Icarus 211 (2011) 1204–1214

Contents lists available at ScienceDirect

Icarus

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

Sequence and timing of conditions on early Mars

Caleb I. Fassett ⇑, James W. HeadDepartment of Geological Sciences, Brown University, 324 Brook Street, Box 1846, Providence, RI 02912, USA

a r t i c l e i n f o a b s t r a c t

Article history:Received 21 July 2010Revised 5 November 2010Accepted 9 November 2010Available online 19 November 2010

Keywords:MarsMars, SurfaceGeological processesAstrobiology

0019-1035/$ - see front matter � 2010 Elsevier Inc. Adoi:10.1016/j.icarus.2010.11.014

⇑ Corresponding author.E-mail address: Caleb_Fassett@brown.edu (C.I. Fas

The geological record of early Mars displays a variety of features that indicate fundamental differencesfrom more recent conditions. These include evidence for: (1) widespread aqueous alteration and phyllos-ilicate formation, (2) the existence of an active magnetic dynamo, (3) the erosion of extensive valley net-works, some thousands of kilometers long, (4) a much more significant role of impact cratering, formingstructures up to the scale of large basins, and (5) the construction of much of the Tharsis volcanic prov-ince. Mars also is likely to have had a much thicker atmosphere during this early period. We discuss andreview the temporal relationships among these processes and conditions. Key observations from thisanalysis suggest the following: (1) the last large impact basins, Argyre, Isidis, and Hellas, all pre-datethe end of valley network formation, potentially by several hundred million years, (2) the magneticdynamo is likely to be ancient (pre-Hellas), since the center of Hellas and other young basins lack mag-netic remanence, and (3) the period of phyllosilicate formation is not readily connected to the period ofvalley network formation. Concepts for the possible formation and evolution of life on Mars shouldaddress this time sequence of conditions.

� 2010 Elsevier Inc. All rights reserved.

1. Introduction

Spacecraft data indicate that the early environment of Mars dif-fers from recent conditions in a multitude of important ways (see,e.g., Solomon et al., 2005; Carr and Head, 2010). Before the mid-Hesperian, Mars appears to have had higher impact flux (Hartmannand Neukum, 2001), a wetter surface (e.g., Carr, 1996; Craddockand Howard, 2002), more volcanic resurfacing (Tanaka et al.,1987), neutral-pH aqueous alteration (Bibring et al., 2006; Murchieet al., 2009), an intense magnetic dynamo (Acuña et al., 1999), andpossibly a denser atmosphere (e.g., Jakosky and Philllips, 2001).More speculatively, Mars may have had an ocean on its Noachiansurface (Baker et al., 1991; Clifford and Parker, 2001; Di Achilleand Hynek, 2010); direct evidence strongly favors the existenceof many large lakes (see, e.g., Irwin et al., 2002; Fassett and Head,2008a). Each of these factors, with the possible exception of a high-er impact flux, is broadly consistent with a more habitable Mars inthe Noachian to Early Hesperian than at present. The potential hab-itability of the ancient planet has helped motivate an explorationstrategy predicated on examining geological materials from thisearly period (e.g., Grotzinger, 2009).

However, given the fact that conditions on early Mars appeardistinct from those observed today, it is common to assume thatthere is a discrete geological period perhaps of some length whenall of these conditions existed simultaneously (active magnetic

ll rights reserved.

sett).

field, valley formation, erosion and transport, aqueous alteration,etc.). Although such a scenario is possible, a variety of observationsconstraining the timing of these processes suggests that it may notbe the most probable scenario. In this paper we review the con-straints on the timing of various conditions based on stratigraphy,crater counting, inferences from the martian meteorite ALH84001,and a variety of orbital observations.

1.1. Crater statistics and the age of surfaces and materials

Given our current lack of samples acquired from known loca-tions on Mars, the primary technique for deriving ages of surfacesor geomorphic features is to measure their superposed crater size-frequency distribution (e.g., Hartmann, 1966; Soderblom et al.,1974; Neukum and Wise, 1976; McGill, 1977; Tanaka, 1986;Barlow, 1988, 1990; Strom et al., 1992; Hartmann and Neukum,2001; Neukum et al., 2010). Since craters are presumed to accumu-late in a spatially random process, at least insofar as the crater pop-ulation is dominated by primary impactors, areas with higherspatial densities of craters are interpreted to be older. If a reason-able model for the rate at which craters are accumulating can beobtained (e.g., Hartmann and Neukum, 2001), absolute ages canbe estimated. These are model dependent, and are therefore lessdefinitive than relative age determinations.

A complicating factor in using craters to derive relative or abso-lute ages is that the number of craters observed in a given region isnot independent of its geological history. Gradation, erosion, andexhumation can remove craters from the surface population or

C.I. Fassett, J.W. Head / Icarus 211 (2011) 1204–1214 1205

expose the remnants of craters from an earlier era (e.g., Grant andSchultz, 1990; Malin and Edgett, 2000; Hartmann and Neukum,2001). The fact that crater counting measurements are a combina-tion of surface age and crater retention mediated by geological pro-cesses has been emphasized by Hartmann (1966), who proposedthe useful concept of ‘‘crater retention age.’’ Crater retention agesof a given surface can vary substantially (by an order of magnitudeor more) when examining craters in different size ranges or at dif-ferent reference diameters. The crater population at larger diame-ters is more representative of unit emplacement ages than thepopulation of smaller craters (<1–2 km), which are easier toremove from the record and commonly reflect rates of regolithprocesses rather than the age of underlying bedrock (e.g.,Hartmann, 2005). If we are primarily interested in the age ofemplacement for surface units, this imposes practical limitationson the minimum area where formation age can be ascertained, sincelarge craters form considerably less frequently than small craters.However, since many of the major stratigraphic markers of impor-tant events in early Mars history subtend large areas, such as largeimpact basins, this limitation does not preclude developing a reason-ably strong relative sequence of events from crater statistics alone.

2. Basin and impact flux, and basin sequence

The impact flux at Mars and throughout the inner Solar Systemis inferred to have been relatively constant over the last 3 Gyr, per-haps within a factor of 1.5 of the modern rate, but prior to that timeis thought to have increased rapidly (e.g., Hartmann, 1972;Neukum and Wise, 1976; Guinness and Arvidson, 1977; Neukumet al., 2001). At the Noachian–Hesperian boundary the impact ratemay be a factor of �80 greater than rates at present (Neukum et al.,2001; Hartmann and Neukum, 2001). Absolute age estimates forthe Noachian/Hesperian boundary are TNH = 3.5–3.75 Gyr, depend-ing on the model age system being used (see Hartmann andNeukum, 2001; Ivanov, 2001; Hartmann, 2005; Fassett and Head,2008b). Earlier, the rate of impacts is unknown and depends onnumerous factors, including whether there was a focused periodof heavy bombardment that affected Mars at �3.9 Gyr (seeChapman et al., 2007).

The flux model of Hartmann and Neukum (2001) assumes thatthere was no impact ‘spike’ per se, but impacts at 3.9 Gyr are stillpresumed to be a factor of 3� greater than at 3.74 Gyr (and a factorof �250 times the present rate). If the early impact flux was suffi-ciently high during the first 600 Myr of Mars history (>�4 Gyr), it ispossible that the terrain may have been cratered to saturationequilibrium, a condition where on average every new crater erasesa pre-existing crater of comparable size. Recent modeling andobservational evidence suggests that this condition was achievedon the Moon (Richardson, 2009; Head et al., 2010). On Mars, terrainfrom this period where the crust was saturated with impacts maybe described as ‘‘Pre-Noachian’’ (Frey et al., 2003; Nimmo andTanaka, 2005); the pre-Noachian/Noachian boundary is assignedto the Hellas impact. Even when saturation is reached, preservationof features on the surface is likely to be scale-dependent: thesignature of large impact basins (>�500 km in size), such as theirlong-wavelength topography, persist even when other geologicalfeatures from this time are entirely obliterated. An example of sucha persistent topographic signature is the martian dichotomy bound-ary, since it pre-dates the oldest mappable surface units and demar-cates a major difference in crustal thickness (Solomon et al., 2005).Much of the rock mass making up the martian crust probably alsopre-dates the surface age of materials that are exposed, as recordedin the ancient age of the meteorite ALH84001, 4.09 ± 0.03 Gyr(Lapen et al., 2010) (note that earlier estimates imply an age forALH84001 closer to crustal formation; �4.50 ± 0.13 Gyr; Nyquistet al., 1995).

Because of this much enhanced impact flux on the early planet,large basin formation (impacts producing craters with diameter>500–750 km) appears to have occurred only during the early his-tory of Mars, from its formation until the end of the Noachian. Thisobservation may be slightly at odds with the models for crater fluxat Mars that are usually applied. Using Poisson statistics and thecratering models of Hartmann (H; Hartmann, 2005) and Neukum(N; from Ivanov (2001)), we calculate a large chance of a basin laterin Mars history: a 60% (H) and 72% (N) chance of at least one crater750 km or larger since the Noachian/Hesperian boundary and an89% (H) and 99% (N) chance for craters larger than 500 km. Thesecalculations would imply an expected number of cratersP500 km between 2 (H) and 4 (N) since the end of the Noachian.

Although small number statistics (N = 0) are unreliable, the lackof any obvious candidates for post-Noachian impact basins largerthan �500 km suggest that the formation rate of very large cratersmay be overestimated late in Mars history by these models. Thisargument assumes that if the expected basins formed during thepost-Noachian, they would stand a good chance of surviving as rec-ognizable impact structures, which is plausible since the rates ofcrater modification and erosion in the Hesperian and Amazonianon Mars appear low (e.g., Craddock and Maxwell, 1993; Golombeket al., 2006).

This apparent absence of late large basins is in agreement with,though does not prove, the hypothesis that the impactor size pop-ulation changed at �3.8 Gyr (Strom et al., 2005), perhaps related tothe end of the Late Heavy Bombardment. Strom et al. (2005) arguethat this change in impactor population is reflected by an increasedfrequency of larger impacts (greater than�10–20 km) compared tosmaller impacts (�10–20 km and smaller) prior to the end of theheavy bombardment on old terrains of the Moon, Mercury, andMars compared to younger plains. Recent observations consistentwith this change have been made on the Moon (Head et al.,2010), Mercury (Fassett et al., manuscript in preparation) andMars, where Werner (2008) noted that it appears that ‘‘[t]hebasin-forming projectile population is most likely different fromthe general impactor population’’.

Regardless of the uncertainties in the cratering flux and impac-tor populations, it is possible to use the superposed visible craterpopulations to directly assess the formation of the largest well-exposed basins on Mars in relative terms. Independent cratercounts on the best preserved rim regions of Argyre, Isidis, Hellasand other basins by Schultz and Rogers (1984), Werner (2008),and us, all suggest that the sequence of the largest, well-preservedimpact basins was Hellas, Isidis, then Argyre (see Table 1 andFig. 1); note that Tanaka (1986) would have Isidis before Hellas.Based on our data, Hellas is at the base of the Noachian, and Isidisand Argyre are Early-to-Mid Noachian.

Along with their sequence, basins represent important strati-graphic markers on the surface which can be used to directly inferenvironmental changes. One factor that is important is that all ofthese basins have been incised by valley networks on their interioror immediate exterior, implying that substantial fluvial activitytook place after their formation (Fig 1). This is consistent with cra-ter counting evidence that valley networks continued to be activeuntil at least the end of the Late Noachian or possibly into the EarlyHesperian (Fassett and Head, 2008b). However, Argyre appears tohave the best preserved basin-related facies (e.g., Schultz, 1986;see also Fig. 1); thus, the inferred sequence of basins from craterstatistics is also supported by the preservation state of the basins.

Several recent studies have extended the search for the visiblecrater population to use topography and gravity data to map qua-si-circular depressions (QCDs), circular crustal thickness anomalies(CTAs), and ghost craters, which are interpreted as buried or highlydegraded impact structures superposed on many martian surfaces(Frey et al., 2002; Head et al., 2002; Frey, 2006, 2008; Edgar and

Table 1Crater measurements of the Hellas, Isidis, and Argyre. N(X) is the cumulative number of craters PX, normalized to an area of 106 km2; errors in N(X)are from ±r =

pN/A. The basins are clearly distinguishable from each other on the basis of the relative density differences for craters larger than

20 km. Age estimates are model ages from Hartmann (2005) isochrons (AH) and Neukum isochrons (AN) (reproduced in Ivanov (2001)). Thecomputed ‘model ages’ are very insensitive to changes in crater frequency because of the high flux of impacts assumed in these absolute age modelsearly in Mars history (as can be seen by comparing the frequencies and derived ages for Hellas and Argyre). Statistical fit errors for given model agesare �±0.01 Gyr. In reality, the age is not known nearly this well; uncertainty in age estimates is dominated by the systematic uncertainty in theabsolute age calibration/impactor flux (see, e.g., Hartmann and Neukum, 2001; Werner, 2008). Count areas are shown in Fig. 1 along with the cratersize-frequency distributions for Hartmann and Neukum isochrons.

Basin Count area (km2) N(20) N(64) AH AN Period

Hellas 7.3 � 105 151 ± 14 27 ± 6 4.02 4.04 Base of the NoachianIsidis 4.7 � 105 117 ± 16 11 ± 5 3.96 3.97 Mid-to-Early NoachianArgyre 1.6 � 106 88 ± 7 10 ± 2.5 3.92 3.94 Mid (to Early?) Noachian

1206 C.I. Fassett, J.W. Head / Icarus 211 (2011) 1204–1214

Frey, 2008). The recognition of quasi-circular depressions andghost craters has been particularly important for inferring that be-neath the upper few hundred meters of the northern plains, thereis an older (Hesperian?) ridged plains unit that is not deeply buried(Withers and Neumann, 2001; Head et al., 2002) and that belowthis surface, the basement of the northern plains are as old asthe Noachian-aged highlands (Frey et al., 2002).

More recently, Frey (2008) and Lillis et al. (2008a) suggested thatthe combined population of possible QCDs and CTAs may providethe best data for considering the age of the largest basins (using fea-tures of 300 km and more in diameter). However, we are cautiousabout relying on this approach, since their data would imply thatArgyre is older than Isidis by an appreciable margin, and that Isidisis as young as the beginning of the Late Noachian (�3.8 Gyr).

This sequence of basins (Hellas, Argyre, Isidis) is in direct dis-agreement with the measured population of visible craters (Table 1and Fig. 1). Across a wide range of visible crater diameters (Table 1and Fig. 1), there are fewer superposed craters on Argyre thanIsidis, and fewer craters on Isidis than Hellas. Moreover, the rimregion of Isidis (the Libya Montes) has a crater population datingto the Mid-to-Early Noachian boundary, which is inconsistent witha Late Noachian formation for the basin. Finally, the inferredsequence of these youngest fresh impacts is inconsistent with therelative preservation state of the basins (Schultz, 1986). We suggestthat possible reasons for this discrepancy include one or more of thefollowing factors: (1) small number statistics: since these young andwell-preserved basins never had many >300 km craters to beginwith (visible, degraded, QCDs or CTAs), inferring their relative agefrom craters of this size may lead to errors, (2) some QCDs or CTAsmay not be impact structures, and/or (3) there may be different de-grees of basin floor filling that affects the number of QCDs/CTAs thatcan be recognized (Isidis is potentially more filled than Argyre)(Head et al., 2002; Howenstine and Kiefer, 2005).

In summary, for the youngest, well-exposed basins on Mars, weprefer to rely on the superposed visible crater population forassessment of their timing and relative sequence – first Hellas,then Isidis, then Argyre.

3. Valley networks and surface erosion

Valley networks provide morphological evidence for fluvialactivity, erosion, sedimentary transport, and a hydrological cycleon early Mars (Carr, 1996). Valley networks have numerous tribu-taries (Hynek et al., 2010), often begin near drainage divides(Craddock and Howard, 2002), and were interconnected acrossgreat distances, at least during their period of peak activity (see,e.g., Irwin et al., 2005; Fassett and Head, 2008a). Paleo-lakes onMars appear to have been relatively common features (e.g., Fassettand Head, 2008a, and references therein), and certain valleys suchas Ma’adim Valles, which initially appeared to come from localizedsources (e.g., Gulick, 2001), appear to have formed as thesepaleo-lakes overtopped confining topography (Irwin et al., 2002).

Groundwater-driven valley erosion alone seems inconsistent withmany valley characteristics, particularly the dendritic, high-ordertributaries that extend to drainage divides (Hynek et al., 2010). Evenif some valleys formed as the result of groundwater discharge, pre-cipitation-based recharge seems to have been necessary to close thehydrological cycle, as basic calculations suggest that subsurfacewater reservoirs would need to be recharged many times to erodethe valley networks observed (Goldspiel and Squyres, 1991; Gulick,2001). The characteristics of valley networks thus seem to require,at minimum, time periods when precipitation on the surface waspossible, water was cycled through the early Mars atmosphere,and water was stable or metastable at the martian surface(Craddock and Howard, 2002; Hynek et al., 2010).

Several independent studies have attempted to estimate whenthe most extensive period of valley network formation occurred(e.g., Pieri, 1980; Carr and Clow, 1981; Fassett and Head, 2008b;Hoke and Hynek, 2009), using stratigraphic and crater countinganalysis to date the termination of valley network activity. Thesestudies suggest that regional-to-global-scale valley formation per-sisted until approximately the Noachian/Hesperian boundary orinto the Early Hesperian at the latest (Fassett and Head, 2008b;Hoke and Hynek, 2009). Note that this ‘regional-to-global’ scaleformation excludes certain regions that are thought to be localexceptions, such as valleys on certain volcanoes (e.g., Gulick andBaker, 1990; Fassett and Head, 2006, 2007), in association with gla-ciation (Dickson et al., 2009; Fassett et al., 2010), and within, or inthe vicinity of, young, large craters (e.g., Williams and Malin, 2008;Tornabene et al., 2008; Morgan and Head, 2009).

In our study (Fassett and Head, 2008b), we suggested that twopossible interpretations were consistent with our craters statistics:either (1) global termination of valley activity near the Noachian/Hesperian boundary or (2) persistence of some valleys into theEarly Hesperian.

Increasing evidence has been put forth for erosion in at leastsome major valley networks were active well into the EarlyHesperian or possibly the Late Hesperian (e.g., Mangold and Ansan,2006; Ansan and Mangold, 2006; Bouley et al., 2009, 2010). Insome of this work, a younger period of activity is derived than inFassett and Head (2008b), primarily due to differences in analyticalchoices, particularly: (1) how count regions are aggregated, (2) dif-ferent stratigraphic interpretations and, most importantly, (3) theeffective diameter used to compare observed crater populationswith isochrons. At some level, these factors are coupled, since lar-ger diameter craters require greater aggregation of area to achievemeaningful statistics, at the expense of the ability to discern reallocal variation if it exists (as noted by Bouley et al. (2010)). As de-scribed above, reliance on smaller craters may result in youngerages due to crater retention. In summary, age data continue to sup-port the idea that regional to global-scale valley network formationterminated in the Early Hesperian, although new evidence hasbolstered the interpretation that valley formation lasted into thisperiod (Bouley et al., 2009, 2010).

Fig. 1. (Top two rows) Crater size-frequency measurements in incremental (top row) and cumulative (second row) plots on terrains related to Argyre, Isidis, and Hellas (thirdrow), which result in Early-to-Mid-Noachian ages. These basins have �3–5� the crater density superposed on valley networks, which in aggregate have a frequency near theNoachian/Hesperian boundary, valley network formation mostly terminated in the Early Hesperian. The bottom row shows examples of valleys superposed on each of thesemajor, young impact basins.

C.I. Fassett, J.W. Head / Icarus 211 (2011) 1204–1214 1207

1208 C.I. Fassett, J.W. Head / Icarus 211 (2011) 1204–1214

Given the observations on the timing of valley network forma-tion, three key questions remain to be definitively addressed: (1)how active was the period of valley network formation in the Noa-chian to the Early Hesperian, (2) how common/continuous werethe periods when valleys were forming on early Mars, and (3)was water required to be stable over an extended period of time?

Estimates from modeling of valley network-associated sedi-mentary deposits imply emplacement times that are geologicallyquite short, of order 1–1000 years (Jerolmack et al., 2004;Kleinhans, 2005; Lewis and Aharonson, 2006; Kraal et al., 2008;Kleinhans et al., 2010). However, these estimates are based on con-tinuous activity and sediment transport; estimates that assumeterrestrial-like intermittency or sediment supply unsurprisinglyinfer much longer periods of time (Moore et al., 2003; Fassettand Head, 2005).

Drainage basin characteristics provide some of the strongestarguments for valley network formation over an extended periodof time (�>105 years) (Barnhart et al., 2009). Barnhart et al.(2009) synthetically reconstructed pre-erosion topography of theParana drainage basin, and applied a variety of erosion scenariosto examine their consistency with the topography we actuallyobserve. They found that an intense period of fluvial erosion andprecipitation lasting �103–104 Earth years would be sufficient toerode the valleys that are observed. However, these intense erosionscenarios resulted in a pattern of erosion far more integrated (withmore crater rims breached) than what we observe on the surface.Thus, models that have greater episodicity, with runoff distributedover �105–106 Earth years, are interpreted to be more consistentwith the drainage pattern observed in the Parana region (Barnhartet al., 2009).

4. Volcanism

Volcanism is known to be a major factor in the long-term cli-mate evolution of Mars, as eruptions liberate volatile species fromthe planetary interior to the atmosphere (e.g., Jakosky and Philllips,2001; Phillips et al., 2001; Craddock and Greeley, 2009). For thisreason, volcanism has commonly been inferred to be closely linkedto changes in the surface environment. The formation of the massof Tharsis in particular has been implicated in a transition from aphyllosilicate-forming era (phyllosian) to a sulfate-forming era(theiikian) (Bibring et al., 2006; Bibring and Langevin, 2008).

Constraining the timing of Tharsis volcanism is critical tounderstanding whether this conclusion is reasonable. On the basisof the fact that its emplacement and load on the Mars lithosphereinfluenced the orientation of Late Noachian/Early Hesperian valleynetworks, Phillips et al. (2001) argued that the bulk of the Tharsisvolcanic was emplaced during the Noachian. Further evidence thatTharsis construction is ancient also comes from mapping, cratercounts, and analysis of the tectonic record (Plescia and Saunders,1982; Anderson et al., 2001), as well as from observations that por-tions of Tharsis are magnetized, even at high elevations (�7 km)(Johnson and Phillips, 2005).

On the other hand, the interpretation that the bulk of Tharsis isNoachian has been disputed by Craddock and Greeley (2009), whopoint out that the lack of craters on much of Tharsis means thatmost of its surface is Hesperian or Amazonian, and requires signif-icant post-Noachian resurfacing. Craddock and Greeley (2009) esti-mate that lava deposits up to �10 km in thickness are required toremove a sufficient number of craters to reset the terrain age.

It is plausible that these two views can be reconciled in a sce-nario where the majority of the crust at Tharsis is constructed inthe Noachian (crustal thickness �50–100 km; Neumann et al.,2004), but where extensive volcanic resurfacing persists throughHesperian and Amazonian times (see also Solomon and Head,1982). However, the observation that a substantial amount of

Tharsis-building is ancient (e.g., back to the Mid-Noachian orbefore) remains credible, as the existence of ancient, Noachianregions is clear, particularly in the Thaumasia highlands (Plesciaand Saunders, 1982). Given that the magnetization of parts of Thar-sis (Johnson and Phillips, 2005), early volcanism in these regionsmay pre-date Hellas (see Section 6). The interpretation that theconstruction of Tharsis near the end of the Noachian led to secularchanges which caused Mars to transition from a planet wherephyllosilicate formation was common to one dominated by sulfateformation (Bibring et al., 2006; Bibring and Langevin, 2008) maynot be consistent with the fact the bulk of Tharsis may be old.

Hesperian and younger volcanism on Mars is also importantregardless of the timing of Tharsis. In particular, volcanic plainsemplacement, particularly focused in the northern lowlands, resur-faced �30% of the surface of Mars in this period (Head et al., 2002).Estimates from Viking mapping suggests that more than half of thevolcanic resurfacing on Mars is Early Hesperian or younger (Tanakaet al., 1987; Greeley and Schneid, 1991); higher resolution observa-tions with recent data would imply that this is conservative, be-cause small patches of volcanic plains have been increasinglyrecognized in the highlands (Fassett and Head, 2008a).

In summary, the volcanic history of Mars should be closely cor-related with a number of other conditions on the planet, includingthe density of the atmosphere, atmospheric chemistry and volatileinventory. As far as it can be determined however, the timing ofvolcanism (e.g., Tanaka et al., 1987) does not imply a one-to-onelink between volcanism and surface conditions. No evidence existsthat a declining volcanic fluxes correlates well with atmosphericloss, or that periods of Noachian volcanism helped facilitate tran-sient clement conditions. Instead, the Hesperian volcanic depositsthat resurfaced 30% of Mars are volumetrically significant andstrikingly uneroded. Based on our current understanding of thetiming of volcanic deposits, secular changes in volcanism or majorvolcanic events can not be directly connected to transitions in sur-face conditions.

5. Aqueous alteration

The recognition of alteration products on Mars has been revolu-tionized by observations in the last decade across the electromag-netic spectrum, first in the thermal infrared by TES and THEMIS(e.g., Christensen et al., 2001; Wyatt and McSween, 2002; Osterlooet al., 2008), and more recently, in the visible to near-infrared, byOMEGA (Gendrin et al., 2005; Poulet et al., 2005; Bibring et al.,2006; Bibring and Langevin, 2008) and CRISM (Milliken et al.,2008; Mustard et al., 2008; Ehlmann et al., 2009; Wray et al.,2009).

These data have resulted in the recognition of at least ten dis-tinct environments where aqueous alteration products are ob-served (Murchie et al., 2009). Based on observations of hydratedminerals, particularly Fe–Mg phyllosilicates, it appears that neu-tral-pH alteration on Mars was an important process in the Noa-chian (Bibring et al., 2006). Murchie et al. (2009) examined thestratigraphic constraints on these deposits; we independently havereexamined these environments from a crater counting and strati-graphic perspective (Table 2). The time-stratigraphy of mineral for-mation in many of these environments is complicated. One of themajor issues is that in some of the outcrops where phyllosilicatesare observed, they are likely to be detrital (e.g., Ehlmann et al.,2008a; Murchie et al., 2009; Milliken and Bish, 2010). The timingof the aqueous alteration that resulted in the formation of theseclays is thus not preserved – their present state could reflect EarlyNoachian formation and Late Noachian physical weathering, trans-port, and deposition.

Where minerals remain in situ (authigenic alteration), it iseasier to make inferences about the timing of the geochemical

Table 2Time constraints on major aqueous alteration environments on Mars. The type outcrops that show aqueous alteration on Mars are modified after Murchie et al. (2009) (hisTable 3) and re-ordered in approximate chronological order based on our independent evaluation of the stratigraphic constraints on these outcrops from crater counts and localrelationships. Most neutral-pH alteration is conceivably quite old (before the Late Noachian), which may pre-date valley network activity, or at least the termination of VNformation. Evaporite deposits/chemical precipitates may be more common late in Mars history (Bibring et al., 2006; Murchie et al., 2009).

Aqueous environments Type area (s) Timing of aqueous alteration for the type area (s)

Layered phyllosilicates Nili Fossae Pre-Noachian or Early Noachian. Most of the deep crustal alteration/phyllosilicate is interpreted to be pre-Isidis (Mustard et al., 2007, 2009;Mangold et al., 2007). Exhumation is important

Mawrth Valles Pre- or Early-to-Mid Noachian. Age constraint on phyllosilicate bedrockis from craters in the Mawrth Valles region that appear to post-datethe phyllosilicate-bearing material (Michalski and Noe Dobrea, 2007).Exhumation is important (craters turned into knobs; phyllosilicatesare exposed from underneath an eroded caprock)

Deep phyllosilicates Exposures by cratersin the highlands

Pre-Noachian or Noachian, Difficult to Constrain More Specifically.Multitude of exposures in central peaks, rims, and walls are excavatedcrust, so limits on timing are hard to come by. Formation/alterationwas conceivably at depth (e.g., Parmentier et al., 2008)

Carbonate-bearing outcrops Nili Fossae Early-to-Mid Noachian. Outcrops are associated with olivine units(Ehlmann et al., 2008b, 2010) that are interpreted to be directly relatedto the Isidis basin-forming event (e.g., Mustard et al., 2009)

Serpentine-bearing outcrops

Intracrater clay-sulfates Columbus Crater Mid-Noachian? (Late-to-Early). Rim of Columbus crater hasN(5) � 440 ± 179 (N = 6), implying a (uncertain) Mid-Noachian age forthe interbedded clay and sulfates described by Wray et al. (2009)

Phyllosilicates in intracrater fans Jezero Crater, Holden Crater,Eberswalde Crater

Unconstrained. Presumably detrital. In the case of Jezero crater (e.g.,Ehlmann et al., 2008a), the source of sediments includes Early (or Pre?)Noachian phyllosilicates and Mid-to-Early Noachian carbonates in thewatershed

Plains sediments (chlorides) Terra Sirenum Late Noachian/Early Hesperian. Crater counting of the type areasuggests has a LN/EH-boundary age for the THEMIS ‘glowing’ terrain(Osterloo et al., 2008). Chlorides are presumably evaporitic in origin;associated phyllosilicates may be detrital

Meridiani-type layered deposits Meridiani Planum Late Noachian to Hesperian. These sulfate-rich deposits retain cratersrather poorly. In Meridiani, sulfate plains clearly embay highlands andhave an Early Hesperian crater density, which is thus a minimum agefor the observed water–rock interaction. ILDs are likely Hesperian inage based on stratigraphy and crater counting (Quantin et al., 2010)

Valles-type layered deposits Valles Marineris ILDs

Siliceous layered deposits Plains above VM Hesperian to Amazonian. Deposits are superposed on Late Hesperian toEarly Amazonian surfaces

Polar gypsum deposits Basal unit and surrounding dunes Unconstrained. Sand in dunes and basal unit; period of alteration isunbounded

C.I. Fassett, J.W. Head / Icarus 211 (2011) 1204–1214 1209

environments where these formed. From Murchie et al.’s (2009)classification of distinct aqueous environments, the most likelyexamples of outcrops with in situ aqueous mineral formation are(1) deep phyllosilicates (common highlands exposures usually incrater rims, central peaks, or ejecta; Mustard et al., 2008); (2) lay-ered phyllosilicates (such as Mawrth Vallis; e.g., Poulet et al.,2005); (3) certain carbonate-bearing outcrops (and, more recentlydiscovered, serpentine-bearing outcrops; Ehlmann et al., 2010) sit-uated with their ultramafic precursors (Ehlmann et al., 2008b),and, (4) environments with chemical precipitates or evaporites(chloride-bearing plains sediments, Osterloo et al., 2008; hydratedsilica deposits, e.g., Milliken et al., 2008; layered sulfates such asthose found in Meridiani Planum and Valles Marineris; e.g., Gen-drin et al., 2005; sulfates interbedded with phyllosilicates on a cra-ter interior; e.g., Wray et al., 2009). For this final class in particular,chemical sedimentation may be a result of groundwater-driveninteractions with the upper crust, rather than surface precipitation,runoff, and weathering; preservation of jarosite at these locationsalso suggests that long-term arid conditions existed after theemplacement of these chemical sediments (Elwood Maddenet al., 2004, 2009).

Despite the fact that both valley networks and phyllosilicateclays are predominantly in Noachian terrains, evidence that dem-onstrates that valley networks and these alteration products arecharacteristics of the same environment and formed at the same

time is limited. Water–rock interactions that formed clays mayhave mostly ended by the time of the Isidis impact in Nili Fossae(Mustard et al., 2007; Mangold et al., 2007), and much of the ob-served neutral-pH alteration may have occurred in very ancienttimes (Poulet et al., 2005; see also Table 2). If this is the case, thephyllosilicates may be older than the Late Noachian to Early Hes-perian valley systems where clay-bearing sediments were trans-ported and deposited, such as in Eberswalde, Holden, and Jezerocraters (e.g., Ehlmann et al., 2008a; Milliken and Bish, 2010).

Along with broad global trends, there are also differences in thecharacter of aqueous alteration around the youngest large impactbasins Isidis (Mustard et al., 2007, 2009; Mangold et al., 2007;Ehlmann et al., 2009) and Argyre (Buczkowski et al., 2010).Buczkowski et al. (2010) observe that although iron/magnesium-bearing phyllosilicates are exposed within and by the Argyre basinstructure, less mineralogical diversity is present than in a compara-ble setting at Isidis. Buczkowski et al. (2010) interpret the alterationminerals of Argyre as primarily pre-dating the basin-forming event,which acted to expose pre-existing alteration products in theNoachian crust. The greater diversity of alteration products in theNili Fossae area associated with Isidis requires multiple alterationevents in distinct weathering environments (Ehlmann et al., 2009).This distinction is consistent with Argyre being younger than Isidis(Section 2) and with a hypothesized global decline in neutral-pH,high-water–rock ratio aqueous alteration as a function of time.

1210 C.I. Fassett, J.W. Head / Icarus 211 (2011) 1204–1214

In summary, a transition in the character of aqueous alterationfrom widespread neutral-pH aqueous alteration to more localizedacidic aqueous alteration is suggested by observations of hydratedminerals on Mars (Bibring et al., 2006; Bibring and Langevin, 2008;Murchie et al., 2009). This paradigm of high-water–rock ratio alter-ation followed by more water-limited alteration later in Mars his-tory (Bibring et al., 2006; Hurowitz and McLennan, 2007) seemsborne out by new data even as many new alteration minerals havebeen recognized on the surface of Mars.

6. Magnetic anomalies and cessation of the magnetic field

Observations from the Mars Global Surveyor magnetometerexperiment demonstrated that there are crustal magnetic anoma-lies observed over much of the surface, with the strongest anoma-lies concentrated in the southern highlands (Acuña et al., 1999).These crustal anomalies imply the existence of a core dynamo onMars early in its history. The existence of this magnetic field mayhave played an important role in arresting the loss of the earlyMars atmosphere by solar wind sputtering (e.g., Jakosky and Phill-lips, 2001), as well as shielding the surface from energetic cosmicrays (e.g. Molina-Cuberos et al., 2001).

There are two compelling constraints on the timing of the Marsmagnetic field. First, crustal magnetic anomalies are largely absentin the interiors of Hellas, Argyre, Isidis, and Utopia, as well acrossmost of Tharsis and most volcanic edifices on Mars, with the excep-tions of Hadriaca Patera (Lillis et al., 2006) and Apollonaris Patera(Hood et al., 2010). The simplest explanation for the lack of magne-tization in these basins and volcanoes is that they post-date thecessation of the magnetic field. If this interpretation is correct,the core dynamo must have ended in the pre-Noachian beforethe formation of Hellas (Lillis et al., 2008a,b; see Schubert et al.(2000) and Hood et al. (2010) for alternative interpretations ofthe timing of the magnetic field). It has been suggested that theformation of large earlier basins that are now buried, such as Uto-pia, may have contributed to this termination (Roberts et al., 2009).

Second, a further possible constraint on the timing of the mag-netic field comes from ALH84001, which has remanent magnetiza-tion consistent with acquisition in a magnetic field caused by acore dynamo with strength 0.1–10� the present Earth dynamo(Kirschvink et al., 1997; Weiss et al., 2002). This interpretation ispreliminary, however, as it is not entirely clear whether themagnetization in this meteorite was acquired from a dynamo orfrom pre-existing crustal fields (Gattacceca and Rochette, 2004).If it was from a core dynamo, and the age of ALH84001 is4.091 ± 0.03 Gyr as recently suggested (Lapen et al., 2010), thiswould provide direct evidence of the persistence of a magneticdynamo until �4.09 Gyr.

If these suppositions are correct, and ALH84001 preserves acore field and Hellas formed after the dynamo ended on Mars, thisalso bounds the formation of Hellas to after 4.09 Gyr (consistentwith crater counting model ages, Table 1). Regardless of the evi-dence from ALH84001, the lack of magnetization within Hellasstrongly suggests termination of the magnetic field before thebasin formed, well before the end of valley network formation.The termination of the magnetic field before the valley networkactivity in the Late Noachian/Early Hesperian is consistent with:(1) crater counting results, which are imprecise but suggest apotentially long gap between Hellas and the end of valley forma-tion and (2) stratigraphy, which irrefutably demonstrates that val-ley formation continued after Hellas, but provides no informationabout the length of time between Hellas and the end of valley for-mation. Thus, if a magnetic dynamo was playing an importantshielding role for the surface and/or atmosphere, the shield mayhave been removed well before water stopped playing an impor-

tant geomorphic role on the martian surface (in contrast to thetimeline in Jakosky and Philllips (2001)).

One observation that complicates this scenario is the apparentcomplex magnetization that is observed in other Mars meteorites(e.g., Collinson, 1986; Collinson et al., 1997). Because the shergot-tites (�180 Ma) and nakhlites (�1.3 Ga) are much younger thanHellas (e.g., McSween, 1994), this requires that when magnetiza-tion is observed in these younger meteorite samples, it was notacquired by cooling in the presence of a dynamo. Other processesthat are plausible include shock magnetization (Cisowski andFuller, 1978), acquisition from the Mars crustal field, or by contam-ination by terrestrial fields. The alternative is that the interpreta-tion that Hellas and other non-magnetized basins formed in theabsence of a core dynamo is wrong. Given our understanding ofthe spatial distribution of magnetic remanence on Mars, thepost-dynamo acquisition of magnetization in these samples isthe simplest explanation, consistent with a scenario where the‘‘SNCs [were] more likely magnetized during or after impact thanduring the initial magmatic cooling’’ (Rochette et al., 2005). Recentmeasurements of the nakhilite Yamato 000593 support this inter-pretation, consistent with the absence of a global magnetic field onMars when Yamato 000593 formed, �1.8 Gyr (Funaki et al., 2009).

Complicating the interpretation of the magnetic record furtheris the fact that the observed pattern of crustal magnetization isheterogeneous, with virtually all of the strong remanent crustalmagnetism observed in the southern hemisphere and with onlyweak magnetic signatures north of the dichotomy boundary.One explanation for this heterogeneity is that hydrothermal alter-ation may have been critical in establishing where magnetizationin the crust is observed today (Solomon et al., 2005). If hydrother-mal alteration of the crust was preferentially concentrated inlow-lying regions, such as the largest impact basins and northernlowlands, the lack of magnetic signatures in the large, youngimpact basins may be a result of this demagnization process, evenif the active dynamo persisted after their formation (Solomonet al., 2005).

Alternatively, the hemispheric difference in observed crustalremanence may reflect a single-hemisphere dynamo (Stanleyet al., 2008), perhaps resulting from degree-one convection (e.g.,Zhong and Zuber, 2001). A hemispheric dynamo does not affectthe overall constraints on timing, since Hellas and Argyre aresurrounded by crust with strong remanent magnetization, so thesingle hemispheric dynamo should still have affected these basins.Thus, in the absence of other modifying influences, the lack of mag-netization in these basins would still imply that they post-date thecessation of the magnetic field, even if the remanent magnetizationwas a result of a one-hemisphere dynamo.

Lower-altitude measurements of the Mars crustal magneticfield would be very useful to help test which scenario is the bestexplanation for the observed magnetic anomalies (Langlais andAmit, 2008).

7. Atmosphere and possible atmospheric loss

Direct constraints on both the density of the early Mars atmo-sphere and its loss are somewhat limited. Some invocations ofhigher atmospheric pressure early in Mars history have been basedsimply on the need to explain valley network formation (e.g.,Pollack et al., 1987). Many such modeling efforts assume that sur-face conditions when valley networks were formed must havebeen above 273 K (averaged over a Mars year), and investigatorshave built various models with different atmospheric pressuresand constituents to explore how such a requirement might bemet (see, e.g., Haberle, 1998 and references therein).

Isotopic measurements provide the strongest indication thatthe early atmosphere was substantially denser than today, perhaps

C.I. Fassett, J.W. Head / Icarus 211 (2011) 1204–1214 1211

by a factor of 10 or more, and subsequently removed (summarizedin Jakosky and Philllips (2001)). These isotopic measurements con-strain the early atmosphere by comparing the abundance of lighterisotopic species, which are more efficiently stripped away by sput-tering and hydrodynamic escape, to heavier species. The factor of>10� higher density atmosphere early in Mars history that theseobservations require is a minimum estimate, since impact erosionof the atmosphere (e.g., Melosh and Vickery, 1989) does not frac-tionate lighter and heavier isotopes and may also have been animportant factor in atmospheric loss. As described above, the endof valley network formation clearly post-dates the period whenthe largest basins were formed and the period of highest impactflux; thus, the majority of valleys on the surface post-date the per-iod when this mechanism would have been most effective.

Although the change in climate associated with the loss of a sig-nificant atmosphere may be correlated with the end of the periodof valley network formation on Mars (e.g., Pollack et al., 1987;Jakosky and Philllips, 2001), this should be seen as an assumptionrather than an observation. On the basis of atmospheric argonobservations, Craddock and Greeley (2009) suggest that the mar-tian atmosphere may have been similar to the present atmosphereas far back as the Mid-Noachian. Moreover, atmospheric escaperates due to solar wind interactions have been measured to be cur-rently quite low, which suggests that even over billions of yearsonly a few millibars of CO2 and minimal water would be removedfrom Mars atmosphere (Barabash et al., 2007). This very slow lossis in agreement with theoretical expectations from atmosphericphotochemistry (Hodges, 2002).

On the other hand, models incorporating both impact erosionand sputtering suggest the loss of 95–99% of the atmosphere(Brain and Jakosky, 1998) since the beginning of the geological

Fig. 2. A schematic of the sequence of various planetary conditions on Early Mars basedexistence of a pre-Noachian period defined as the time before the Hellas impact, from whiTanaka, 2005). Along with the conditions we show, other important environmental condLate Noachian; before that time it may have had a peak (during the Late Heavy Bombarddensity, for which the time-history is poorly understood, though evidence suggests thatfor the timing and history of the core dynamo are particularly complex (see the text for mmagnetization of ALH 84001 was frozen in an active dynamo; the ‘‘Early Scenario’’ wodiscards the idea that the large basins post-date the magnetic field and requires a differenwhy they lack apparent crustal magnetization. Evidence suggests that the beginning ofPhillips, 2005) and that the bulk of the Tharsis load was in place by the period of valley

record (Early Noachian). Reconciling scenarios for atmosphericstate, climate, and surface erosion remains an important goalfor further research. Observations of the Mars atmosphere byMars Science Laboratory (e.g., Mahaffy et al., 2009) and later bythe Mars Atmosphere and Volatile Evolution Mission (Jakosky,2008) should help address these questions during the next decadeby providing improved measurements of atmospheric isotoperatios, trace gases, and interactions of the atmosphere with thesolar wind.

8. Synthesis

The observations of individual processes outlined above allowus to draw some inferences on the most likely sequence for variousconditions (Fig. 2). We summarize the relationships between theseconditions here:

1. It is likely that the magnetic field responsible for the crustalmagnetism observed pervasively in the Noachian highlands(Acuña et al., 1999) was: (a) still active at 4.09 ± 0.03 Gyr(because of magnetization in ALH84001; Kirschvink et al.,1997; Weiss et al., 2002; age from Lapen et al. (2010)) and (b)terminated by the time of the formation of Hellas (Lillis et al.,2008a,b). On the basis of our discussion of the magnetic field,this is what we call the baseline scenario in Fig. 2.

Two other scenarios for the dynamo history are reasonable: ifthe magnetization in ALH84001 post-dates a core dynamo (per-haps because it was acquired from pre-existing local crustalremanent fields), or if ALH84001 is much older than recentlydetermined, an ‘‘early scenario’’ is possible. Or, if the lack of amagnetic signature in Utopia, Hellas, Isidis and Argyre is not a

on the information described in the text. Note that in this diagram we accept thech no known surface units date on the modern surface (Frey et al., 2003; Nimmo anditions were (1) the general impact flux, which is thought to have declined since thement) or simply a monotonic rise (see discussion in text), and (2) the atmospheric

the atmosphere was denser during early periods than it is at present. The scenariosore discussion and references). The baseline scenario shown here assumes that the

uld require its magnetization from pre-existing crustal fields. The ‘‘Late Scenario’’t explanation (e.g., thin crust, lack of magnetic carriers, hydrothermal alteration) forthe construction of Tharsis pre-dates the termination of the dynamo (Johnson andnetwork formation (Phillips et al., 2001).

1212 C.I. Fassett, J.W. Head / Icarus 211 (2011) 1204–1214

result of their formation after the core dynamo terminated (e.g.,Schubert et al., 2000; Solomon et al., 2005; Hood et al., 2010), a‘‘late scenario’’ is possible. In this case, the timing of the coredynamo is not bounded except by the lack of a global magneticfield from a dynamo today.

2. Hellas is plausibly younger than 4.09 ± 0.03 Gyr on the basis ofits non-magnetization and ALH84001’s magnetic signature. Onits face, this is consistent with crater model ages of 4.02–4.04 Gyr (though the systematic calibration is far more uncer-tain than this range suggests). This upper bound on the absoluteage of Hellas goes away if the core dynamo terminated beforeALH84001 (the ‘early scenario’) or if the new ALH84001 age(Lapen et al., 2010) is too young and it is actually older (as orig-inally thought). Unless one of the hypothesized reasons requir-ing a ‘late scenario’ is correct, the relative age sequence wherethe magnetic dynamo terminated before Hellas remains.

3. Large basins like Hellas, Isidis, and Argyre pre-date the end ofvalley network formation, and hence the magnetic field likelydoes so as well. The gap between the termination of the mag-netic field and the formation of the Late Noachian to Early Hes-perian valleys could be appreciable, depending on the absolutelength of the Noachian; current impact models would suggest aperiod of 0.3–0.5 Gyr between the Hellas impact and end ofwidespread valley formation. Neither the loss of the magneticfield, nor a decline in the rate of volcanism or the impact rate,connects in a one-to-one manner with the decline in valley net-work formation.

4. Similarly, formation of much of the phyllosilicate record thatindicates that pervasive aqueous alteration on Mars is difficultto connect temporally to the period of valley network forma-tion; many of the alteration products that are observed arelikely to be older than at least the last period of widespread val-ley formation.

5. Portions of Tharsis are magnetized (Johnson and Phillips, 2005),suggesting that Tharsis construction began in the Early-to-MidNoachian or before. This is consistent with ancient tectonicactivity in parts of Tharsis (e.g., Plescia and Saunders, 1982)and with the observation that the bulk of the Tharsis load wasin place before valley network formation (Phillips et al., 2001).A secular change in the Mars environment linked to Tharsis for-mation cannot be connected in a one-to-one manner withobservations of the shift in the nature of aqueous alterationenvironment.

A few other implications from these timing constraints areapparent. As has been discussed before (Fassett and Head, 2008b;Hynek et al., 2010), the obvious large basins (>�500–600 km) onMars appear too old to be the direct cause of valley formation, incontrast to the original scenario described by Segura et al.(2002), where >100 km impactors lead to surface warming and val-ley formation. If the impact hypothesis described by Segura et al.(2002) is to work, smaller impactors are more likely to be the causeof valley networks (see also Toon et al., 2010). Timing constraintsalone allow for this possibility, although whether it is possible toreconcile the observed erosion with the erosion that impacts mightproduce still seems uncertain.

Second, these results suggest that if the magnetic field of Marswas necessary for protecting life at surface of Mars, valley sedi-ments and even phyllosilicates that date to the Late Noachian orEarly Hesperian such as those in Holden, Eberswalde, or Jezero cra-ters may have been formed in conditions that had already becomeless than favorable for life. Even though such sedimentary sitesprovide invaluable information about surface hydrology and havethe advantage of clear stratigraphic context, their deposition inthe Late Noachian or Early Hesperian may have occurred on a sur-

face subject to a radiation environment that was similar to that ofMars today. If the presence of the magnetic field was a necessaryrequirement for habitability, and exploring habitable conditionsis the goal, this would imply that locations with more ancientmaterials may give us the best hope for detecting traces of life fromearly Mars.

Acknowledgments

We thank Jay Dickson, Bethany Ehlmann, and Ian Garrick-Bethell for helpful discussions. Reviews by Bob Craddock and ananonymous reviewer improved the final manuscript. We gratefullyacknowledge financial assistance from NASA in support of co-investigator participation on the ESA Mars Express High ResolutionStereo Camera Team (JPL Contract 1237163).

References

Acuña, M.H. et al., 1999. Global distribution of crustal magnetization discovered bythe Mars Global Surveyor MAG/ER experiment. Science 284, 790–793.doi:10.1126/science.284.5415.790.

Anderson, R.C., Dohm, J.M., Golombek, M.P., Haldemann, A.F.C., Franklin, B.J.,Tanaka, K.L., Lias, J., Peer, B., 2001. Primary centers and secondaryconcentrations of tectonic activity through time in the western hemisphere ofMars. J. Geophys. Res. 106, 20563–20585. doi:10.1029/2000JE001278.

Ansan, V., Mangold, N., 2006. New observations of Warrego Valles, Mars: Evidencefor precipitation and surface runoff. Planet. Space Sci. 54, 219–242. doi:10.1016/j.pss.2005.12.009.

Baker, V.R., Strom, R.G., Gulick, V.C., Kargel, J.S., Komatsu, G., Kale, V.S., 1991.Ancient oceans, ice sheets and the hydrological cycle on Mars. Nature 352, 589–594.

Barabash, S., Federov, A., Lundin, R., Sauvaud, J.-A., 2007. Martian atmosphericerosion rates. Science 315, 501–503. doi:10.1126/science.1134358.

Barlow, N.G., 1988. Crater size-frequency distributions and a revised martianrelative chronology. Icarus 75, 285–305.

Barlow, N.G., 1990. Constraints on early events in martian history as derived fromthe cratering record. J. Geophys. Res. 95, 14191–14201.

Barnhart, C.J., Howard, A.D., Moore, J.M., 2009. Long-term precipitation and late-stage valley network formation: Landform simulations of Parana Basin, Mars. J.Geophys. Res. 114, E01003. doi:10.1029/2008JE003122.

Bibring, J-P. et al., 2006. Global mineralogical and aqueous Mars history derivedfrom OMEGA/Mars Express data. Science 312, 400–404.

Bibring, J.P., Langevin, Y., 2008. Mineralogy of the martian surface from MarsExpress OMEGA observations. In: Bell, J.F. (Ed.), The Martian Surface. CambridgeUniv. Press, New York, pp. 153–168.

Bouley, S., Ansan, V., Mangold, N., Masson, Ph., Neukum, G., 2009. Fluvialmorphology of Naktong Vallis, Mars: A late activity with multiple processes.Planet. Space Sci. 57, 982–999. doi:10.1016/j.pss.2009.01.015.

Bouley, S., Craddock, R.A., Mangold, N., Ansan, V., 2010. Characterization of fluvialactivity in Parana Valles using different age-dating techniques. Icarus 207, 686–698. doi:10.1016/j.icarus.2009.12.030.

Brain, D.A., Jakosky, B.M., 1998. Atmospheric loss since the onset of the martiangeologic record: Combined role of impact erosion and sputtering. J. Geophys.Res. 103, 22689–22694.

Buczkowski, D.L., Murchie, S., Clark, R., Seelos, K., Seelos, F., Malaret, E., Hash, C.,2010. Investigation of an Argyre basin ring structure using MRO/CRISM. J.Geophys. Res., in press. doi:10.1029/2009JE003508.

Carr, M.H., 1996. Water on Mars. Oxford University Press, New York. 229pp.Carr, M.H., Clow, G.D., 1981. Martian channels and valleys: Their characteristics

distribution, and age. Icarus 48, 91–117.Carr, M.H., Head, J.W., 2010. Geologic history of Mars. Earth Planet. Sci. Lett. 294,

185–203. doi:10.1016/j.epsl.2009.06.042.Chapman, C.R., Cohen, B.A., Grinspoon, D.H., 2007. What are the real constraints on

the existence and magnitude of the late heavy bombardment? Icarus 189, 233–245. doi:10.1016/j.icarus.2006.12.020.

Christensen, P.R., Morris, R.V., Lane, M.D., Bandfield, J.L., Malin, M.C., 2001. Globalmapping of martian hematite mineral deposits: Remnants of water-drivenprocesses on early Mars. J. Geophys. Res. 106, 23873–23885. doi:10.1029/2000JE001415.

Clifford, S.M., Parker, T.J., 2001. The evolution of the martian hydrosphere:Implications for the fate of a primordial ocean and the current state of thenorthern plains. Icarus 154, 40–79.

Cisowski, S., Fuller, M., 1978. The effect of shock on the magnetism of terrestrialrocks. J. Geophys. Res. 83, 3441–3458.

Collinson, D.W., 1986. Magnetic properties of Antarctic shergottite meteorites EETA79001 and ALHA 77005: Possible relevance to a martian magnetic field. EarthPlanet. Sci. Lett. 77, 159–164.

Collinson, D.W., 1997. Magnetic properties of martian meteorites: Implications foran ancient martian magnetic field. Meteorit. Planet. Sci. 32, 803–811. 10.1111/j.1945-5100.1997.tb01571.

C.I. Fassett, J.W. Head / Icarus 211 (2011) 1204–1214 1213

Craddock, R.A., Maxwell, T.A., 1993. Geomorphic evolution of the martian highlandsthrough ancient fluvial processes. J. Geophys. Res. 98, 3453–3468.

Craddock, R.A., Howard, A.D., 2002. The case for rainfall on a warm, wet early Mars.J. Geophys. Res. 107, 5111. doi:10.1029/2001JE001505.

Craddock, R.A., Greeley, R., 2009. Minimum estimates of the amount and timing ofgases released into the martian atmosphere from volcanic eruptions. Icarus 204,512–526. doi:10.1016/j.icarus.2009.07.026.

Di Achille, G., Hynek, B.M., 2010. Ancient ocean on Mars supported by globaldistribution of deltas and valleys. Nature Geosci. doi:10.1038/ngeo891.

Dickson, J.L., Fassett, C.I., Head, J.W., 2009. Amazonian-aged fluvial valley systems ina climatic microenvironment on Mars: Melting of ice deposits on the interior ofLyot Crater. Geophys. Res. Lett. 36, L08201. doi:10.1029/2009GL037472.

Edgar, L.A., Frey, H.V., 2008. Buried impact basin distribution on Mars:Contributions from crustal thickness data. Geophys. Res. Lett. 35, L02201.doi:10.1029/2007GL031466.

Ehlmann, B.L., Mustard, J.F., Fassett, C.I., Schon, S.C., Head, J.W., Des Marias, D.J.,Grant, J.A., Murchie, S.L., 2008a. Clay-bearing minerals and organic preservationpotential in sediments from a martian delta environment, Jezero crater, NiliFossae, Mars. Nature Geosci. 1, 355–358. doi:10.1038/ngeo207.

Ehlmann, B.L. et al., 2008b. Orbital identification of carbonate-bearing rocks onMars. Science 322, 1828–1832. doi:10.1126/science.1164759.

Ehlmann, B.L. et al., 2009. Identification of hydrated silicate minerals on Mars usingMRO-CRISM: Geologic context near Nili Fossae and implications for aqueousalteration. J. Geophys. Res. 114, E00D08. doi:10.1029/2009JE003339.

Ehlmann, B.L., Mustard, J.F., Murchie, S.L., 2010. Geologic setting of serpentinedeposits on Mars. Geophys. Res. Lett. 37, L06201. doi:10.1029/2010GL042596.

Elwood Madden, M.E., Bodnar, R.J., Rimstidt, J.D., 2004. Jarosite as an indicator ofwater-limited chemical weathering on Mars. Nature 431, 821–823.doi:10.1038/nature02971.

Elwood Madden, M.E., Madden, A.S., Rimstidt, J.D., 2009. How long was MeridianiPlanum wet? Applying a jarosite stopwatch to constrain the duration ofdiagenesis. Geology 37, 635–638. doi:10.1130/G25639A.1.

Fassett, C.I., Head, J.W., 2005. Fluvial sedimentary deposits on Mars: Ancient deltasin a crater lake in the Nili Fossae region. Geophys. Res. Lett. 32, L14201.doi:10.1029/2005GL023456.

Fassett, C.I., Head, J.W., 2006. Valleys on Hecates Tholus, Mars: Origin by basalmelting of summit snowpack. Planet. Space Sci. 54, 370–378. doi:10.1016/j.pss.2005.12.011.

Fassett, C.I., Head, J.W., 2007. Valley formation on martian volcanoes in theHesperian: Evidence for melting of summit snowpack, caldera lake formation,drainage and erosion on Ceraunius Tholus. Icarus 189, 118–135. doi:10.1016/j.icarus.2006.12.021.

Fassett, C.I., Head, J.W., 2008a. Valley network-fed, open-basin lakes on Mars:Distribution and implications for Noachian surface and subsurface hydrology.Icarus 198, 37–56. doi:10.1016/j.icarus.2008.06.016.

Fassett, C.I., Head, J.W., 2008b. The timing of martian valley network activity:Constraints from buffered crater counting. Icarus 195, 61–89. doi:10.1016/j.icarus.2007.12.009.

Fassett, C.I., Dickson, J.L., Head, J.W., Levy, J.S., Marchant, D.R., 2010. Supraglacialand proglacial valleys on Amazonian Mars. Icarus 208, 86–100. doi:10.1016/j.icarus.2010.02.021.

Frey, H., Roark, J.H., Shockey, K.M., Frey, E.L., Sakimoto, S.E.H., 2002. Ancientlowlands on Mars. Geophys. Res. Lett. 29, 1384. doi:10.1029/2001GL013832.

Frey, H.V., Frey, E.L., Hartmann, W.K., Tanaka, K.L., 2003. Evidence for buried ‘‘Pre-Noachian’’ crust pre-dating the oldest observed surface units on Mars. LunarPlanet. Sci. 34. Abstract No. 1848.

Frey, H.V., 2006. Impact constraints on, and a chronology for, major events in earlyMars history. J. Geophys. Res. 111, E08S91. doi:10.1029/2005JE002449.

Frey, H., 2008. Ages of very large impact basins on Mars: Implications for the lateheavy bombardment in the inner Solar System. Geophys. Res. Lett. 35, L13203.doi:10.1029/2008GL033515.

Funaki, M., Hoffman, V., Imae, N., 2009. Estimate of the magnetic field of Mars basedon the magnetic characteristics of the Yamato 000593 nakhlite. Meteorit.Planet. Sci. 44, 1179–1191. 10.1111/j.1945-5100.2009.tb01216.x.

Gattacceca, J., Rochette, P., 2004. Toward a robust normalized magnetic paleointensitymethod applied to meteorites. Earth Planet. Sci. Lett. 227, 377–393. doi:10.1016/j.epsl.2004.09.013.

Gendrin, A. et al., 2005. Sulfates in martian layered terrains: The OMEGA/MarsExpress View. Science 307, 1587–1591. doi:10.1126/science.1109087.

Goldspiel, J.M., Squyres, S.W., 1991. Ancient aqueous sedimentation on Mars. Icarus89, 392–410.

Golombek, M.P. et al., 2006. Erosion rates at the Mars Exploration Rover landingsites and long-term climate change on Mars. J. Geophys. Res. 111, E12S10.doi:10.1029/2006JE002754.

Grant, J.A., Schultz, P.H., 1990. Gradational epochs on Mars – Evidence from west–northwest of Isidis Basin and Electris. Icarus 84 (1), 166–195.

Greeley, R., Schneid, B.D., 1991. Magma generation on Mars: Amounts, rates, andcomparisons with Earth, Moon, and Venus. Science 254, 996–998.

Grotzinger, J., 2009. Beyond water on Mars. Nature Geosci. 2, 231–233. doi:10.1038/ngeo480.

Guinness, E.A., Arvidson, R.E., 1977. On the constancy of the lunar cratering fluxover the past 3.3 � 109 years. Proc. Lunar Sci. Conf. 8, 3475–3494.

Gulick, V.C., Baker, V.R., 1990. Origin and evolution of valleys on martian volcanoes.J. Geophys. Res. 95, 14325–14344.

Gulick, V.C., 2001. Origin of the valley networks on Mars: A hydrological perspective.Geomorphology 37, 241–268. doi:10.1016/S0169-555X(00)00086-6.

Haberle, R.M., 1998. Early Mars climate models. J. Geophys. Res. 103, 28467–28479.Hartmann, W.K., 1966. Martian cratering. Icarus 5, 565–576.Hartmann, W.K., 1972. Paleocratering of the Moon: Review of post-Apollo data.

Astrophys. Space Sci. 17, 48–64.Hartmann, W.K., 2005. Martian cratering. 8: Isochron refinement and the chronology

of Mars. Icarus 174, 294–320.Hartmann, W.K., Neukum, G., 2001. Cratering chronology and the evolution of Mars.

Space Sci. Rev. 96, 165–194.Head, J.W., Kreslavsky, M.A., Pratt, S., 2002. Northern lowlands of Mars: Evidence for

widespread volcanic flooding and tectonic deformation in the Hesperian period.J. Geophys. Res. 107, 5003. doi:10.1029/2000JE001445.

Head, J.W., Fassett, C.I., Kadish, S.J., Smith, D.E., Zuber, M.T., Neumann, G.A.,Mazarico, E., 2010. The global distribution of large craters (D P 20 km) on theMoon: Implications for resurfacing history and impactor populations. Science329, 1504–1507. doi:10.1126/science.1195050.

Hodges Jr., R.R., 2002. The rate of loss of water from mars. Geophys. Res. Lett. 29,1038. doi:10.1029/2001GL013853.

Hoke, M.R.T., Hynek, B.M., 2009. Roaming zones of precipitation on ancient Mars asrecorded in valley networks. J. Geophys. Res. 114, E08002. doi:10.1029/2008JE003247.

Hood, L.L., Harrison, K.P., Langlais, B., Lillis, R.J., Poulet, F., Williams, D.A., 2010.Magnetic anomalies near Apollinaris Patera and the Medusae Fossae Formationin Lucus Planum, Mars. Icarus 208, 118–131.

Howenstine, J.B., Kiefer, W.S., 2005. Morphometry of large martian impact craters.Lunar Planet. Sci. 36. Abstract No. 1742.

Hurowitz, J.A., McLennan, S.M., 2007. A �3.5 Ga record of water-limited, acidicweathering conditions on Mars. Earth Planet. Sci. Lett. 260, 432–443.doi:10.1016/j.epsl.2007.05.043.

Hynek, B.M., Beach, M., Hoke, M.R.T., 2010. Updated global map of martian valleynetworks and implications for climate and hydrologic processes. J. Geophys.Res. 115, E09008. doi:10.1029/2009JE003548.

Irwin, R.P., Maxwell, T.A., Howard, A.D., Craddock, R.A., Leverington, D.W., 2002. Alarge Paleolake Basin at the head of Ma’adim Vallis, Mars. Science 296, 2209–2212. doi:10.1126/science.1071143.

Irwin, R.P., Howard, A.D., Craddock, R.A., Moore, J.M., 2005. An intense terminalepoch of widespread fluvial activity on early Mars: 2. Increased runoff andpaleolake development. J. Geophys. Res. 110, E12S15. doi:10.1029/2005JE002460.

Ivanov, B.A., 2001. Mars/Moon cratering rate ratio estimates. Space Sci. Rev. 96, 87–104.Jakosky, B.M., 2008. The Mars atmosphere and volatile evolution (MAVEN) Mars

Scout Mission. In: 3rd Int. Workshop on the Mars Atmosphere: Modeling andObservations (Abstract No. 9036).

Jakosky, B.M., Philllips, R.J., 2001. Mars’ volatile and climate history. Nature 412,237–244.

Jerolmack, D.J., Mohrig, D., Zuber, M.T., Byrne, S., 2004. A minimum time for theformation of Holden Northeast fan, Mars. Geophys. Res. Lett. 31, L21701.doi:10.1029/2004GL021326.

Johnson, C.L., Phillips, R.J., 2005. Evolution of the Tharsis region of Mars: Insightsfrom magnetic field observations. Earth Planet. Sci. Lett. 230, 241–254.doi:10.1016/j.epsl.2004.10.038.

Kirschvink, J.L., Maine, A.T., Vali, H., 1997. Paleomagnetic evidence of a low-temperature origin of carbonate in the martian meteorite ALH84001. Science275, 1629–1633. doi:10.1126/science.275.5306.1629.

Kleinhans, M.G., 2005. Flow discharge and sediment transport models forestimating a minimum timescale of hydrological activity and channel anddelta formation on Mars. J. Geophys. Res. 110, E12003. doi:10.1029/2005JE002521.

Kleinhans, M.G., van de Kasteele, H.E., Hauber, E., 2010. Palaeoflow reconstructionfrom fan delta morphology on Mars. Earth Planet. Sci. Lett. 294. doi:10.1016/j.epsl.2009.11.025.

Kraal, E., van Dijk, M., Postma, G., Kleinhans, M.G., 2008. Martian stepped deltaformation by rapid water release. Nature 451, 973–976. doi:10.1038/nature06615.

Langlais, B., Amit, H., 2008. The past martian dynamo (perspective). Science 321,1784–1785. doi:10.1126/science.1162874.

Lapen, T.J., Righter, M., Brandon, A.D., Debaille, V., Beard, B.L., Shafer, J.T., Peslier,A.H., 2010. A younger age for ALH84001 and its geochemical link to shergottitesources in Mars. Science 328, 347–351.

Lewis, K.W., Aharonson, O., 2006. Stratigraphic analysis of the distributary fan inEberswalde crater using stereo imagery. J. Geophys. Res. 111, E06001.doi:10.1029/2005JE002558.

Lillis, R.J., Manga, M., Mitchell, D.L., Lin, R.P., Acuña, M.H., 2006. Unusual magneticsignature of Hadriaca Patera volcano: Implications for early Mars. Geophys. Res.Lett. 33, L03202. doi:10.1029/2005GL024905.

Lillis, R.J., Frey, H.V., Manga, M., 2008a. Rapid decrease in martian crustalmagnetization in the Noachian era: Implications for the dynamo and climateof early Mars. Geophys. Res. Lett. 35, L14203. doi:10.1029/2008GL034338.

Lillis, R.J., Frey, H.V., Manga, M., Mitchell, D.L., Lin, R.P., Acuña, M.H., Bougher, S.W.,2008b. An improved crustal magnetic field map of Mars from electronreflectometry: Highland volcano magmatic history and the end of the martiandynamo. Icarus 194, 575–596. doi:10.1016/j.icarus.2007.09.032.

Mahaffy, P.R., Cabane, M., Conrad, P.G., Webster, C.R., and the SAM Team, 2009.Sample analysis at Mars (SAM) instrument suite for the 2011 Mars ScienceLaboratory. Lunar Planet. Sci. 40. Abstract No. 1088.

Malin, M.C., Edgett, K.S., 2000. Sedimentary rocks of early Mars. Science 290, 1927–1937. doi:10.1126/science.290.5498.1927.

1214 C.I. Fassett, J.W. Head / Icarus 211 (2011) 1204–1214

Mangold, N., Ansan, V., 2006. Detailed study of an hydrological system of valleys, adelta and lakes in the Southwest Thaumasia region, Mars. Icarus 180, 75–87.doi:10.1016/j.icarus.2005.08.017.

Mangold, N. et al., 2007. Mineralogy of the Nili Fossae region with OMEGA/MarsExpress data: 2. Aqueous alteration of the crust. J. Geophys. Res. 112, E08S04.doi:10.1029/2006JE002835.

McGill, G.E., 1977. Craters as ‘‘fossils’’: The remote dating of planetary surfacematerials. Geol. Soc. Am. Bull. 88, 1102–1110.

McSween, H.Y., 1994. What we have learned about Mars from SNC meteorites.Meteoritics 29, 757–779.

Melosh, H.J., Vickery, A.M., 1989. Impact erosion of the primordial martianatmosphere. Nature 338, 487–489. doi:10.1038/338487a0.

Michalski, J.R., Noe Dobrea, E.Z., 2007. Evidence for a sedimentary origin of clayminerals in the Mawrth Vallis region, Mars. Geology 35, 951–954. doi:10.1130/G23854A.1.

Milliken, R.E. et al., 2008. Opaline silica in young deposits on Mars. Geology 36, 847–850. doi:10.1130/G24967A.1.

Milliken, R.E., Bish, D.L., 2010. Sources and sinks of clay minerals on Mars. Philos.Mag. 90, 2293–2308. doi:10.1080/14786430903575132.

Molina-Cuberos, G.J., Stumptner, W., Lammer, H., Kömle, N.I., O’Brien, K., 2001.Cosmic ray and UV radiation models on the ancient martian surface. Icarus 154,216–222. doi:10.1006/icar.2001.6658.

Moore, J.M., Howard, A.D., Dietrich, W.E., Schenk, P.M., 2003. Martian layered fluvialdeposits: Implications for Noachian climate scenarios. Geophys. Res. Lett. 30,2292. doi:10.1029/2003GL019002.

Morgan, G.A., Head, J.W., 2009. Sinton crater, Mars: Evidence for impact into aplateau icefield and melting to produce valley networks at the Hesperian–Amazonian boundary. Icarus 202, 39–59. doi:10.1016/j.icarus.2009.02.025.

Murchie, S.L. et al., 2009. A synthesis of martian aqueous mineralogy after one Marsyear of observations from the Mars Reconnaissance Orbiter. J. Geophys. Res.114, E00D06. doi:10.1029/2009JE003442.

Mustard, J.F., Poulet, F., Head, J.W., Mangold, N., Bibring, J.-P., Pelkey, S.M., Fassett,C.I., Langevin, Y., Neukum, G., 2007. Mineralogy of the Nili Fossae region withOMEGA/Mars Express data: 1. Ancient impact melt in the Isidis basin andimplications for the transition from the Noachian to Hesperian. J. Geophys. Res.112, E08S03. doi:10.1029/2006JE002834.

Mustard, J.F. et al., 2008. Hydrated silicate minerals on Mars observed by the CRISMinstrument on MRO. Nature 454, 305–309.

Mustard, J.F., Ehlmann, B.L., Murchie, S.L., Poulet, F., Mangold, N., Head, J.W., Bibring,J.-P., Roach, L.H., 2009. Composition, morphology, and stratigraphy of Noachiancrust around the Isidis basin. J. Geophys. Res. 114, E00D12. doi:10.1029/2009JE003349.

Neukum, G., Wise, D.U., 1976. Mars: A standard crater curve and possible new timescale. Science 194, 1381–1387.

Neukum, G., Ivanov, B.A., Hartmann, W.K., 2001. Cratering records in the inner SolarSystem in relation to the lunar reference system. Space Sci. Rev. 96, 55–86.doi:10.1023/A:1011989004263.

Neukum, G., Basilevsky, A.T., Kneissl, T., Chapman, M.G., van Gasselt, S., Michael, G.,Jaumann, R., Hoffmann, H., Lanz, J.K., 2010. The geologic evolution of Mars:Episodicity of resurfacing events and ages from cratering analysis of image dataand correlation with radiometric ages of martian meteorites. Earth Planet. Sci.Lett. 294, 204–222. doi:10.1016/j.epsl.2009.09.006.

Neumann, G.A., Zuber, M.T., Wieczorek, M.A., McGovern, P.J., Lemoine, F.G., Smith,D.E., 2004. Crustal structure of Mars from gravity and topography. J. Geophys.Res. 109, E08002. doi:10.1029/2004JE002262.

Nimmo, F., Tanaka, K., 2005. Early crustal evolution of Mars. Annu. Rev. Earth Planet.Sci. 33, 133–161. doi:10.1146/annurev.earth.33.092203.122637.

Nyquist, L.E., Bansal, B.M., Wiesmann, H., Shih, C.-Y., 1995. ‘‘Martians’’ young andold: Zagami and ALH 84001. Lunar Planet. Sci. 26, 1065–1066.

Osterloo, M.M., Hamilton, V.E., Bandfield, J.L., Glotch, T.D., Baldridge, A.M.,Christensen, P.R., Toranbene, L.L., Anderson, F.S., 2008. Chloride-bearingmaterials in the southern highlands of Mars. Science 319, 1651–1654.doi:10.1126/science.1150690.

Parmentier, E.M., Mustard, J.F., Ehlmann, B.L., Roach, L.H., 2008. Deep hydrothermalcirculation and implications for the early crustal compositional and thermalevolution of Mars. Lunar Planet. Sci. 39. Abstract No. 1544.

Phillips, R.J., Zuber, M.T., Solomon, S.C., Golombek, M.P., Jakosky, B.M., Banerdt, W.B.,Smith, D.E., Williams, R.M.E., Hynek, B.M., Aharonson, O., Hauck, S.A., 2001.Ancient geodynamics and global-scale hydrology on Mars. Science 291, 2587–2591. doi:10.1126/science.1058701.

Pieri, D.C., 1980. Martian valleys: Morphology, distributions, age, and origin. Science210, 895–897.

Plescia, J.B., Saunders, R.S., 1982. Tectonic history of the Tharsis region, Mars. J.Geophys. Res. 87, 9775–9791.

Pollack, J.B., Kasting, J.F., Richardson, S.M., Poliakoff, K., 1987. The case for a wet,warm climate on early Mars. Icarus 71, 203–224.

Poulet, F., Bibring, J.-P., Mustard, J.F., Gendrin, A., Mangold, N., Langevin, Y.,Arvidson, R.E., Gondet, B., Gomez, C., 2005. Phyllosilicates on Mars andimplications for early martian climate. Nature 438, 623–627. doi:10.1038/nature04274.

Quantin, C., Mangold, N., Hauber, E., Flahaut, J., Le Doit, L., Fueten, F., Zegers, T., andISSI International Team on ILDs, 2010. Timing constrains of interior layereddeposit emplacement in Valles Marineris. In: 1st International Conf. on MarsSed. and Strat. (Abstract No. 6048).

Richardson, J.E., 2009. Cratering saturation and equilibrium: A new model looks atan old problem. Icarus 204, 697–715.

Roberts, J.H., Lillis, R.J., Manga, M., 2009. Giant impacts on early Mars and thecessation of the martian dynamo. J. Geophys. Res. 114, E04009. doi:10.1029/2008JE003287.

Rochette, P., Gattacceca, J., Chevrier, V., Hoffmann, V., Lorand, J.P., Funaki, M.,Hochleitner, R., 2005. Matching martian crustal magnetization and magneticproperties of martian meteorites. Meteorit. Planet. Sci. 40, 529–540.

Schubert, G., Russell, C.T., Moore, W.B., 2000. Geophysics: Timing of the martiandynamo. Nature 408, 666–667.

Schultz, P.H., Rogers, J., 1984. Evolution of erosional styles of martian impact basins.Lunar Planet. Sci. 15, 734–735.

Schultz, P.H, 1986. The martian atmosphere before and after the Argyre impact.Reports of Planetary Geology and Geophysics Program – 1985, NASA TM-88383,pp. 188–189.

Segura, T.L., Toon, O.B., Colaprete, A., Zahlne, K., 2002. Environmental effects of largeimpacts on Mars. Science 298, 1977–1980. doi:10.1126/science.1073586.

Soderblom, L.A., Condit, C.B., West, R.A., Herman, B.M., Kreidler, T.J., 1974. Martianplanetwide crater distributions: Implications for geologic history and surfaceprocesses. Icarus 22, 239–263.

Solomon, S., Head, J., 1982. Evolution of the Tharsis province of Mars: Theimportance of heterogeneous lithospheric thickness and volcanic construction.J. Geophys. Res. 87, 9755–9774.

Solomon, S.C. et al., 2005. New perspectives on ancient Mars. Science 307, 1214–1220. doi:10.1126/science.1101812.

Stanley, S., Elkins-Tanton, L., Zuber, M.T., Parmentier, E.M., 2008. Mars’paleomagnetic field as the result of a single-hemisphere dynamo. Science321, 1822–1825. doi:10.1126/science.1161119.

Strom, R.G., Croft, S., Barlow, N., 1992. The martian impact cratering record. In:Kieffer, H. et al. (Eds.), Mars. Univ. Arizona Press, Tucson, pp. 383–423.

Strom, R.G., Malhotra, R., Ito, T., Yoshida, F., Kring, D.A., 2005. The origin of planetaryimpactors in the inner Solar System. Science 309, 1847–1850. doi:10.1126/science.1113544.

Tanaka, K.L., 1986. The stratigraphy of Mars. Proc. Lunar Sci. Conf. 17, J. Geophys.Res. 91, E139–E158.

Tanaka, K.L., Isbell, N.K., Scott, D.H., Greeley, R., Guest, J.E., 1987. The resurfacinghistory of Mars – A synthesis of digitized, viking-based geology. Proc. Lunar Sci.Conf. 18, 665–678.

Tornabene, L.L., McEwen, A.S., and the HiRISE Team, 2008. Recent channel systemsemanating from Hale crater ejecta: Implications for the Noachian landscapeevolution of Mars. Lunar Planet. Sci. 39. Abstract No. 2180.

Toon, O.B., Segura, T., Zahlne, K., 2010. The formation of martian river valleys byimpacts. Annu. Rev. Earth Planet. Sci. 38, 303–322. doi:10.1146/annurev-earth-040809-152354.

Weiss, B.P., Vali, H., Baudenbacher, F.J., Kirschvink, J.L., Stewart, S.T., Shuster, D.L.,2002. Records of an ancient martian magnetic field in ALH84001. Earth Planet.Sci. Lett. 201, 449–463. doi:10.1016/S0012-821X(02)00728-8.

Werner, S.C., 2008. The early martian evolution—Constraints from basin formationages. Icarus 195, 45–60. doi:10.1016/j.icarus.2007.12.008.

Williams, R.M.E., Malin, M.C., 2008. Sub-kilometer fans in Mojave Crater, Mars.Icarus 198, 365–383. doi:10.1016/j.icarus.2008.07.013.

Withers, P., Neumann, G., 2001. Enigmatic northern plains of Mars. Nature 410, 651.doi:10.1038/35070640.

Wray, J.J., Murchie, S.L., Squyres, S.W., Seelos, F.P., Tornabene, L.L., 2009. Diverseaqueous environments on ancient Mars revealed in the southern highlands.Geology 37, 1043–1046. doi:10.1130/G30331A.1.

Wyatt, M.B., McSween, H.Y., 2002. Spectral evidence for weathered basalt as analternative to andesite in the northern lowlands of Mars. Nature 417, 263–266.doi:10.1038/417263.

Zhong, S., Zuber, M.T., 2001. Degree-1 mantle convection and the crustal dichotomyon Mars. Earth Planet. Sci. Lett. 189, 75–84. doi:10.1016/S0012-821X(01)00345-4.