For Submission to the Journal of Geophysical Research Planets

1 Last edited December 28, 2014

2 For submission to the Journal of Geophysical Research – Planets

3 Modeling the Thermal and Physical Evolution of Mount Sharp’s Sedimentary

4 Rocks, Gale Crater, Mars: Implications for Diagenetic Minerals on the MSL

5 Curiosity Rover Traverse

7 Caue S. Borlina1,2, and Bethany L. Ehlmann2,3

9 1 Department of Atmospheric, Oceanic and Space Sciences, University of Michigan, Ann

10 Arbor, Michigan, USA

11 2 Division of Geological and Planetary Sciences, California Institute of Technology,

12 Pasadena, California, USA

13 3 Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California,

14 USA

16 Corresponding author: [email protected]; 2215 Space Research Building, 2455

17 Hayward St., Ann Arbor, MI 48109

19 Keywords: burial diagenesis, sedimentation, erosion, heat flow, Gale Crater, Mars

20 Science Laboratory (MSL)

23 24 Abstract

25 Gale Crater, landing site of the NASA Mars Science Laboratory (MSL) mission, has a

26 unique sedimentary stratigraphy, preserved in a 5 km high mound called Aeolis Mons

27 (informally known as Mt. Sharp). Understanding the processes of Mt. Sharp

28 sedimentation, erosion, and secondary mineral formation can reveal important

29 information regarding past geologic processes, the climate and habitability of early Mars,

30 and the organic preservation potential of these sedimentary deposits. Two endmember

31 scenarios for Mt. Sharp's formation are: (1) a complete filling of Gale crater followed by

32 partial removal of deposited sediments or (2) building of a central deposit with

33 morphology controlled by slope winds and only incomplete sedimentary fill of Gale

34 crater. Here we consider depths of burial predicted for both scenarios, coupled with a

35 thermal model to predict the temperature history of particular sediments presently

36 exposed at the surface. We compare our results with chemical and mineralogical analyses

37 by MSL to date and provide sedimentation scenario-dependent predictions of diagenetic

38 mineralogy in locations between Yellowknife Bay and upper Mt. Sharp. Results show

39 modeled erosion and deposition rates range from 5-42 μm/yr across different scenarios

40 and timing scales, consistent with or slightly higher than rates calculated for other

41 locations on Mars. Mineralogical predictions for locations close to Yellowknife Bay vary

42 with the choice of a cold or warm early Mars and the filling scenario. For (1),

43 temperatures experienced by sediments should decrease monotonically over the traverse

44 and up Mt. Sharp stratigraphy, whereas for (2) maximum temperatures are reached in the

45 lower units of Mt. Sharp and thereafter decline or hold roughly constant. If surface

46 temperatures on early Mars were similar to today, (1) predicts temperatures experienced 47 by sediments ranging from 50-100°C while for (2) only specificelect scenarios permit

48 diagenetic fluids (T>0°C) at select locations. In contrast, if surface temperatures on early

49 Mars were near the melting point of water, diagenetic temperatures were achieved over

50 the entire Curiosity traverse with temperatures ranging from 100-150°C under scenario

51 (1) and 0-100°C for variations of scenario (2). Under nearly all scenarios, evidence of

52 diagenesis at higher temperatures is expected as MSL moves towards Mt. Sharp.

53 Comparing our predictions with future MSL results on secondary mineral assemblages,

54 their spatial variation as a function of location on the traverse, and age of any authigenic

55 phases will constrain the timing and timescale of Mt. Sharp formation, paleosurface

56 temperature, and the availability and setting of liquid water on early Mars.

58 59 1. Introduction

60 The study of sedimentation processes at specific regions on Mars helps time order

61 aqueous mineral formation and defines the ancient geologic history of the red planet.

62 Such analysis can reveal important information about past habitability, presence of water

63 at or near the surface of the planet, and potential for the long-term preservation of organic

64 materials. Gale crater (137.4oE, -4.6oN), the landing site of the NASA Mars Science

65 Laboratory (MSL) mission, has a unique sedimentary stratigraphy, which permits

66 examining ancient Martian environmental conditions and aqueous alteration. Recent

67 studies have shown that Gale once hosted a fluvial-lacustrine environment, capable of

68 supporting life [Grotzinger et al., 2014]. The stratigraphic rock record of the area, and

69 thus its geologic history, is preserved in a 5 km high mound called Aeolis Mons

70 (informally, Mt. Sharp) located in Gale crater [Grotzinger et al., 2012] (Figure 1a).

72 Materials comprising Mt. Sharp have a low thermal inertia and subhorizontally layered

73 units, which implicate a sedimentary origin [Pelkey et al., 2004; Anderson & Bell, 2010;

74 Thomson et al., 2011]. More recent work has suggested the lowermost units are cross-

75 bedded sandstones formed by cementation and lithification of sand dunes [Milliken et al.,

76 2014]. The lower mound includes distinctive sedimentary beds with hydrated sulfates,

77 iron oxides and Fe/Mg smectite clay minerals [Millken et al., 2010; Fraeman et al.,

78 2014]. Spectral signatures associated with the upper mound do not permit unique

79 identification of secondary mineral phases, though eroded boxwork structures suggest

80 precipitation of minerals during fluid flow within the sediments [Siebach and Grotzinger,

81 2014]. 82

83 Many hypotheses have been developed to explain Mt Sharp formation, variously

84 invoking airfall dust or volcanic ash, lag deposits from ice/snow, aeolian and

85 fluviolacustrine sedimentation [for review, see Anderson & Bell, 2010; Wray, 2013;

86 LeDeit et al., 2013]. Nevertheless, regardless of the process(es) delivering sediments, two

87 endmember scenarios describe the time-evolution, i.e., growth and subsequent erosion, of

88 Mt. Sharp. In Scenario 1, Gale Crater was completely filled with layered sediments then

89 partially exhumed, leaving a central mound [Malin and Edgett, 2000] (Figure 1b). In

90 Scenario 2, an aeolian process characterized by slope winds created a wind-topography

91 feedback enabling growth of a high mound without complete fill of the crater [Kite et al.,

92 2013] (Figure 1c).

94 The history of Mt. Sharp’s sedimentation and mineralization provides key constraints on

95 environmental conditions on early Mars, including the availability of liquid water and the

96 nature of geochemical environments. Understanding sediment deposition and possible

97 diagenesis is also crucial to establishing the preservation potential through time for

98 organic carbon of biological or abiotic origin trapped in sedimentary rock strata. The

99 thermal history of sediments and their exposure to fluids exerts strong control on the

100 persistence of organic compounds in the sedimentary record (e.g., Harvey et al., 1995;

101 Lehmann et al., 2002). Initial studies of the diagenesis of Martian sediments pointed out

102 the apparent ubiquity of the “juvenile” sediments with smectite clays, amorphous silica

103 and a paucity of evidence for illite, chlorite, quartz and other typical products of

104 diagenesis, which are common in the terrestrial rock record, thus concluding that 105 diagenetic processes on Mars were uncommon, perhaps limited by water availability

106 [Tosca & Knoll, 2009]. Since then, a growing number of studies have identified clay

107 minerals such as illite, chlorite, and mixed layer clays that can form from via diagenesis

108 [Ehlmann et al., 2009, 2011a, 2011b; Milliken & Bish, 2010; Carter et al., 2013]. So far,

109 minerals identified in Mt. Sharp from orbit do not include these phases. However, in situ

110 rover data at Yellowknife Bay implies at least early diagenetic reactions to form

111 mineralized veins, nodules, and filled fractures within the mudstones [Stack et al., 2014;

112 Siebach et al., 2014], including exchange of interlayer cations in smectite clays or

113 incipient chloritization [Vaniman et al., 2014; Rampe et al., 2014; Bristow et al., 2014].

115 Here we model the potential diagenetic history of the sediments comprising Mt. Sharp

116 and accessible in rock units along Curiosity’s traverse. We couple the two sedimentation

117 scenarios [Malin and Edgett, 2000; Kite et al., 2013] with a thermal model for ancient

118 Martian heat flow and timescales for Mt. Sharp sedimentary deposition and erosion

119 constrained by crater counts. We model temperature variations experienced within the

120 region between Yellowknife Bay, the base of Mt. Sharp, and the lower unit/upper unit

121 Mt. Sharp unconformity, and we compare them with select key temperature thresholds

122 relevant for diagenesis: stability of liquid water (0oC) and the a temperature for efficient

123 smectite to illite conversion (~40oC). We also analyze the time-temperature integral, an

124 alternative method for establishing the likelihood and extent of mineral diagenesis as

125 described by Tosca and Knoll [2009]. The final results are compared to mineralogical and

126 chemical findings obtained by MSL to date. Potential mineralogical findings across

127 MSL’s traverse are discussed and correlated with implications for sedimentation 128 processes on Mars, timescales, early Mars temperatures, liquid water availability, and the

129 organic preservation potential of Gale sediments.

131 2. Methodology

132 In this section we describe how we defined the pristine Gale basement and final Mt.

133 Sharp profile, the two different endmember sedimentation scenarios and variations in

134 parameters used during the modeling, timescales and timing of sedimentation assumed,

135 the derivation of the thermal model for surface heat flow, and the topographic profile of

136 the rover traverse that defines the studied region.

138 2.1 Pristine Gale Basement & Modern Topography

139 In order to evaluate how Gale Crater changed we needed to first determine the ancient

140 (starting point) and modern (ending point) topographies of the crater. Gale crater has

141 been both eroded and filled relative to its original topographic profile. Consequently, a

142 combination of theoretical constraints and observational constraints from the Mars

143 Orbiting Laser Altimeter Altitude (MOLA) dataset was used to set the initial conditions

144 for Gale's crater shape, i.e. its initial topographic profile. Observed crater depth-diameter

145 relationships on Mars predict a range of initial crater depths for Gale crater (D ~154 km)

146 ranging from 4.2 km to 5.4 km [Garvin et al., 2003; Boyce and Garbeil, 2007; Robbins

147 and Hynek, 2007]. Kalynn et al. [2013] empirically have shown Martian central peak

148 heights of ~1 km for ~100 km craters. Our examination of craters on Mars, better

149 preserved than Gale, with diameters ranging from 131 km to 155 km (at 16ºW, 43ºS;

150 36ºW, 36ºS; 45ºE, 42ºN) yielded lower bounds on central peak height ranging from 1.0- 151 2.1 km that did not scale in a straightforward way with diameter, and these central peak

152 heights may be underestimates due to later crater fill. Hence, we set the initial shape of

153 Gale Crater to be 154 km in diameter and 5 km deep with a central peak height of 1.55

154 km. In order to have realistic slopes, we scaled the average topographic profile from

155 Moreux crater (45E, 42N) to fit Gale's parameters.

157 Modern Gale crater has a highly asymmetric central mound, its cross-sectional profile

158 varying with azimuth. The average Gale profile shown (Figure 2) was compiled from

159 multiple roughly NW-SE cross-sections of present-day crater topography and then

160 averaged after removing local geologic features. The average Gale profile is used for

161 estimation of an average overburden over the Curiosity rover traverse. Scenarios 2a and

162 2b were tuned to produce final mounds of width approximately equivalent to the width of

163 the average profile.

165 2.2 Mt. Sharp Sedimentation Scenarios

166 We develop two geological scenarios for the time evolution of Mt. Sharp filling/removal:

167 (1) a complete filling of the crater followed by partial removal leaving a central mound,

168 [Malin and Edgett, 2000]; and (2) an aeolian process, where the mound only grew near

169 the center of the crater with continuous presence of strong mound-flank slope winds

170 capable of eroding the mound and keeping it away from the sides of the crater [Kite et al.,

171 2013].

173 2.2.1 Scenario 1 174 Scenario 1 is characterized by a complete fill of the crater to the height of the highest

175 peak of Mt. Sharp, followed by partial erosion of the crater, leaving the modern shape of

176 Mt. Sharp as final output. We use an average Gale profile as the final shape (see section

177 2.1). This scenario is strongly dependent on timescales defined in section 2.3.

178 Sedimentation and erosion rates are computed linearly based on the defined time period

179 for erosion/sedimentation processes and necessary burial/erosion height, i.e., deposition

180 rate is computed as distance from pristine basement to the top of the crater, divided by

181 deposition time. Erosion rate computed as distance from top of the crater (completely

182 filled crater) to the average modern profile, divided by erosion time.

184 2.2.2 Scenario 2

185 Scenario 2 is an aeolian process where the mound grew close to the center of the crater

186 with the surrounding topography creating an environment for the presence of strong

187 mound-flank slope winds capable of eroding the mound. We use the model and code

188 developed by Kite et al. [2013] to generate the evolution of the topographic profile with

189 time. Briefly, in Kite et al.’s model a saltation model approximation is used to determine

190 the balance between the deposition D (set at the beginning of the simulation), and the

191 erosion E (time varying). The result is the computation of dz/dt (altitude variation over

192 time) for every time step. The following set of equations define the scenario

(1) 193

(2) 194

(3) 195 196 where k is an erodibility factor; α is a parameter responsible for sand transport, soil

197 erosion and rock abrasion on the crater; U is a unitless dimensionless expression related

198 to the magnitude of the shear velocity (dx’, a mesh unit in Kite et al., is 1i.e. at t = 0 the

199 basement is represented by a mesh with spacing dx’ of unit value); U0 is the component

200 of background bed shear velocity; z’ is the height; x is the location within the crater (0 <

201 x < 154 km, the crater diameter); x’ is the distance half-way between the values of x

202 starting at x = 1.5 km and ending at x = 153.5 km; and L is a length scale that accounts for

203 the effects of inertia. A special parameter R/L is used which correlates the width of the

204 crater wall to the crater size. For small craters we should expect small values of R/L,

205 around 0.1-1 (for a complete analysis of R/L verify the supplementary material from Kite

206 et al., [2013]).

208 Is important to notice that . U and z’ are tiime varying values. Eq. (3) is evaluated for

209 each time step considering left and right slopes of the winds: for a value of x, we compute

210 the integral for left slopes (values less than x) and right slopes (values greater than x and

211 less than 154 km). The operator max[ ] selects the slope with the highest value, which

212 then permits evaluation of Eq. (2) at each time step. Eq. (1) is also then evaluated in each

213 timestep, using E from Eq. (2) and a value for D determined from the user-set parameter

214 D’ (D=D’E0, where E0 is the initial erosion at the start of the simulation)). In Kite et al.

215 [2013] and our Scenario 2a, D is fixed in the first time step and is constant thereafter. In

216 our Scenario 2B, D is set initially then declines linearly.

217 We experimented with multiple choices of parameter combinations and choose to

218 work with two different sets of parameters that yield topographic profiles most similar to 219 Gale. Parameters used by Kite et al. [2013] and this study are given in Table 1. We name

220 the two different scenarios as 2a and 2b. Scenario 2a is defined with a constant deposition

221 rate. The mound grows tall with a relatively constant width over time. Scenario 2b has a

222 linearly decreasing deposition rate over time, and the mound grows wide then steepens

223 and narrows. Both scenarios converge on a similar final output though the amount of

224 sediment overburden as a function of time depends on the intermediate steps. The output

225 has no time dependency; we implement timescales in the final output of the model by

226 scaling the output to our specified durations after iterations converged. A special

227 parameter R/L is introduced at this point, which correlates the width of the crater wall to

228 the crater size. For small craters we should expect small values of R/L, around 0.1-1 (for

229 a complete analysis of R/L verify the supplementary material from Kite et al., [2013]).

231 2.3 Timescales

232 Crater counts on the ejecta blanket of Gale crater constrain its formation age to Late

233 Noachian/Early Hesperian, approximately 3.8 to 3.6 Ga, and place an upper limit on the

234 period of infilling Mt. Sharp sediments [Thomson et al., 2011; Le Deit et al., 2013].

235 Similarly, an upper limit to the age and extent of lower Mt. Sharp can be obtained using

236 the superposition relationship of the topographically lower but stratigraphically higher

237 deposits of Aeolis Palus, which have estimated ages ranging from early Hesperian to

238 early Amazonian [Thomson et al., 2011; Le Deit et al., 2012; Grant et al., 2012], i.e.,

239 from ~3.2 to ~3.5 Ga. Thus, most of the formation and erosion of Mt. Sharp to its present

240 extent took place during the Hesperian, though processes continued to shape the form of

241 the mound during the Amazonian. 242

243 While surface ages based on crater counts are superb for relative age dating, pinning in

244 absolute time is challenged by the existence of different chronology models relating the

245 density of craters with time [e.g., Werner & Tanaka, 2011]. Yet, numerical ages

246 constraining the start and end of major episodes of Mt. Sharp erosion and deposition are

247 required to tie burial history to models of the secular cooling of Mars (section 2.4).

248 Consequently, we examine three different temporal scenarios for the fill and exhumation

249 of Mt. Sharp: (1) a standard model, (2) a maximum diagenesis model, and (3) a minimum

250 diagenesis model. For (1), Mt. Sharp formation begins at 3.7 Ga, reaches 5 km in height,

251 and is then exhumed to reach approximately its present extent by 3.3 Ga. For (2), Gale

252 crater and Mt. Sharp form early, 3.85 Gyr, and Mt. Sharp is exhumed late, 3.0 Gyr, thus

253 providing a maximum for heat flow and duration of burial. For (3), Mt. Sharp forms late,

254 3.6 Gyr and is quickly exhumed by 3.4 Gyr.

256 2.4 Thermal Model

257 The thermal model used here defines temperature as a function of depth and time. We

258 construct it by using the one-dimensional steady-state heat conduction solution that

259 describes the temperature T as a function of the depth, z, and time, t, as

(1) 260 where T0 is mean surface temperature, q(t) is the heat flow as a function of time, k is the

261 thermal conductivity, ρ is the density and H(t) the heat production as a function of time.

262 We define ρ = 2500 kg/m3, k = 2 W/mC, values typical for average sedimentary rocks on

263 Earth [Beardsmore and Cull, 2001]. 264 The mean surface temperature for early Mars is still an unknown. Here we adopt

o o 265 two possibilities: T0 = 0 C and T0 = -50 C in order to analyze how different values of T0

266 can impact the results. The first presumes a warmer early Mars where temperatures

267 routinely exceed the melting point of water during large portions of the Martian year; the

268 latter represents the modern-day average equatorial temperature. Finally, q(t) and H(t)

269 were estimated by curve-fitting to geophysical models for the evolution of heat flow

270 [Parmentier and Zuber, 2007] and crustal heat production, respectively, through time

271 [Hahn et al., 2011; their Figure 2]. Results for q(t) and H(t) are shown in Figures 3a and

272 3b. Parmentier and Zuber [2007] estimated values for q to be ~60mW/m2 at 3.5 Gyr for a

273 variety of cooling scenarios and 20mW/m2 at present; measurements of modern heat flow

274 will soon be obtained by the InSight mission.. These values are similar to those

275 independently predicted by Morschhauser et al. [2011]. Figure 3c shows derived

276 temperatures as a function of depth for the two different mean surface temperatures and

277 several time periods.

279 Gale may have had additional heating from hydrothermal sources such as residual

280 heat following the Gale-forming impact or local volcanic sources, but we do not include

281 these in our modeling. Were additional sources of heat present, heat flow would be higher

282 and temperatures higher. Thus our estimates for temperatures reached are lower bounds.

284 2.5 Yellowknife Bay to Mt. Sharp: MSL’s Traverse

285 After landing, MSL headed toward Yellowknife Bay, a region in the opposite direction of

286 the expected path toward Mt. Sharp. The rocks at Yellowknife Bay preserve evidence for 287 a fluvio-lacustrine environment [Grotzinger et al., 2014] and several minerals related to

288 aqueous alteration, including Mg smectites and hydrated calcium sulfates, were identified

289 [Vaniman et al., 2014].

291 Given sections 2.1-2.4 above, we model the sedimentary and thermal history along the

292 Curiosity traverse, obtained between Yellowknife Bay, the Murray buttes entrance to Mt.

293 Sharp, and predicted future locations of MSL Curiosity after climbing Mt. Sharp. Figure

294 4 shows Bradbury Landing, Yellowknife Bay, Hidden Valley and a potential future final

295 location of MSL. We picked the final future location for modeling to be the

296 unconformity, marking the boundary between upper and lower Mt Sharp. In order to

297 locate Yellowknife Bay in our model, we compute the ratio between the distance of the

298 closest rim to Yellowknife Bay and the distance of the rim to the foothill of Mt. Sharp.

299 This is necessary because different models output mounds of different widths, and proper

300 location of the rover relative to the mound is crucial for computation of overburden rates.

301 The ratio computed here is ~0.93. For the final output of our models we identify

302 Yellowknife Bay at 0.93 of the distance between the closest rim and the foothill of the

303 output mound. Yellowknife Bay and Hidden Valley (~ Sol 722) have similar altitudes

304 and rim distances. The unconformity is ~1000 m higher than Yellowknife Bay: we then

305 conclude that ~ Sol 722 is represented by the location that is 1000 m higher than the

306 location set to be Yellowknife Bay (at the 0.93 ratio to the crater). and XX further from

307 the crater rim toward the interior of the crater. Results for the thermal history are

308 subsequently presented as a range of values between Yellowknife Bay and the

309 unconformity in Mt. Sharp. This range represents MSL’s future traverse range. 310

311 2.6 Key diagenetic thresholds

312 Once deposited, subsequent fluid circulation through sediments can lead to textural

313 changes as well as alteration of existing minerals and formation of new minerals.

314 Phyllosilicates, sulfates, iron oxides, and silica phases may form and/or undergo

315 mineralogic transitions, depending upon the chemistry of diagenetic fluids and

316 temperature. It is beyond the scope of this work to track all diagenetic transitions [for

317 review, see Mackenzie, 2005]; we instead focus on reporting modeled temperatures. For

318 reference, however, we do track three key diagenetic thresholds: the freezing point of

319 water at 0ºC, the illite to smectite conversion at ~40ºC, and the time-temperature integral.

321 Smectites are swelling clays that form during the reaction of water with silicates. At

322 elevated temperatures, however, smectites are no longer the thermodynamically favored

323 phase, and conversion of smectites to other phyllosilicate phases like illite or chlorite that

324 lack interlayer water begins. The temperature, the fluid composition, the amount of water

325 available, and burial duration are dictate resultant mineralogy. Given the presence of

326 smectite clay, either from in situ formation [Vaniman et al., 2014; Bristow et al., 2014] or

327 transport from a watershed with smectite clays [Ehlmann & Buz, accepted] in

328 sedimentary rocks at Gale crater, we track two thresholds for determining whether

329 smectite would be expected to have converted to another phase: a single temperature

330 threshold (40ºC) commonly cited for the smectite-illite conversion temperature [CITE]

331 and a time-temperature integral with thresholds based on the terrestrial clay mineral rock

332 record [Tosca & Knoll, 2009]. Tracking the 0ºC threshold also provides a useful 333 parameterization of water availability for alteration processes (if it is present). Fluids

334 may, of course, be available at lower temperatures due to freezing point depression from

335 dissolved salts.

338 3. Results

339 In this section we provide results for the scenario modeling erosion/deposition rates,

340 followed by temperature results for different scenarios and locations. , and Wwe conclude

341 by presenting the time-temperature integrals for the different models.

343 3.1 Topographic Evolution and Erosion/Deposition Rates

344 Figures 5a and 5d show the evolution of topography and overburden values for scenario

345 1, complete fill and exhumation, under the standard timing model; maximum diagenesis

346 and minimum diagenesis models have the same overall evolution, albeit with

347 temperatures achieved at different points in time. Figure 5 also shows the same

348 information for Sscenarios 2a (Figure 5b and 5e) and 2b (figure 5c and 5f). For the

349 topography variation the blue line represents the most ancient topography, while the red

350 line represents the most recent one. In Figure 5, a solid line represents Yellowknife Bay

351 while a dashed line represents the unconformity described in section 2.5.

353 If the complete filling and exhumation to present-day topography is assumed (Scenario

354 1), with the time period equally split into an interval of net erosion followed by interval

355 of net deposition, the average rates of erosion and deposition are 6-27 µm/yr and 8-34 356 µm/yr, respectively. Assuming a model where strong mound-flank slope winds are

357 capable of eroding the mound (Scenario 2), erosion and deposition rates were calculated

358 based on where elevation gradients were either negative (erosion) or positive

359 (deposition). For Scenario 2a average erosion rates fall between 5-21 µm/yr, while

360 average deposition rates range from 5-21 µm/yr. For Scenario 2b average erosion rates

361 are within 7-29 µm/yr while deposition rates range from 8-35 µm/yr. Table 2 summarizes

362 the obtained results.

364 3.2 Temperature

365 The results obtained in this section are based on coupling the topographic evolution of

366 Mt. Sharp with the thermal model for Mars’ changing geothermal gradient. Figure 6

367 shows the temperature variation with time for two locations (Yellowknife Bay and Mt.

368 Sharp), considering three different timing scenarios and three different sedimentary

369 models. Figures 6a, 6b and 6c show results for the different timing constraints

370 considering a cold early Mars (-50oC), while Figures 6d, 6e and 6f show the same results

371 but for a warm early Mars (0oC).

373 As a general observation, the timing has little effect on the peak temperatures achieved

374 and overall range of temperatures experienced by the sediments. It does, however,

375 significantly affect the time-temperature integral, which determines the expected degree

376 of diagenesis under Tosca & Knoll [2009]’s suggested approach to evaluating sediment

377 maturity. Our summary of results below assumes ground waters or ice were present. For

378 more on this issue, see section 4.2, discussion. 379

380 3.2.1 Cold Early Mars

381 For a cold early Mars, scenario 1 leads to conditions where water would be liquid and

382 illite could form from smectite everywhere from Yellowknife to Mt. Sharp. Maximum

383 temperatures reached are ~100ºC and ~70ºC at Yellowknife Bay and Mt. Sharp,

384 respectively (Figure 6a, 6b and 6c). Scenario 2a does not generate conditions above 0ºC;

385 there would be no alteration or diagenesis in situ unless driven by salty brines. Scenario

386 2a implies that swelling clays (if formed by another process and part of the sedimentary

387 deposit) should be the dominant clay from Yellowknife Bay and Mt. Sharp. Scenario 2b

388 predicts liquid water at Mt. Sharp that could facilitate diagenetic transitions, though

389 temperatures above 40ºC are not reached. Under Sscenario 2b, at Yellowknife Bay the

390 0ºC threshold is not reached, and only freezing point-depressed brines are permitted by

391 the temperature data.

393 Using the time-temperature integral as a metric of diagenesis (Figure 7a, 7b and 7c),

394 Scenario 1 predicts the complete transformation of smectite to illite at Yellowknife Bay

395 for the standard and maximum diagenesis models and at Mt. Sharp for the maximum

396 diagenesis model only. In all other Scenario 1 cases, transformation of smectite to illite

397 occurs, but is partial. Scenario 2b implies partial smectite to illite conversion at Mt.

398 Sharp, but not at Yellowknife Bay. Scenario 2a does not predict that any conversion

399 process occurs.

401 3.2.2 Warm Early Mars 402 For warm early Mars, liquid water that might cause alteration and diagenesis would be

403 available everywhere between Yellowknife Bay and Mt. Sharp under all scenarios

404 (Figures 6d, 6e and 6f). Maximum temperatures greater than 100ºC occur at both

405 Yellowknife Bay and Mt Shap in scenario 1. In both Sscenarios 2a and 2b, temperatures

406 above 40ºC are reached at Mt. Sharp, while in Sscenario 2b, this threshold is only

407 exceeded for Mt. Sharp.

409 Using the time-temperature integral, complete conversion of smectite to illite is predicted

410 for all scenarios and locations except for 2a at Yellowknife Bay. Surface temperatures at

411 0 o C would imply that any smectites should not have persisted elsewhere from the base to

412 unconformity of Mt. Sharp. Plots are shown in Figures 7d, 7e and 7f.

414 4. Discussion

415 In this section we present a discussion based on the comparison of our results with the

416 prior work, including MSL Curiosity data, acquired to date. We start by discussing

417 predicted erosion and deposition rates. We then proceed to describe the current

418 knowledge regarding the presence of clays throughout MSL’s path, and predictions for

419 what the rover might encounter under each scenario.

421 4.1 Martian Deposition and Erosion Rates

422 Overall the calculated averages obtained in section 3.1 and summarized in Table 2 fall

423 near the ranges of 10-100 µm/yr for estimated Martian sedimentary deposition [Lewis &

424 Aharonson, 2014]. The calculated average ranges moderately exceed estimated erosion 425 rates of Noachian and Hesperian terrains, 0.7-10 µm/yr, but are at the lower end of

426 erosion rates from Earth, 2-100µm/yr [Golombek et al., 2006]. Notably, typical Martian

427 net erosion rates are not compatible with the growth and erosion of Mt. Sharp entirely

428 during the Hesperian, as suggested by surface ages based on crater counts; at ~1 µm/yr,

429 the erosion of 4 km of material would take 4 Gyr. Having erosion and deposition rates

430 falling within a reasonable range derived from the literature confirms the plausibility of

431 our assumptions and the models described in this work.

433 It should be noted that deposition/erosion under Scenario 2a and 2b is computed

434 somewhat differently than for Scenario 1 in which the erosion and deposition and equally

435 scaled to add and then remove the necessary materials over the specified time period.

436 While in Scenario 2, a discrete erosion rate is computed in every time step iteration (, the

437 Kite et al. [2013] model computes deposition rate using a constant value based on the

438 initial time step that does not change during the simulation, regardless of Mt Sharp form).

439 For scenario 2a, this assumption is used; however, with our choice of crater wall height, 5

440 km instead of 10 km in Kite et al. [2013], this leads to a thin central mound. For scenario

441 2b, we decrease the deposition rate linearly with time. Several behaviors of deposition

442 rate could be used here. For a distinct endmember, we choose one that yielded results

443 similar to what we see at Gale, e.g. a mound height and shape that matches modern

444 topography and that also steepens with time and retreats back from the wall toward the

445 center, a sequence originally proposed in Kite et al. [2013]. A step function could be

446 used here as well: a simulation of volcanic interference while the aeolian process happens 447 inside the crater. Our analysis with this case showed a final form similar to the constant

448 deposition case.

450 4.2 Mineralogical Implications from Ground Measurements and Modeling

451 4.2.1 Yellowknife Bay

452 Vaniman et al. [2014] identified trioctahedral smectites at Yellowknife Bay in the

453 samples John Klein and Cumberland. No illites were positively recognized. Interestingly,

454 XRD patterns indicate that the smectite interlayers in John Klein are collapsed, while

455 those in Cumberland are held open, perhaps by metal-hydroxides [Bristow et al., in

456 press]. This suggests an additional episode(s) of fluid interaction with the smectite clays

457 after formation. This is consistent with the overall history of the Yellowknife Bay

458 sedimentary rock which shows, raised ridges, nodules, and calcium sulfate veins inferred

459 to result from post-depositional processes [Grotzinger et al., 2014; Siebach et al., 2014;

460 Stack et al., 2014].

462 Our analysis of models for the temperatures experienced by the Yellowknife Bay

463 sediments indicates either Yellowknife Bay was either never buried by >1 km of fill and

464 mean Hesperian Mars surface temperatures were substantially below zero or water was

465 unavailable during most of Yellowknife Bay’s burial. Any scenario where Yellowknife

466 Bay is buried under 5 km of fill (Scenario 1) leads to conversion of smectite to illite if

467 water is available. The conversion is complete in all except the minimum diagenesis, Tsurf

468 = 50 C case. Scenarios where the crater is partially filled (Scenarios 2a, 2b; up to XX km)

469 do not predict any smectite to illite conversion if average mean surface temperatures were 470 substantially below 0ºC because, even with burial, subsurface temperatures do not exceed

471 0ºC. In a warmer early Mars, where Tsurf ≥ 0ºC, partial to complete conversion of smectite

472 to illite would be expected at Yellowknife Bay, even with partial burial, so long as water

473 is available.

475 Thus, if the deposition of the Yellowknife Bay sediments pre-dates Mt. Sharp formation,

476 a cold early Mars and slope wind model with relatively little burial is implicated.

477 However, the stratigraphic relationships between Yellowknife Bay sediments and other

478 units are ambiguous, and the deposit might instead represent a late-stage unit from Peace

479 Vallis alluvial activity emplaced atop materials left behind during Mt. Sharp’s retreat.

480 Continued analysis of orbital and in situ data to understand contact relationships is

481 crucial. Moreover, these data highlight the importance of analysis of the mineralogy of

482 Mt. Sharp sediments where the stratigraphic relationships are clear to determine the

483 environmental history of Gale crater deposits.

485 4.2.2. Implications for Diagenesis along the Traverse and Lower Mt. Sharp

486 As Curiosity continues its trek toward Mt Sharp, continued collection of mineralogic data

487 will test whether paleotemperatures were higher or lower in Mt. Sharp compared to

488 Yellowknife Bay and whether the smectite to illite conversion took place. Figure 8

489 provides maximum temperatures experienced sedimentary rocks at locations between

490 Yellowknife Bay and Mt. Sharp, considering different scenarios and timing scales.

491 Scenario 1 has a maximum temperature at every location that decreases from

492 Yellowknife Bay to Mt. Sharp. Scenario 2a has an increasing maximum temperature 493 across MSL’s traverse, and scenario 2b has a peak maximum temperature located

494 somewhere between Yellowknife Bay the Mt. Sharp unconformity, near an elevation of

495 2.6 km. These distinctive patterns provide crucial, testable constraints for the history of

496 sedimentary deposition because these patterns of temperature increase would be recorded

497 in the sediments, if there were waters within Gale crater during the Hesperian. If Gale

498 crater was completely filled then exhumed, diagenesis should decrease with distance up

499 Mt. Sharp, whereas incomplete fill under a slope wind model scenario would show

500 increased diagenesis toward the center part of the mound. If temperature can also be

501 discerned from diagenetic mineral assemblages, these constrain Hesperian surface

502 temperatures, and thus paleoclimate, at Gale crater.

504 5. Conclusions

505 Understanding the conditions that led to sediment deposition, erosion, and secondary

506 mineral formation in Gale crater advances our understanding of the evolution of

507 environmental conditions on ancient Mars. We developed a coupled sedimentation-

508 thermal model that traced temperature as a function of time for two different scenarios: 1)

509 complete fill of Gale crater followed by partial erosion and 2) partial fill dictated by slope

510 winds as described by Kite et al., [2013]. Determining maximum temperatures of

511 sediments and computation of time-temperature integrals provides a set of distinctive

512 predictions for the extent of diagenesis, testable by what we are seeing today by the MSL

513 mission. Our models have erosion and deposition rates within expected Martian

514 deposition and erosion values. We find that in the presence of water during Mt. Sharp

515 formation and erosion, the smectite to illite conversion should have occurred at all levels 516 in lower Mt. Sharp if Gale crater had been completely filled. Moreover, either

517 Yellowknife Bay was never buried by >1km of overburden, or there was insufficient

518 water available for diagenesis when it was buried, due to either aridity or Hesperian mean

519 annual surface temperatures well below zero. Complete fill of Gale crater means the

520 rover should observe decreasing diagenesis as it climbs Mt. Sharp, whereas a slope wind

521 model [Kite et al., 2013] would predict increasing diagenesis as the rover climbs lower

522 Mt. Sharp. Characterization of diagenetic mineral assemblages observed by Mt. Sharp as

523 well as their spatial pattern along the rover traverse thus collectively can reveal the

524 paleoenvironmental and sedimentary histories of Gale crater’s Mt. Sharp.

526 Acknowlegements

527 We thank Edwin Kite for sharing his Matlab code for the wind slope model for

528 generation of central mounds of discrete widths. This work was partially funded by an

529 MSL Participating Scientist grant to B.L.E. The Caltech Summer Undergraduate

530 Research Fellowship program provided programmatic support to C.B.

531 References

532 Anderson, R. C., and J. F. Bell III (2010), Geologic mapping and characterization of Gale

533 Crater and implications for its potential as a Mars Science Laboratory landing

534 site, Mars, 5, 76–128, doi:10.1555/mars.2010.0004.

535 Beardsmore, G. R., & Cull, J. P. (2001). Crustal heat flow: A guide to measurement and

536 modelling Cambridge University Press. 537 Boyce, J. M., & Garbeil, H. (2007). Geometric relationships of pristine martian complex

538 impact craters, and their implications to mars geologic history. Geophysical

539 Research Letters, 34(16)

540 Bristow, T.F. et al., The origin and implications of clay minerals from Yellowknife Bay,

541 Gale crater, Mars. American Mineralogist, in press

542 Carter, J., F. Poulet, J.-P. Bibring, N. Mangold, and S. Murchie (2013), Hydrous minerals

543 on Mars as seen by the CRISM and OMEGA imaging spectrometers: Updated global

544 view, J. Geophys. Res. Planets, 118, doi: 10.1029/2012JE004145.

545 Ehlmann, B.L. and J. Buz. Mineralogy and Fluvial History of the Watersheds of Gale,

546 Knobel, and Sharp craters: A Regional Context for MSL Curiosity’s Exploration,

547 Geophysical Research Letters, accepted

548 Ehlmann, B. L., et al. (2009), Identification of hydrated silicate minerals on Mars using

549 MRO-CRISM: Geologic context near Nili Fossae and implications for aqueous

550 alteration, J. Geophys. Res., 114, E00D08, doi:10.1029/2009JE003339.

551 Ehlmann, B. L., J. F. Mustard, S. L. Murchie, J.-P. Bibring, A. Meunier, A. A. Fraeman,

552 and Y. Langevin (2011a), Subsurface water and clay mineral formation during the

553 early history of Mars, Nature, 479, 53–60, doi:10.1038/nature10582.

554 Ehlmann, B. L., J. F.Mustard, R. N. Clark, G.A. Swayze, and S. L.Murchie (2011b),

555 Evidence for low-grade metamorphism, hydrothermal alteration, and diagenesis on

556 mars from phyllosilicate mineral assemblages, Clays Clay Miner., 590(4), 359–377,

557 doi:10.1346/CCMN.2011.0590402. 558 Fraeman, A., et al. (2013). A hematite-bearing layer in gale crater, mars: Mapping and

559 implications for past aqueous conditions. Geology, 41(10), 1103-1106.

560 Garvin, J., Sakimoto, S., & Frawley, J. (2003). Craters on mars: Global geometric

561 properties from gridded MOLA topography. Sixth International Conference on

562 Mars, , 1 3277.

563 Golombek, M., et al. (2006). Erosion rates at the mars exploration rover landing sites and

564 long‐term climate change on mars. Journal of Geophysical Research: Planets

565 (1991–2012), 111(E12)

566 Grant, J. A., et al. (2011). The science process for selecting the landing site for the 2011

567 mars science laboratory. Planetary and Space Science, 59(11), 1114-1127.

568 Grotzinger, J. P., et al. (2012). Mars science laboratory mission and science

569 investigation. Space Science Reviews, 170(1-4), 5-56.

570 Grotzinger, J. P., Sumner, D. Y., Kah, L. C., Stack, K., Gupta, S., Edgar, L. et al. (2014).

571 A habitable fluvio-lacustrine environment at Yellowknife Bay, Gale Crater,

572 mars. Science (New York, N.Y.), 343(6169), 1242777. doi:10.1126/science.1242777.

573 Harvey H. R., J.H. Tuttle, J.T. Bell (1995) Kinetics of phytoplankton decay during

574 simulated sedimentation: Changes in biochemical composition and microbial activity

575 under oxic and anoxic conditions. Geochim. Cosmochim. Acta 59(16), 3367–3377.

576 Kalynn, J., Johnson, C. L., Osinski, G. R., & Barnouin, O. (2013). Topographic

577 characterization of lunar complex craters. Geophysical Research Letters, 40(1), 38-

578 42. 579 Kite, E. S., Lewis, K. W., Lamb, M. P., Newman, C. E., & Richardson, M. I. (2013).

580 Growth and form of the mound in gale crater, mars: Slope wind enhanced erosion

581 and transport. Geology, 41(5), 543-546.

582 Le Deit, L., E. Hauber, F. Fueten, M. Pondrelli, A. P. Rossi, and R. Jaumann (2013),

583 Sequence of infilling events in Gale Crater, Mars: Results from morphology,

584 stratigraphy, and mineralogy, J. Geophys. Res. Planets, 118, 2439–2473,

585 doi:10.1002/2012JE004322.

586 Lehmann, M.F., S.M. Bernasconi, Barbieri, J.A. McKenzie (2002), Preservation of

587 organic matter and alteration of its carbon and nitrogen isotope composition during

588 simulated and in situ early sedimentary diagenesis. Geochimica et Cosmochimica

589 Acta, 66(20), 3573-3584.Lewis, K. W., & Aharonson, O. (2014). Occurrence and

590 origin of rhythmic sedimentary rocks on mars. Journal of Geophysical Research:

591 Planets, 119, 1432–1457, doi:10.1002/2013JE004404.

592 Mackenzie, F. T., ed. (2005). Sediments, Diagenesis, and Sedimentary Rocks: Treatise

593 on Geochemistry, Second Edition, Volume 7, Elsevier, Oxford, UK, 446 pp.

594 Malin, M. C., & Edgett, K. S. (2000). Sedimentary rocks of early

595 mars. Science, 290(5498), 1927-1937.

596 Milliken, R. E., and D. L. Bish (2010), Sources and sinks of clay minerals on Mars.

597 Philos. Mag., 90, 2293–2308, doi:10.1080/14786430903575132.

598 Milliken, R. E., J. P. Grotzinger, and B. J. Thomson (2010), Paleoclimate of Mars as

599 captured by the stratigraphic record in Gale Crater, Geophys. Res. Lett., 37, L04201,

600 doi:10.1029/2009GL041870. 601 Milliken, R. E., R. C. Ewing, W. W. Fischer, and J. Hurowitz (2014), Wind-blown

602 sandstones cemented by sulfate and clay minerals in Gale Crater, Mars, Geophys.

603 Res. Lett., 41, 1149–1154, doi:10.1002/2013GL059097.

604 Morschhauser, A., Grott, M., & Breuer, D. (2011). Crustal recycling, mantle dehydration,

605 and the thermal evolution of mars. Icarus, 212(2), 541-558.

606 Pelkey, S.M., B.M., Jakosky, P.R. Christensen (2004) Surficial properties in Gale Crater,

607 Mars, from Mars Odyssey THEMIS data. Icarus, 167, 244-270.

608 Rampe, Am Min

609 Robbins, S. J., & Hynek, B. M. (2012). A new global database of mars impact craters≥ 1

610 km: 1. database creation, properties, and parameters. Journal of Geophysical

611 Research: Planets (1991–2012), 117(E5)

612 Siebach, K. L., and J. P. Grotzinger (2014), Volumetric estimates of ancient water on

613 Mount Sharp based on boxwork deposits, Gale Crater, Mars, J. Geophys. Res.

614 Planets, 119, 189–198, doi:10.1002/2013JE004508.

615 Siebach, K. L., and J. P. Grotzinger (2014), Volumetric estimates of ancient water on

616 Mount Sharp based on boxwork deposits, Gale Crater, Mars, J. Geophys. Res.

617 Planets, 119, 189–198, doi:10.1002/2013JE004508.

618 Stack, K. M., et al. (2014), Diagenetic origin of nodules in the Sheepbed member,

619 Yellowknife Bay formation, Gale crater, Mars, J. Geophys. Res. Planets, 119,

620 1637–1664, doi:10.1002/2014JE004617. 621 Thomson, B., et al. (2011). Constraints on the origin and evolution of the layered mound

622 in Gale crater, Mars using Mars Reconnaissance Orbiter data. Icarus,214(2), 413-

623 432.

624 Tosca, N. J., & Knoll, A. H. (2009). Juvenile chemical sediments and the long term

625 persistence of water at the surface of mars. Earth and Planetary Science

626 Letters, 286(3), 379-386.

627 Vaniman, D. T., Bish, D. L., Ming, D. W., Bristow, T. F., Morris, R. V., Blake, D. F. et

628 al. (2014). Mineralogy of a mudstone at yellowknife bay, gale crater, mars. Science

629 (New York, N.Y.), 343(6169), 1243480. doi:10.1126/science.1243480 [doi]

630 Wray, J. J. (2012), Gale crater: The Mars Science Laboratory/Curiosity Rover Landing

631 Site, Int. J. Astrobiol., doi:10.1017/S1473550412000328.

632 Werner, S., & Tanaka, K. (2011). Redefinition of the crater-density and absolute-age

633 boundaries for the chronostratigraphic system of mars. Icarus, 215(2), 603-607.

635 Tables

641 Figure Captions 642 Figure 1. (A) Overview of 154-km diameter Gale crater (137.4oE, -4.6oN). The star

643 indicates MSL’s Bradbury landing site. Aeolis Mons (Mt. Sharp) is located on center of

644 the figure. Two scenarios for Mt. Sharp’s formation are analyzed in this paper: (B)

645 complete fill of the crater to the rim, followed by partial exhumation; (C) the Kite et al.,

646 [2013] model of wind slopes creating feedback with the sides of the crater, enabling

647 mound formation in the center with only incomplete fill.

649 Figure 2. Ancient Gale profile (solid black line) is compared with the modern (dashed

650 gray line) and average (solid gray line) Gale profiles. The ancient Gale profile represents

651 the initial state of the model for every scenario. Here we try to explain how the ancient

652 Gale profile turned out to be the modern Gale profile.

654 Figure 3. Depth to temperature relation considering two different early Mars

655 temperatures (0oC and -50oC) as well as four period of time (Noachian/Hesperian, Early

656 Amazonian, Late Amazonian and Modern).

659 Figure 4. Bradbury Landing, Yellowknife Bay (YKB), Hidden Valley and a potential

660 future MSL location (an unconformity at Mt. Sharp) are identified in the figure. The

661 change in altitude from Yellowknife Bay to the unconformity is of 1000 m. One should

662 also notice that the distance from Yellowknife Bay to the Hidden Valley, longitudinally

663 speaking, is small when compared to the distance from Yellowknife Bay to the

664 unconformity. 665

666 Figure 5. Topography variation and overburden rates for scenario 1 (Figures 5a and 5d),

667 scenario 2a (Figures 5b and 5e) and scenario 2b (Figures5c and 5f) are presented here.

669 Figure 6. Temperature as a function of timing scenarios for sedimentary models 1, 2a

670 and 2b considering an early mean surface temperature of -50oC and 0oC. Solid lines

671 represent location at Yellowknife Bay; dashed lines represent location at Mt. Sharp.

673 Figure 7. Time-temperature integral as a function of timing scenarios for sedimentary

674 models 1, 2a and 2b considering an early mean surface temperature of -50oC and 0oC.

675 Solid lines represent location at Yellowknife Bay; dashed lines represent location at Mt.

676 Sharp.

678 Figure 8. Maximum and minimum temperatures experienced by MSL’s traverse

679 considering different sedimentary scenarios and timescales for cold and warm early Mars.