0.5 )2SiO4. Figures C and E show mixing models between a groundwater-derived component (rich in Fe, Mg,Ca, and S) and a variety of siliciclastic protoliths. Figures D and F are used to validate the mixing relationships in C and E (see text), andhighlight the potential diagenetic disturbance of some samples of the Burns Formation (samples denoted Group 3 ). The arrow on Fig. 1Eindicates the vector from RAT-ed to undisturbed surfaces of Shoemaker suevite breccias.

J.A. Hurowitz, W.W. Fischer/ Geochimica et Cosmochimica Acta 127 (2014) 2538 29

8/12/2019 Horowitz & Fischer 2014

6/14

deviation observed in the soil data from Meridiani Planumand Gusev crater must have been caused by a process thatadded differing amounts of sulfur without signicant mobi-lization and fractionation of cations.

On Fig. 1A we show the results of modeling a simpliedchemical weathering process in which olivine, in this caseFo 50 Fa 50 , is mathematically subtracted from the analysisHumprey_RAT1 (the rock analysis with the highest SiO 2 /SO3). This intermediate olivine composition is consistentwith measured and modeled olivine compositions forAdirondack Class basalts from Gusev Crater ( McSweenet al., 2008), and we emphasize olivine dissolution as it isthe most kinetically labile primary igneous phase presentin the Adirondack class basalts ( Stopar et al., 2006; Zolotovand Mironenko, 2007; Hausrath et al., 2008 ). In this calcu-lation, for every mole of MgO, FeO, and SiO 2 subtracted, amole of MgSO 4, FeSO 4 , and SiO 2 was added back into theresidual rock chemical composition. Effectively, this algo-rithm captures the dissolution of olivine by a S-bearinguid (e.g., sulfuric acid), followed by the precipitation of

divalent metal-sulfates and silica, and produces a horizontaltrend away from the RAT-ed interior surfaces of Adiron-dack class basalts that explains the remaining data well.Olivine removal in the absence of secondary mineral precip-itation has also been modeled on Fig. 1A and is, in part, re-quired to explain the observation that the brushed rocksurfaces are offset vertically on Fig. 1A relative to theRAT-ed rock surfaces. This analysis reveals that the differ-ences in chemical composition between the interior andexterior of rock surfaces is best explained by the formationof thin (mm-scale) weathering rinds on rock surfaces underconditions in which olivine is removed and sulfur is addedwithout signicant cation fractionation relative to the pro-tolith. Altogether, these results support chemical weather-

ing processes hypothesized in earlier works to explain thealteration of rock surfaces at Gusev crater ( Haskin et al.,2005; Ming et al., 2006; Hurowitz et al., 2006b; Hausrathet al., 2008; King and McLennan, 2010 ). In our updatedanalysis, however, the importance of the precipitation of secondary minerals at the site of primary mineral dissolu-tion is highlighted.

We note that the actual choice of secondary productformed is immaterial to the alteration trajectory realizedunder these conditions. For instance, a scenario in whichFe 3+ -oxide is substituted for Fe 2+ -sulfate again results ina horizontal trajectory, but in this case, half the SO 3 isadded for every mole of olivine dissolved. In effect, thisdoubles the amount of cation-conservative alterationneeded to cause a unit change in SiO 2/SO 3 relative to themodel depicted on Fig. 1A. We can also envision a scenarioin which S-bearing acid volatiles are added to the systemforming a suite of amorphous secondary mineraloids fromthe dissolution of any one of the primary igneous mineralphases present in the host rock (e.g., plagioclase, pyroxene,olivine). Indeed, the horizontal trend in the data could evenbe explained simply as sulfur addition, with no chemicalinteractions between sulfur and the basaltic substrate towhich it is added, though such a scenario seems geologicallyimplausible. What can be stated with condence is that ahorizontal array is produced when sulfur is added to the

system and primary and secondary phases are not physi-cally separated from each other following interactionbetween sulfur and basalt. In this way, the Al 2O3 /(FeO T +MgO + CaO + Na 2O + K 2O) ratio remains relativelyunfractionatedfromthatof theprotolith, while SiO 2/SO 3 de-creases. Changing the primary and secondary minerals in-volved simply changes the degree of cation-conservativeweathering needed to move a given distance along the hori-zontal trajectory. Interestingly, undisturbed soil analysesfrom Gusev Crater and Meridiani Planum fall further alongthe continuum of cation-conservative weathering, indicatinga commonality of weathering processes in the generation of alteration rinds on rocks and the regolith on Mars.

Additional processes can be recognized from the soilchemical data, which are plotted alone with a linear abscis-sa in Fig. 1B. Here, we re-calculate the cation-conservativeweathering trajectory starting from an analysis of a soil tar-get named El Dorado . This target was part of a eld of soil ripples encountered on the south-facing slope of Hus-band Hill in Gusev Crater. The nature of the ripple eld

was fully described in Sullivan et al. (2008); the salient pointfor our analysis here is that this soil represents an endmem-ber composition in SiO 2/SO 3 space, and given what isknown about its mineralogical and geochemical properties(Ming et al., 2008; Morris et al., 2008 ), appears to be oneof the least chemically altered soils encountered at eitherMER landing site. Accordingly, it presents an appropriateendmember from which to examine the effects of weather-ing processes recorded by soils. The cation-conservativeweathering trajectory from El Dorado also matches theremaining soil data well, with some vertical scatter aboutthis trend. It has been proposed that hydrodynamic sorting,particularly of olivine, plays an important role in determin-ing the chemical and mineralogical properties of soils at the

MER landing sites ( McGlynn et al., 2012 ). The olivine losstrajectory on Fig. 1B can effectively be cast as either a phys-ical (hydrodynamic) removal mechanism or a chemical (dis-solution) removal mechanism. We also model the effect of olivine addition on Fig. 1B, which might occur as a resultof physical concentration during transport. While olivineloss/gain can certainly account for the variation from a hor-izontal cation-conservative weathering trajectory, it cannotaccount for the widespread observation of sulfur additionto the system. This suggests that hydrodynamic loss andgain processes played a subordinate role to cation-conser-vative weathering in determining the chemical compositionof these soils.

Soils exposed in the three wheel trenches that were dugby the Spirit rover on the traverse across the basaltic plainsof Gusev Crater reveal an interesting but different trend,specically due to the accumulation of excess secondarymineral components. In broad agreement with the conclu-sions of Wang et al. (2006) , the chemical composition of subsurface soils at these trench investigation sites impliesthe addition of a mixture of sulfate salts and silica, here cal-culated as 1/3 SiO 2 and 2/3 (Mg, Fe, Ca)SO 4 to an analysisof soil in a trench wall, which has a composition similar tothat of undisturbed surface soils (target name: Road-Cut_Wall). Such salt accumulations provide an importantexample of open system hydrological processes on a local

30 J.A. Hurowitz, W.W. Fischer/ Geochimica et Cosmochimica Acta 127 (2014) 2538

8/12/2019 Horowitz & Fischer 2014

7/14

8/12/2019 Horowitz & Fischer 2014

8/14

chemically-weathered version of the modern basaltic rego-lith that is omnipresent at the MER landing sites. Previouswork has suggested that the Burns Formation is composedof distinct chemical and siliciclastic endmembers, with thesiliciclastic component represented by basalt that had

5060% of its soluble Ca, Mg, and Fe removed by leach-ing (Squyres et al., 2006b). Interestingly, these results sug-gest that far less leaching is required to explain the data,as the mixing array crosses the weathering array at 15%cation-conservative weathering. We note again, however,that estimates of the absolute degree of cation-conservativeweathering implied by our modeling are dependent on boththe choice of starting composition and secondary mineralsformed. This nding of a lower degree of alteration is there-fore non-unique, and should not be taken as necessarilyinconsistent with the conclusions reached by Squyreset al. (2006b) regarding the degree to which the siliciclasticcomponent of the Burns Formation has been altered. Thesedifferences highlight the difficult problem of assessing theprecise chemical characteristics of the various components

in a ne-grained sedimentary rock using bulk chemical dataalone.In order to assess the quality of the results from our mix-

ing analysis, we recast the Burns Formation data onFig. 1D, which plots (FeO T + MgO + CaO + Na 2O + K 2-O)/SO 3 against SiO 2/SO 3 . In this space, data exhibiting atrue 2-component mixing relationship should display linearbehavior, and each data point should plot on the mixing ar-ray (dashed line) in the same relative order as they do onFig. 1C (Langmuir et al., 1978 ). We nd that the mixingrelationship between Gagarin and MacKenzie fails the lat-ter test for 2-component mixing, and that MacKenzie (thelow SO3 endmember in the mixing relationship) is displacedfrom the remaining samples of the Burns Formation. In to-

tal, there are six such samples that display anomalous geo-chemical behavior similar to MacKenzie. Four of them, theaforementioned MacKenzie , Inuvik , and DiamondJenness (which was analyzed after a rst RAT abrasionand again after a second, deeper abrasion) are analysesfrom RAT-ed rocks that lie beneath a recognized strati-graphic boundary in Endurance Crater (informally namedthe Wellington contact ). This contact separates the over-lying sand sheet and interdune facies from the dune facies atthe bottom of the stratigraphic section. Two additionalsamples, named Cha and Gilbert , also cluster withthese four samples; notably, Gilbert was the deepest sampleanalyzed in the stratigraphic section analyzed at VictoriaCrater. These six samples are all relatively depleted inSO3 and enriched in SiO 2, Al2O3 and K 2O. These traitsare not unexpected given that these samples lie closer tothe siliciclastic end of the mixing array on Fig. 1C, but thesesamples are also anomalously low in MgO (ranging from5.1 to 6.5 wt%), which displaces them from the remainingBurns Formation samples (MgO = 6.89.2 wt%) onFig. 1D, a trait noted previously by Clark et al. (2005).The CaO, Na 2O, and FeO T concentrations of these six sam-ples are indistinguishable from the rest of the Burns Forma-tion. These six samples all share a unique anddistinguishing textural trait: all appear to have undergonea recrystallization process that gave rise to the presence of

blocky isopachous cements, overgrowth precipitates onhematitic concretions, and nodular textures ( Clark et al.,2005; McLennan et al., 2005; Herkenhoff et al., 2008 ). Thissuggests that the loss of MgO in these samples may havebeen the result of diagenetic recrystallization processes.

Another attribute that these six samples have in com-mon is elevated Cl concentrations, ranging from 0.90 to1.90 wt%. Notably, however, there is another group of 9samples in the dataset that also have similar high Cl con-centrations, ranging from 1.45 to 1.70 wt%, but that other-wise exhibit no particularly unusual characteristics relativeto the remaining 21 samples of the Burns Formation, whichhave Cl concentrations between 0.46 and 0.91 wt%.

On Fig. 2, we show Microscopic Imager mosaics of theRAT-ed samples Gagarin ( Fig. 2A), Dramensfjord(Fig. 2B), and MacKenzie ( Fig. 2C) in order to illustratethe texture exhibited by the most SO 3-rich endmember(Gagarin), the texture associated with a high chlorine sam-ple (Dramensfjord), and the previously described blockyrecrystallization texture (MacKenzie). Here it can be seen

that MacKenzie is dramatically different in appearance.On Fig. 2D, we show that in (FeO T + MgO + CaO + Na 2-O + K 2O)/SO 3 versus SiO2/SO 3 space, the anomalous geo-chemical behavior of those samples that exhibit nodularrecrystallization textures is consistent with the loss of MgSO 4 .

Based on these geochemical and textural attributes, wehave separated the Burns Formation into Group 1 , Group 2 , and Group 3 on Fig. 1C and D. Group 1are normal Burns Formation samples, Group 2 are the9 samples with high chlorine abundance, but which donot exhibit disturbed geochemical characteristics orblocky, nodular textures, and Group 3 are the remainingsix samples that do exhibit such textural and geochemical

disturbance. For the remainder of the discussion we willignore the six Group 3 samples, as they appear not to faith-fully record the chemistry of the endmembers in the 2-com-ponent mixing relationship that denes the geochemistry of the rest of the Burns Formation (Groups 1 and 2). Finally,one other sample that is anomalously high in FeO T , namedPenrhyn, will also be ignored in our analysis.

We can now recalculate the mixing array using as analternate endmember the most SO 3-poor analysis that isnot offset from the linear array displayed on Fig. 1D(Cercedilla, 19.1 wt% SO 3, Group 1). An MI mosaic fromthis sample is also shown on Fig. 2 which shows no evi-dence of blocky, nodular texture. Returning to Fig. 1C,the updated mixing array (shown by a solid line) pointsto the same groundwater-derived endmember, but thesiliciclastic endmember could now plausibly be a materialhaving a chemical composition similar to that of modernbasaltic regolith. From a parsimony perspective, this is anenticing possibility: that the Burns Formation simply repre-sents a mixture between lightly weathered basaltic regolith,similar to that observed in abundance at the modern-daysurface of Mars, and a groundwater-derived evaporativesalt component that acted as a cement to bind the basalticgrains together. However, in light of the fact that modernbasaltic regolith at the MER sites contains abundant oliv-ine (modeled at 1015% of the modal mineralogy,

32 J.A. Hurowitz, W.W. Fischer/ Geochimica et Cosmochimica Acta 127 (2014) 2538

8/12/2019 Horowitz & Fischer 2014

9/14

McSween et al., 2010), yet Burns Formation sandstonescontain very little detectable olivine ( Morris et al., 2006 ),this simple relationship seems somewhat untenable. A cal-culation based on the average FeO T concentration of RAT-ed Burns Formation outcrop (15.9 wt% FeO T , Elec-tronic Annex ) and the average abundance of olivine in

RAT-ed Burns Formation (2% of all Fe is present as faya-lite, Morris et al., 2006 ), indicates a maximum of 0.32 wt%of FeO T in the Burns Formation could come from Fe inolivine. This is not consistent with a mixture in which mod-ern basaltic soil is a 50% component, because the expectedcontribution from Fe in olivine would amount to

0 1 2 3 4 50

1

2

3

4

SiO 2 / SO 3

5

( F e

O T

+ M g

O +

C a

O +

N a

2 O

+ K

2 O ) / S O

3

M g S O 4 R e m

o v a lMgSO 4

Fig. 2. Microscopic Imager mosaics of the RAT-ed Burns Formation targets Gagarin (A), Dramensfjord (B) and MacKenzie (C), taken onsols 403, 162, and 177, respectively. These samples are members of geochemical Group 1, Group 2, Group 3, respectively (see text for details).MacKenzie exhibits blocky, nodular textures, potentially associated with a diagenetic recrystallization event(s) that the other type examplesdid not experience. This recrystallization process also resulted in a change in the geochemical properties of Group 3 samples, most notably aloss of MgO, consistent with the removal of MgSO 4 , as shown on Fig. 2D, which plots (FeO T + MgO + CaO + Na 2O + K 2O)/SO 3 versusSiO2/SO 3. Fig. 2E is a microscopic imager mosaic, taken in partial shadow, of the Group 1 sample

Cercedilla on sol 1182. Each MI mosaiccontains 4 images, resulting in mosaics that are 5 cm 5 cm in size. Image credit: NASA/JPL/Cornell/USGS.

J.A. Hurowitz, W.W. Fischer/ Geochimica et Cosmochimica Acta 127 (2014) 2538 33

8/12/2019 Horowitz & Fischer 2014

10/14

2.13.1 wt% of FeOT (assuming 1015% olivine having thecomposition Fo 50 Fa 50 ). Therefore, we conclude that a silic-iclastic component having the composition of modernbasaltic regolith is an unlikely participant in the mixingrelationships observed.

More recently, Opportunity has begun exploration of Noachian bedrock of the Shoemaker Formation, whichunderlies the Burns Formation and is exposed in outcropsrimming the 22 km diameter Endeavour Crater. Thus far,the rocks encountered are reported to have textures consis-tent with impact-derived lithologies, including shallow sue-vite breccias and deeper lithic breccias, both of which mayhave been affected by hydrothermal alteration processes(Squyres et al., 2012). Analyses from these two lithologiesare plotted on Fig. 1E and F and reveal that these breccias,particularly the lithic breccias, are a plausible siliciclasticendmember for the Burns Formation. If correct, a geneticdetrital link between the Burns Formation and the underly-ing Shoemaker Formation can be drawn, indicating thatsediment derived from the Shoemaker Formation was ce-

mented and lithied by upwelling evaporating groundwaterto form the siliciclastic component of the Burns Formation.It should be noted that, to date, only one analysis of aRAT-ed Shoemaker Formation rock has been published;the remaining analyses are of undisturbed surfaces. Thereis a signicant offset between RAT-ed and undisturbed sue-vite breccia analyses (Fig. 1E), indicating that the undis-turbed surface analyses may not properly record the bulkcompositions of the rock interiors. Accordingly, furtherexploration of the possible linkages between the siliciclasticcomponent of the Burns Formation and the underlyingShoemaker Formation will await the results of Opportu-nitys future exploration of this new geologic terrane.

5. CONCLUSIONS

Geochemical data from weathered basaltic rock surfacesand soils provide important examples of a style of cation-conservative chemical alteration that occurs when acidiedwaters come into contact with basaltic substrates underwhat are very likely water-limited environmental conditions(e.g., Banin et al., 1997; Tosca et al., 2004; King andMcSween, 2005; Hurowitz and McLennan, 2007; Kingand McLennan, 2010 ). Under these conditions, alteringuid enters the system and interacts with rock and soil,but does not migrate away to a signicant extent, allowingfor the precipitation of secondary phases at or very nearbythe site of chemical alteration, and preserving the clear geo-chemical ngerprint of cation-conservative weathering pro-cesses. It seems sensible that for those interactions thatresult in S-addition, the pH of the uid was lower than istypical for terrestrial waterrock interactions, given thatsulfuric and sulfurous acids are stronger acids than car-bonic acid. It is highly likely, however, that on a planet witha CO 2 atmosphere and intermittent volcanic activity, waterthat was not acidied by S-volatiles has interacted withrock and soil. In such cases, it is reasonable to suspect thatother secondary mineral phases with compositions that theAPXS is not particularly sensitive to (e.g., carbonates) areformed. Under these conditions, sulfate minerals may be

recycled through the weathering process, and the long-termstability of a given secondary mineral assemblage will bedictated by factors such as: relative mineral solubilities,the integrated quantity of water and fresh acid to thatof soil or rock, and the degree to which redox-based acidchemistries have been titrated.

In marked contrast, the geochemical data collected fromthe ancient sedimentary rocks of the Burns Formation pre-serve evidence for a hydrological environment that wascapable of fractionating soluble cations from basalt, trans-porting them in a sulfur-rich solution, and precipitating sul-fate salts and silica that acted as a cementing agent for astill somewhat enigmatic, siliceous sediment that wasundergoing physical transport in a predominantly dry aeo-lian surface environment. An intriguing possibility raisedby our analysis is that the source of this sediment was re-cently discovered by the Opportunity rover in the impactbreccias of the Shoemaker Formation that underlies theBurns. It is important to note that study of terrestrial ana-logues, experiments, and models of Burns Formation chem-

istry and mineralogy have already demonstrated that undercertain conditions, acidic groundwaters are an expectedproduct of basaltgroundwater interaction on Mars ( Toscaet al., 2005, 2008; Baldridge et al., 2009; Marion et al., 2009;Hurowitz et al., 2010 ). Accordingly, our results supporthypotheses that the groundwater-derived component inthe Burns Formation obtained its chemical compositionas a result of interactions between S-charged groundwaterand the basaltic crust of Mars under open system hydro-logic conditions, fractionating elements along a path lengthfrom the catchment to the basin. These results do not sup-port those models in which the sedimentary rocks wereformed under water-limited, cation-conservative condi-tions. We conclude, therefore, that the Burn Formation re-

cords evidence of a Martian climate that was capable of supporting a hydrologic system characterized by ground(or surface) water recharge and transport, placing a valu-able temporal constraint on the activity and availabilityof water on the late Noachian to early Hesperian surfaceof Mars.

ACKNOWLEDGEMENTS

This research was carried out in part at the Jet Propulsion Lab-oratory, California Institute of Technology, under a contract withthe National Aeronautics and Space Agency (J.A.H.). This workwas supported by NASA award NNX10AM84G to J.A.H. Theauthors thank associate editor Penny King, Mikhail Zolotov, andtwo anonymous reviewers for their helpful comments, which sub-stantially improved this manuscript. The authors thank Ken Her-kenhoff (USGS) for producing the Microscopic Imager mosaicson Fig. 2, and Edwin Kite for feedback on an early draft of themanuscript.

APPENDIX A. SUPPLEMENTARY DATA

Supplementary data associated with this article can befound, in the online version, at http://dx.doi.org/10.1016/ j.gca.2013.11.021.

34 J.A. Hurowitz, W.W. Fischer/ Geochimica et Cosmochimica Acta 127 (2014) 2538

8/12/2019 Horowitz & Fischer 2014

11/14

REFERENCES

Amundson R., Ewing S., Dietrich W., Sutter B., Owen J.,Chadwick O., Nishiizumi K., Walvoord M. and McKay C.(2008) On the in situ aqueous alteration of soils on Mars.Geochim. Cosmochim. Acta 72 , 38453864.

Andrews-Hanna J., Zuber M., Arvidson R. and Wiseman S. (2010)

Early Mars hydrology: Meridiani playa deposits and thesedimentary record of Arabia Terra. J. Geophys. Res. 115 ,E06002, doi:06010.01029/02009JE003485.

Andrews-Hanna J. C., Phillips R. J. and Zuber M. T. (2007)Meridiani Planum and the global hydrology of Mars. Nature446 , 163166.

Baldridge A. M., Hook S. J., Crowley J. K., Marion G. M., KargelJ. S., Michalski J. L., Thomson B. J., de Souza C. R., BridgesN. T. and Brown A. J. (2009) Contemporaneous deposition of phyllosilicates and sulfates: using Australian acidic saline lakedeposits to describe geochemical variability on Mars. Geophys.Res. Lett. 36 , 6.

Banin A., Kan H. I. and Cicelsky A. (1997) Acidic volatiles and theMars soil. J. Geophys. Res. 102 , 1334113356.

Bibring J.-P., Langevin Y., Gendrin A., Gondet B., Poulet F.,

Berthe M., Soufflot A., Arvidson R. E., Mangold N., MustardJ. F. and Drossart P.OMEGA-Team (2005) Mars surfacediversity as revealed by the OMEGA/Mars express observa-tions. Science 307 , 15761581.

Blake D. F., Morris R. V., Kocurek G., Morrison S. M., Downs R.T., Bish D., Ming D. W., Edgett K. S., Rubin D., Goetz W.,Madsen M. B., Sullivan R., Gellert R., Campbell I., Treiman A.H., McLennan S. M., Yen A. S., Grotzinger J., Vaniman D. T.,Chipera S. J., Achilles C. N., Rampe E. B., Sumner D., MeslinP.-Y., Maurice S., Forni O., Gasnault O., Fisk M., Schmidt M.,Mahaffy P., Leshin L. A., Glavin D., Steele A., Freissinet C.,Navarro-Gonzalez R., Yingst R. A., Kah L. C., Bridges N.,Lewis K. W., Bristow T. F., Farmer J. D., Crisp J. A., StolperE. M., Des Marais D. J., Sarrazin P. and Team M. S. (2013)Curiosity at Gale Crater, Mars: characterization and Analysis

of the Rocknest Sand Shadow. Science 341 .Carr M. H. (2012) The uvial history of Mars. Philos. Trans. R.Soc. A: Math. Phys. Eng. Sci. 370 , 21932215.

Cassata W. S., Shuster D. L., Renne P. R. and Weiss B. P. (2012)Trapped Ar isotopes in meteorite ALH 84001 indicate Mars didnot have a thick ancient atmosphere. Icarus 221 , 461465.

Clark B. C., Baird A. K., Rose H. J., Toulmin P., Keil K., CastroA. J., Kelliher W. C., Rowe C. D. and Evans P. H. (1976)Inorganic analyses of Martian surface samples at the Vikinglanding sites. Science 194 , 12831288.

Clark B. C., Morris R. V., McLennan S. M., Gellert R., Joliff B. L.,Knoll A. H., Squyres S. W., Lowenstein T. K., Ming D. W.,Tosca N. J., Yen A., Christensen P. R., Gorevan S., BrucknerJ., Calvin W., Dreibus G., Farrand W. H., Klingelhoefer G.,Wanke H., Zipfel J., Bell, III, J. F., Grotzinger J.,

McSween, Jr., H. Y. and Rieder R. (2005) Chemistry andmineralogy of outcrops at Meridiani Planum. Earth Planet. Sci.Lett. 240 , 7394.

Clark B. C. and Van Hart D. C. (1981) The salts of Mars. Icarus45, 370378.

Edgar L., Grotzinger J., Bell III J. and Hurowitz J. (submitted forpublication) Hypotheses for the origin of ne-grained sedimen-tary rocks at Santa Maria Crater, Meridiani Planum. Icarus .

Ehlmann B. L., Mustard J. F., Murchie S. L., Bibring J. P.,Meunier A., Fraeman A. A. and Langevin Y. (2011) Subsurfacewater and clay mineral formation during the early history of Mars. Nature 479 , 5360.

Forget F. and Pierrehumbert R. T. (1997) Warming early Marswith carbon dioxide clouds that scatter infrared radiation.Science 278 , 12731276.

Gaidos E. and Marion G. (2003) Geological and geochemicallegacy of a cold early Mars. J. Geophys. Res. 108 , 19.

Gellert R., Rieder R., Bruckner J., Clark B. C., Dreibus G.,Klingelhoefer G., Lugmair G., Ming D. W., Wanke H., Yen A.,Zipfel J. and Squyres S. W. (2006) The Alpha Particle X-raySpectrometer (APXS): results from Gusev Crater and calibra-tion report. J. Geophys. Res. 111 . http://dx.doi.org/10.1029/2005JE002583.

Glotch T. D., Bandeld J. L., Christensen P. R., Calvin W. M.,McLennan S. M., Clark B. C., Rogers A. D. and Squyres S. W.(2006) Mineralogy of the light-toned outcrop at MeridianiPlanum as seen by the Miniature Thermal Emission Spectrom-eter and implications for its formation. J. Geophys. Res. 111 ,E12S03, doi:10.1029/2005JE002672.

Gorevan S. P., Myrick T., Davis K., Chau J. J., Bartlett P.,Mukherjee S., Anderson R., Squyres S. W., Arvidson R. E.,Madsen M. B., Bertelsen P., Goetz W., Binau C. S. and RichterL. (2003) Rock Abrasion Tool: Mars Exploration Rovermission. J. Geophys. Res. 108 , 8.

Greenwood J. P., Itoh S., Sakamoto N., Vicenzi E. P. andYurimoto H. (2008) Hydrogen isotope evidence for loss of water from Mars through time. Geophys. Res. Lett. 35 , 5.

Grotzinger J., Bell J., Herkenhoff K., Johnson J., Knoll A.,McCartney E., McLennan S., Metz J., Moore J., Squyres S.,Sullivan R., Ahronson O., ArvIdson R., Joliff B., GolombekM., Lewis K., Parker T. and Soclerblom J. (2006) Sedimentarytextures formed by aqueous processes, Erebus crater, MeridianiPlanum, Mars. Geology 34 , 10851088.

Grotzinger J. and Milliken R. (2012) In Sedimentary Geology of Mars (series eds. G.J. Nichols and B. Ricketts). SEPM SpecialPublication # 102. SEPM Society for Sedimentary Geology,Tulsa. p. 270.

Grotzinger J. P., Arvidson R. E., Bell, III, J. F., Calvin W., ClarkB. C., Fike D. A., Golombek M., Greeley R., Haldemann A. F.C., Herkenhoff K. E., Joliff B. L., Knoll A. H., Malin M. C.,McLennan S. M., Parker T., Soderblom L. A., Sohl-DicksteinJ. N., Squyres S. W., Tosca N. J. and Watters W. A. (2005)Stratigraphy and sedimentology of a dry to wet eoliandepositional system, Burns formation, Meridiani Planum,Mars. Earth Planet. Sci. Lett. 240 , 1172.

Halevy I., Zuber M. T. and Schrag D. P. (2007) A sulfur dioxideclimate feedback on early Mars. Science 318 , 19031907.

Haskin L. A., Wang A., McSween H. Y., McLennan S. M., CabrolN. A., Crumpler L. S., Jolliff B. L., Yen A. S., Clark B. C.,DesMarais D. J., Squyres S. W., Arvidson R. E., Ming D. W.,Morris R. V., Tosca N. J., Hurowitz J. A., Gellert R.,Klingelho fer G., Bell, III, J. F., Herkenhoff K. E., ChristensenP. R., Ruff S. W., Blaney D. L., Farmer J. D., Grant J. A.,Greeley R., Grin E. A., Landis G. A., Rice J. W., Richter L.,

Schroder C., Soderblom L. A. and de Souza, Jr., P. A. (2005)Water alteration of rocks and soils from the Spirit rover site,Gusev Crater, Mars. Nature 436. http://dx.doi.org/10.1038/nature03640 .

Hausrath E. M., Navarre-Sitchler A. K., Sak P. B., Steefel C. I. andBrantley S. L. (2008) Basalt weathering rates on Earth and theduration of liquid water on the plains of Gusev Crater, Mars.Geology 36 , 6770.

Hayes A. G., Grotzinger J. P., Edgar L. A., Squyres S. W., WattersW. A. and Sohl-Dickstein J. (2011) Reconstruction of eolian bedforms and paleocurrents from cross-bedded strata at VictoriaCrater, Meridiani Planum, Mars. J. Geophys. Res. 116 , 17.

J.A. Hurowitz, W.W. Fischer/ Geochimica et Cosmochimica Acta 127 (2014) 2538 35

8/12/2019 Horowitz & Fischer 2014

12/14

Herkenhoff K. E., Grotzinger J., Knoll A. H., McLennan S. M.,Weitz C., Yingst A., Anderson R., Archinal B. A., Arvidson R.E., Barrett J. M., Becker K. J., Bell J. F., Budney C., ChapmanM. G., Cook D., Ehlmann B., Franklin B., Gaddis L. R.,Galuszka D. M., Garcia P. A., Geissler P., Hare T. M.,Howington-Kraus E., Johnson J. R., Keszthelyi L., Kirk R. L.,Lanagan P., Lee E. M., Leff C., Maki J. N., Mullins K. F.,Parker T. J., Redding B. L., Rosiek M. R., Sims M. H.,Soderblom L. A., Spanovich N., Springer R., Squyres S. W.,Stolper D., Sucharski R. M., Sucharski T., Sullivan R. andTorson J. M. (2008) Surface processes recorded by rocks andsoils on Meridiani Planum, Mars: Microscopic Imager obser-vations during Opportunitys rst three extended missions. J.Geophys. Res. 113 , 39.

Hurowitz J. A., Fischer W., Tosca N. J. and Milliken R. E. (2010)Origin of acidic surface waters and the evolution of atmosphericchemistry on early Mars. Nat. Geosci. 3, 323326.

Hurowitz J. A. and McLennan S. M. (2007) A 3.5 Ga record of water-limited acidic weathering conditions on Mars. EarthPlanet. Sci. Lett. 260 , 432443.

Hurowitz J. A., McLennan S. M., McSween H. Y., DeSouza P. A.and Klingelhofer G. (2006a) Mixing relationships and theeffects of secondary alteration in the Wishstone and Watch-tower Classes of Husband Hill, Gusev Crater, Mars. J.Geophys. Res. 111 , 18.

Hurowitz J. A., McLennan S. M., Tosca N. J., Arvidson R. E.,Michalski J. R., Ming D. W., Schro der C. and Squyres S. W.(2006b) In-situ and experimental evidence for acidic weatheringof rocks and soils on Mars. J. Geophys. Res. 111 . http://dx.doi.org/10.1029/2005JE002515 .

Hynek B. M. and Phillips R. J. (2008) The stratigraphy of Meridiani Planum, Mars, and implications for the layereddeposits origin. Earth Planet. Sci. Lett. 274 , 214220.

Johnson S. S., Mischna M. A., Grove T. L. and Zuber M. T. (2008)Sulfur-induced greenhouse warming on early Mars. J. Geophys.Res. 113 , 15.

Johnson S. S., Pavlov A. A. and Mischna M. A. (2009) Fate of SO 2in the ancient Martian atmosphere: implications for transientgreenhouse warming. J. Geophys. Res.: Planets 114 , E11011.

King P. L. and McLennan S. M. (2010) Sulfur on Mars. Elements6, 107112.

King P. L. and McSween, Jr., H. Y. (2005) Effects of H 2O, pH, andoxidation state on the stability of Fe-minerals on Mars. J.Geophys. Res. 110 , E12S10, doi:10.1029/2005JE002485, 002005.

Kite E., Halevy I., Kahre M., Wolff M. and Manga M. (2013)Seasonal melting and the formation of sedimentary rocks onMars, with predictions for the Gale Crater mound. Icarus 223 ,181210.

Knauth L. P., Burt D. M. and Wohletz K. H. (2005) Impact originof sediments at the opportunity landing site on Mars. Nature438 , 11231128.

Kounaves S., Catling D., Clark B., DeFlores L., Gospodinova K.,

Hecht M., Kapit J., Ming D. and Quinn R. (2009) Aqueouschemistry of the Martian soil at the Phoenix landing site. 40thLunar and Planetary Science Conference . Houston, TX.

Kring D. A. (2007) The Chicxulub impact event and its environ-mental consequences at the CretaceousTertiary boundary.Paleogeogr. Paleoclimatol. Paleoecol. 255 , 421.

Langmuir C. H., Vocke R. D. J., Hanson G. N. and Hart S. R.(1978) A general mixing equation with applications to Icelandicbasalts. Earth Planet. Sci. Lett. 37 , 380392.

Marion G. M., Crowley J. K., Thomson B. J., Kargel J. S., BridgesN. T., Hook S. J., Baldridge A., Brown A. J., da Luz B. R. andde Souza C. R. (2009) Modeling aluminumsilicon chemistries

and application to Australian acidic playa lakes as analoguesfor Mars. Geochim. Cosmochim. Acta 73 , 34933511.

McCollom T. M. and Hynek B. M. (2005) A volcanic environmentfor bedrock diagenesis at Meridiani Planum on Mars. Nature438 , 11291131.

McGlynn I. O., Fedo C. M. and McSween H. Y. (2012) Soilmineralogy at the Mars Exploration Rover landing sites: anassessment of the competing roles of physical sorting andchemical weathering. J. Geophys. Res. 117 , 16.

McLennan S. M. (2003) Sedimentary silica on Mars. Geology 31 ,315318.

McLennan S. M., Bell, III, J. F., Calvin W. M., Christensen P. R.,Clark B. C., De Souza, Jr., P. A., Farmer J., Farrand W. H.,Fike D. A., Gellert R., Ghosh A., Glotch T. D., Grotzinger J.P., Hahn B. C., Herkenhoff K. E., Hurowitz J. A., Johnson J.R., Johnson S. S., Joliff B. L., Klingelhoefer G., Knoll A. H.,Learner Z. A., Malin M. C., McSween, Jr., H. Y., Pocock J.,Ruff S. W., Soderblom L. A., Squyres S. W., Tosca N. J.,Watters W. A., Wyatt M. B. and Yen A. (2005) Provenance anddiagenesis of the evaporite-bearing Burns formation, MeridianiPlanum, Mars. Earth Planet. Sci. Lett. 240 , 95121.

McSween H. Y., McGlynn I. O. and Rogers A. D. (2010)Determining the modal mineralogy of Martian soils. J.Geophys. Res. 115 , 10.

McSween H. Y., Ruff S. W., Morris R. V., Gellert R., KlingelhoferG., Christensen P. R., McCoy T. J., Ghosh A., Moersch J. M.,Cohen B. A., Rogers A. D., Schroder C., Squyres S. W., CrispJ. and Yen A. (2008) Mineralogy of volcanic rocks in GusevCrater, Mars: Reconciling Mossbauer, Alpha Particle X-raySpectrometer, and Miniature Thermal Emission Spectrometerspectra. J. Geophys. Res. 113 , 14.

McSween H. Y., Taylor G. J. and Wyatt M. B. (2009) ElementalComposition of the Martian Crust. Science 324 , 736739.

McSween H. Y., Wyatt M. B., Gellert R., Bell J. F., Morris R. V.,Herkenhoff K. E., Crumpler L. S., Milam K. A., Stockstill K.R., Tornabene L. L., Arvidson R. E., Bartlett P., Blaney D.,Cabrol N. A., Christensen P. R., Clark B. C., Crisp J. A., DesMarais D. J., Economou T., Farmer J. D., Farrand W., GhoshA., Golombek M., Gorevan S., Greeley R., Hamilton V. E.,Johnson J. R., Joliff B. L., Klingelhofer G., Knudson A. T.,McLennan S., Ming D., Moersch J. E., Rieder R., Ruff S. W.,Schroder C., de Souza P. A., Squyres S. W., Wanke H., WangA., Yen A. and Zipfel J. (2006) Characterization and petrologicinterpretation of olivine-rich basalts at Gusev Crater, Mars. J.Geophys. Res. 111 , 17.

Metz J. M., Grotzinger J. P., Rubin D. M., Lewis K. W., SquyresS. W. and Bell J. F. (2009) Sulfate-rich eolian and wet interdunedeposits, Erebus Crater, Meridiani Planum, Mars. J. Sediment.Res. 79 , 247264.

Michalski J. R. and Bleacher J. E. (2013) Supervolcanoes within anancient volcanic province in Arabia Terra, Mars. Nature 502,4752.

Michalski J. R., Cuadros J., Niles P. B., Parnell J., Rogers A.D. and Wright S. P. (2013) Groundwater activity on Marsand implications for a deep biosphere. Nat. Geosci. 6, 133 138.

Ming D. W., Gellert R., Morris R. V., Arvidson R. E., Brukner J.,Clark B. C., Cohen B. A., dUston C., Economou T., FleischerI., Klingelhofer G., McCoy T. J., Mittlefehldt D. W., SchmidtM. E., Schroder C., Squyres S. W., Treguier E., Yen A. S. andZipfel J. (2008) Geochemical properties of rocks and soils inGusev Crater, Mars: Results of the Alpha Particle X-RaySpectrometer from Cumberland Ridge to Home Plate. J.Geophys. Res. 113 , 28.

36 J.A. Hurowitz, W.W. Fischer/ Geochimica et Cosmochimica Acta 127 (2014) 2538

8/12/2019 Horowitz & Fischer 2014

13/14

Ming D. W., Mittlefehldt D. W., Morris R. V., Golden D. C.,Gellert R., Yen A., Clark B. C., Squyres S. W., Farrand W.H., Ruff S. W., Arvidson R. E., Klingelhoefer G., McSweenH. Y., Rodionov D. S., Schroder C., de Souza, Jr., P. A. andWang A. (2006) Geochemical and mineralogical indicators foraqueous processes in the Columbia Hills of Gusev Crater,Mars. J. Geophys. Res. 111 . http://dx.doi.org/10.1029/2005JE002560.

Morris R. V., Golden D. C., Bell, III, J. F., Shelfer T. D., ScheinostA. C., Hinman N. W., Furniss G., Mertzman S. A., Bishop J.L., Ming D. W., Allen C. C. and Britt D. T. (2000) Mineralogy,composition and alteration of Mars Pathnder rocks and soils:evidence from multispectral, elemental and magnetic data onterrestrial analogue, SNC meteorite, and Pathnder samples. J.Geoph