Mars: The evolutionary history of the northern lowlands based on crater counting and geologic mapping

Planetary and Space Science 59 (2011) 1143–1165

Contents lists available at ScienceDirect

Planetary and Space Science journal homepage: www.elsevier.com/locate/pss

Mars: The evolutionary history of the northern lowlands based on crater counting and geologic mapping S.C. Werner a,, K.L. Tanaka b, J.A. Skinner Jr.b a b

Physics of Geological Processes, University of Oslo, Norway Astrogeology Science Center, U.S. Geological Survey, Flagstaff, Arizona, USA

a r t i c l e i n f o

abstract

Article history: Received 15 April 2010 Received in revised form 18 February 2011 Accepted 29 March 2011 Available online 9 April 2011

The geologic history of planetary surfaces is most effectively determined by joining geologic mapping and crater counting which provides an iterative, qualitative and quantitative method for defining relative ages and absolute model ages. Based on this approach, we present spatial and temporal details regarding the evolution of the Martian northern plains and surrounding regions. The highland–lowland boundary (HLB) formed during the pre-Noachian and was subsequently modified through various processes. The Nepenthes Mensae unit along the northern margins of the cratered highlands, was formed by HLB scarp-erosion, deposition of sedimentary and volcanic materials, and dissection by surface runoff between 3.81 and 3.65 Ga. Ages for giant polygons in Utopia and Acidalia Planitiae are  3.75 Ga and likely reflect the age of buried basement rocks. These buried lowland surfaces are comparable in age to those located closer to the HLB, where a much thinner, post-HLB deposit is mapped. The emplacement of the most extensive lowland surfaces ended between 3.75 and 3.4 Ga, based on densities of craters generally 4 3 km in diameter. Results from the polygonal terrain support the existence of a major lowland depocenter shortly after the pre-Noachian formation of the northern lowlands. In general, northern plains surfaces show gradually younger ages at lower elevations, consistent local to regional unit emplacement and resurfacing between 3.6 and 2.6 Ga. Elevation levels and morphology are not necessarily related, and variations in ages within the mapped units are found, especially in units formed and modified by multiple geological processes. Regardless, most of the youngest units in the northern lowlands are considered to be lavas, polar ice, or thick mantle deposits, arguing against the ocean theory during the Amazonian Period (younger than about 3.15 Ga). All ages measured in the closest vicinity of the steep dichotomy escarpment are also 3.7 Ga or older. The formation ages of volcanic flanks at the HLB (e.g., Alba Mons (3.6–3.4 Ga) and the last fan at Apollinaris Mons, 3.71 Ga) may give additional temporal constraint for the possible existence of any kind of Martian ocean before about 3.7 Ga. It seems to reflect the termination of a large-scale, precipitation-based hydrological cycle and major geologic processes related to such cycling. & 2011 Elsevier Ltd. All rights reserved.

Keywords: Mars Geologic evolution Lowlands Crater statistics Paleao-ocean

1. Introduction One of the dominating features of the Martian physiography is its division into heavily cratered highlands, which occupy the southern hemisphere, and mostly featureless low-lying plains, which occupy the northern hemisphere. The lowland region, roughly centered on the north pole, covers one-third of the planet and is broadly characterized by a smooth, gently sloping surface (at kilometer-scale) (Kreslavsky and Head, 2002), with elevations below the mean planetary radius (e.g., Smith et al., 1999). Topography data reveal a difference in elevation between the

 Corresponding author. Tel.: þ 47 22856472.

E-mail address: [email protected] (S.C. Werner). 0032-0633/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.pss.2011.03.022

northern and the southern hemispheres of about 5 km (e.g., Smith et al., 1998). In the eastern hemisphere along Terrae Sabaea and Cimmeria, the dichotomy boundary is characterized by a prominent scarp and possible extensional and contractional features (Watters, 2003), while other parts in the western hemisphere along Arabia and Tempe Terrae have mainly gentle slopes. Tharsis, one of the large volcanic provinces, and the possible volcanically formed Medusae Fossae formation partially bury the dichotomy escarpment, and several large impact basins appear to have contributed to its current exposure. Various formation mechanisms have been suggested for the northern plains basin; for example, its origin has been attributed to tectonism driven by mantle dynamics (e.g., Wise et al., 1979; McGill and Dimitriou, 1990; Sleep, 1994; Zhong and Zuber, 2001; Zuber, 2001) or to impact cratering (e.g., Wilhelms and Squyres, 1984; Frey and Schultz, 1988;

1144

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

McGill, 1989; Marinova et al., 2008; Nimmo et al., 2008; AndrewsHanna et al., 2008). However, any clear morphologic evidence has been obscured by resurfacing events. Gravity signatures indicate the Martian crust thins northward in the vicinity of the highland– lowland boundary (Janle, 1983; Zuber et al., 2000). Crustal magnetic anomalies are generally stronger in the highlands compared to the lowlands, indicating that the Martian dynamo had ceased prior to (or perhaps as a consequence of) the formation of the northern plains (Connerney et al., 2001). Prior to Mars Global Surveyor (MGS; Orbital insertion date: 11 September 1997. Launch date: 7 November 1996), mapping and stratigraphic analysis of Mars’ global geology was largely based on Viking Orbiter image data (Scott and Tanaka, 1986; Greeley and Guest, 1987; Tanaka and Scott, 1987). However, a clearer understanding of the physiographic characteristics and geologic evolution of the northern lowlands was afforded by topographic data of the Mars Orbiter Laser Altimeter (MOLA) and additional image data of the Mars Orbiter Camera (MOC), both onboard MGS. In particular, Tanaka et al. (2003, 2005) used these datasets to significantly revise the boundaries and ages of geological units in the northern plains. The main definition of the new lowland units was guided by the demarcation between predominantly Noachian ( 4 3.7 Ga) densely cratered materials that characterize the equatorial and southern highlands of Mars and the younger, mostly plains-forming materials that typically lie below an elevation of  2000 m. Major geographical features include the Borealis, Utopia, and Isidis basins, which are largely defined, respectively, by the Vastitas Borealis, Utopia Planitia, and Isidis Planitia plains regions. All geologic units and structures that occur within the boundaries of these lowland topographic basins are distinctive because of their separated geologic evolution since early in Mars’ evolution. As such, these became repositories of regional sequences of materials marked by a unique assemblage of landforms. The post-Viking geological map of the northern plains (Tanaka et al., 2005) resulted in two new definitions applied to the global stratigraphic scheme of Tanaka (1986). First, the pre-Noachian period was defined as the age of primordial crustal rocks underlying and predating but incorporated as fragments within Early Noachian outcrops. The northern plains depression and Utopia basin within it are topographic features that apparently formed during the pre-Noachian as no outcrops of materials predating the features are evident. The second major change was a refinement of the Hesperian/Amazonian boundary (3.15 Ga) as the surface age of Vastitas Borealis marginal and interior units. This modification from previous geological map depictions of the Vastitas Borealis units continues to provide a basis for subdividing primary (depositional) and secondary (modificational) geologic processes within the Martian northern plains. The time-stratigraphic classification of the northern plains geologic units of Tanaka et al. (2005) is based mainly on superposition relations among the geologic units as well as the overall density of craters larger than 5 km in diameter for each unit (Ncum ðD Z 5 kmÞ) from the crater catalog published by Barlow (1988, 2001) and the Ncum ðD Z5 kmÞ epoch boundary values given by Tanaka (1986). Tanaka et al. (2003) discussed the relevance of these crater frequencies with respect to their correctness and stated three problems with their approach: (1) the misregistration between the topographically based geological maps and the Viking image-based crater catalog, which might attribute craters to a unit to which they do not belong or vice versa, affecting smaller units the most; (2) taking the cumulative number for a single reference diameter (here 5 km and larger) instead of the entire crater size-frequency distribution, which inherently obscures the ability to discern resurfacing event(s) within the crater population, particularly of small vertical scale (i.e., those

affecting smaller diameter craters); and (3) the exclusion of highly degraded craters in the crater catalog, so that only relatively pristine craters with preserved ejecta blankets were included (these were interpreted to superpose and thus postdate the unit). By investigating the full cratering record using the crater sizefrequency distribution, it is not necessary to exclude or pre-sort crater populations. Significant resurfacing will be visible as deviations from the fitted crater curve in cumulative crater plots and different events of sufficient magnitude can be separated (Werner, 2009). Northern plains surfaces reflect complex resurfacing histories, which may relate to a variety of processes occurring at a range of rates and temporal overlap. These complexities make the northern plains favorable for detailed analysis through the combination of map- and crater-based stratigraphic interpretations. Here, we present a sample of crater size-frequency distributions for representative localities of major geologic units of the northern plains. In combination with geologic map relations, we discuss the evolution of the northern lowlands and surrounding terrains that is implied by these counts. The main focus in the first part of the study is to detail the modeled ages for major episodes of geologic activity within the lowlands, based on key areas guided by the most recent geological mapping (Tanaka et al., 2005). In the second part of the paper, we present detailed properties of the lowlands’ crater size-frequency distributions observed in Utopia Planitia and giant polygonal terrain of Utopia and Acidalia Planitiae. We discuss these regional observations in the context of the northern plains geologic record, with specific focus on the existence of a postulated Oceanus Borealis (Baker et al., 1991).

2. Absolute ages in resurfaced units—method and uncertainties Remote-sensing based delineation of geological units by relative age, following morphologic or spectral characteristics, relies on superposition principles and other stratigraphic relations among surface units and features (e.g., Hansen, 2000; Tanaka et al., 2005). However, these are largely qualitative assessments and do not provide information on absolute age. Because of the local variability in crater densities, separated outcrops of equivalent units generally cannot provide consistent ages relative to one another. For many planetary surfaces, the relative abundances (and size-frequency distribution) of impact craters distinguish outcrops of geological units quantitatively by relative age. Determination of a surface’s geologic history most effectively depends on not only geologic mapping, which delineates materials by surface and natural breaks in the rock sequence, but also crater counting, which provides a quantitative tool for defining relative ages and modeling absolute ages (e.g., Tanaka, 1986; Hartmann and Neukum, 2001). Although a combination of these two approaches is optimal, the application of crater-based ages is inherently more complex and, thus, requires more knowledge of and adherence to theory, method, and procedure. The theoretical concept and mathematical background of agedating techniques based on crater counts was developed in the 1960s and 1970s and summarized by the Crater Analysis Techniques Working Group (Arvidson et al., 1979). Absolute ages on Mars and other solid surface objects are determined by a crater production function and an applied cratering chronology model. The crater production function describes the expected crater sizefrequency distribution recorded in a geological unit at a specific time, and is, when scaled, comparable between planetary bodies such as the Moon and Mars (e.g., Neukum and Wise, 1976; Neukum and Ivanov, 1994; Ivanov, 2001). The transfer of craterbased functions from the Moon to Mars requires the knowledge of

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

Cum. Crater Frequency N(D> 1km) [km-2]

the impacting body flux ratio between the lunar case to Mars, with consideration of celestial mechanics and how the conditions of crater formation differ from one planet to another. The latter conditions are mainly defined by the respective gravity and target properties of the impacted bodies (Ivanov, 2001). The transfer of relative into absolute ages by the use of a chronology model is based on the assumptions that (1) the crater production function—the original size-frequency distribution—is accurately known (Ivanov, 2001, originally in Neukum et al., 1975), (2) the flux of projectiles onto the surfaces is isotropic, and (3) any target-property variations are negligible. With these assumptions, the crater densities measured on planetary surfaces that have been exposed for the same amount of time with respect to diameter are equivalent in age. Relative ages based on Ncum ðD,tÞ are given for a fixed diameter D (e. g., 1, 2, 5, 16 km); these are commonly referred to as crater-retention ages (Hartmann, 1966). The crater chronology model applied herein is based on measurements of the crater size-frequency distribution on areas of the Moon, which are linked to radiometric ages retrieved from lunar rocks sampled and returned during the Apollo missions. This distribution is transferable to other (inner solar system) planets and solid surface bodies, assuming that all bodies underwent the same bombardment (e.g., Neukum and Wise, 1976; Neukum and Hiller, 1981; Hartmann, 1973, 1977, 1978; and unified in Hartmann and Neukum, 2001). Three phenomena—saturation, contamination by secondary cratering, and geological resurfacing—have the potential to cause deviation from the crater production function. In densely cratered (i.e., old) units, the rate of production and destruction eventually becomes steady, starting at small diameters. Such an equilibrium distribution exhibits a characteristic cumulative slope on a powerlaw fit of b¼  2 at the small crater range (e.g., Woronow, 1977; Richardson, 2009). If measurements are contaminated by secondary craters, the observed crater size-frequency distribution may appear steeper than the crater production function suggests (e.g., Shoemaker, 1965). Other deviation from the crater production function typically implies a resurfacing event (erosion or deposition). Although separation of different surface units might sometimes not be possible due to limited image resolution or lack of stratigraphic contrasts, effects of geologic resurfacing are detectable in the crater size-frequency distribution (as long as there are statistically sufficient crater populations). In such cases, the age of the resurfacing event can be estimated because it effectively

Single Resurfacing

1145

removes or obscures the cratering record starting at the smaller size range. Alternatively, resurfacing can altogether reset the cratering record in instances where the surface is completely renewed. In either case, the determination of a resurfacing event based on crater density is dependent on the event’s magnitude. A distinct drop in the crater size-frequency distribution of a resurfaced geologic unit signifies a resurfacing event (Fig. 1A or B); multiple episodes of resurfacing can be implied by multiple drops in the distribution. In cases of ongoing resurfacing (e.g., on steep slopes of Martian surface structures) or saturation, the crater size-frequency distribution continually crosses isochrons and cannot be linked to a discrete age (Fig. 1C). For units where resurfacing has been observed the determination of ages has to be limited to a certain diameter range (below or above a diameter d , where the resurfacing has influenced the crater size-frequency distribution). The oldest age is derived straightforwardly, by fitting the production function to the large-diameter range. For the resurfacing event, if no correction has been applied, the age is overestimated (dependent on the magnitude of the resurfacing event) due to the cumulative character of the crater size-frequency distribution. This problem is avoided in incremental distribution plots (Arvidson et al., 1979, and customarily used by W. K. Hartmann). The general description of the cumulative crater size-frequency distribution Ncum ðD,tÞ is given by the cumulative number of craters per area for a given diameter D at a certain time t: Z 1Z t 0 0 Ncum ðD,tÞ ¼ gðD0 Þ dD  f ðt 0 Þ dt ð1Þ %

D

0

for an undisturbed geological unit (Fig. 2A) a simple solution can be found Ncum ðD,tÞ ¼ GðDÞ  FðtÞ,

ð2Þ

where G represents the cumulative crater size-frequency distribution and F the flux of projectiles. It is required that the crater size-frequency distribution is stable over time, and only the flux changes (decays, e.g., Hartmann and Neukum, 2001). In a unit where subsequent processes (erosion and deposition) occurred, the smaller crater population has been modified preferentially (compare Fig. 2A and B). Subsequently, the surface accumulates craters as if it is a fresh surface (Fig. 2C). On a planetary surface, a proper remote-sensing method to distinguish both populations is not known (compare Fig. 2C and D). An

Mult iple Resurfacing

Cont inual Resurfacing

Crater Diameter D [km] Fig. 1. Three different schematic crater size-frequency distributions show resurfacing events which have (A) a distinct timing visible in the re-steepening of the crater sizefrequency distribution at smaller sizes, (B) is followed by a second distinct event, and (C) is occurring continuously. At the smallest crater range a flattening of the curve is observed which is related only to the resolution limit of the image data.

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

Cum. Cra equency N (D> 1km) [km-2]

1146

3.3 Ga 3.7 Ga

2.2 Ga 3.3 Ga 3.7 Ga

2.17

Ga 3.70

Crater Diameter D [km]

Ga

Crater Diameter D [km]

Fig. 2. Schematic description of a (A) cratered surface, (B) the same surface affected by erosion erasing craters of smaller diameters, (C) the same surface after continued cratering. In reality the surface would look as shown in (D), where the old and younger population cannot be distinguished (stippled and solid lines in C). (E) depicts the resulting crater size-frequency distribution plotted with isochrons, and (F) shows the same crater size-frequency distribution but also the smaller range crater sizefrequency distribution corrected for the resurfacing event. This results in a reduced age for the resurfacing event compared to the uncorrected one.

example of a measured resurfacing crater size-frequency distribution is given in Fig. 2E. Any attempts to distinguish between erosional state or other parameters of the craters can be strongly misleading (and are subjective). To understand the crater population accumulated on the surface after the resurfacing event, a method is needed to separate both populations. A first approach is described by Werner (2005, 2009), which was later adopted by Michael and Neukum (2010) using a different numerical implementation. An iterative treatment is necessary to determine the age(s) of resurfacing event(s) by fitting the well-known crater production function to the crater-size range below a certain diameter to restore the full crater size-frequency distribution of the resurfacing event and the age of the younger surface (Fig. 2F). In cases where additional resurfacing processes took place, the correction has to be repeated at the next cut-off diameter. Prerequisites for the crater count statistics are (1) careful geologic mapping in order to delineate homogeneous, temporally unique units, and (2) exclusion of sublimation pits, volcanic and secondary craters, and other non-primary impact craters to the extent possible. Contamination due to unrecognized global secondary craters unwittingly included in the measurements is less than 10% (old surfaces), and in most cases less than 5% (Werner, 2005; Werner et al., 2009). For determination of initially relative ages and subsequently absolute ages, the cumulative number at a certain diameter

(Ncum—value of an individual diameter bin) is not taken, but rather the best representing crater-production function (Ivanov, 2001) of the observed crater size-frequency measurements. The fitting procedure follows the Marquardt–Levenberg algorithm, a weighted non-linear least-squarespffiffiffiffi fit (Levenberg, 1944; Marquardt, 1963). The statistical error ( N) of the measurement indicates the uncertainties of the relative ages and the fit quality depends on the number of bins which can be used as fit range. The uncertainty of such a fit for making use of the whole size range for the derivation of a mean N-value at a certain diameter is of the order of the statistical error of 2s and less than the individual bin error of 1s. The level of uncertainty of the scaled cumulative number N per bin is given by  1=2  N 7 sN ¼ 7 ð3Þ A for each bin. Here, the reliability of the ages is improved by statistical measures if a larger area and/or a larger number of craters is counted. The fitted crater frequencies given for a reference diameter are translated to absolute ages by applying a cratering chronology model. In the course of this work the reference diameter is Ncum ¼ Dð Z 1 kmÞ. For Mars, the applied chronology model is transferred from the Moon (Hartmann and Neukum, 2001; Ivanov, 2001). A variety of lunar cratering chronology models existed previously and agreed

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

within a factor of 2–3. These different models have converged to a unified model (Hartmann and Neukum, 2001; Ivanov, 2001). The cratering chronology used in this paper is given by Ncum ð Z1 kmÞ ¼ 3:22  1014 ðe6:93T 1Þ þ4:875  104 T

ð4Þ

according to Hartmann and Neukum (2001). However, they do not present this equation numerically but as a diagram. Any possible uncertainties inherent in the lunar cratering chronology model might be found in the cratering chronology model for Mars (Hartmann and Neukum, 2001; Ivanov, 2001) because of the transfer from Moon to Mars. Therefore, this uncertainty of the Martian cratering chronology, causes a possible systematic error in model ages (up to a factor of 2 for younger, Amazonian surfaces and about 100 Ma for older, Hesperian and Noachian surfaces). Possible inherent uncertainties in the cratering chronology may be significantly larger than our quoted uncertainties (e.g., Hartmann and Neukum, 2001). However, only these statistical uncertainties from fitting the isochrons are listed. Other uncertainties may be introduced if the counting units and crater diameters are determined on inappropriately projected image maps. We have tested the range of uncertainties resulting for the area determination based on different map projections used for our image data and a sinusoidal (equal-area) projection, and found differences o 2% for the area sizes. However, these variations in areal size are insignificantly small compared to the statistical uncertainties of the ages determined here ( o 0:002%). Measurements resulting in considerable age ranges can be caused by complex resurfacing histories (or random statistical fluctuations in crater populations). Crater populations measured on image data of relatively low resolution (in our case 231 m/pxl) do not permit a discussion of any resurfacing events acting on the removal of craters in the small-size range (less than about 1 km in diameter). Craters smaller than 1 km in diameter cannot be detected due to resolution limits, therefore, possible resurfacing effects on these smaller craters cannot be assessed.

3. Northern lowland stratigraphy We selected representative type areas that have simple outlines and contain a single geological unit for crater size-frequency distribution measurements. These areas were extracted from the Viking-based Mars Digital Image Model (MDIM) 2.1 global dataset (a complete orthorectified mosaic at a resolution of 231 m/pxl; Archinal et al., 2003) as well as from the THEMIS daytime IR global mosaic (a mosaic at a resolution of 99 m/pxl available at the USGS). The location of the crater-count areas is shown in Fig. 3. The image clippings and measured crater size-frequency distributions are only collectively presented in Appendix A. Based on the crater size-frequency measurements, the resulting relative and absolute surface ages and the time-stratigraphic relations of the investigated units are described in the following section and will be discussed with respect to previously suggested timestratigraphic relations (e.g., Scott and Tanaka, 1986; Greeley and Guest, 1987; Tanaka and Scott, 1987; Tanaka et al., 2003, 2005). Special focus has been given to four different zones, which are interpreted to represent contrasting geologic characteristics and evolutionary histories within the northern plains, based on results from geologic mapping. These zones include the Chryse region (1), the Utopia and Isidis basins (2), the Amazonis and Elysium Planitiae (3), and the Scandia region (4). Densely cratered units are common at the margins of the northern plains and generally grade morphologically into one another. The Libya Montes is the oldest unit, and forms massifs and other high-standing terrains characteristic of ancient, large impact basin rims. The Noachis Terra unit makes up rugged

1147

terrain characterizing most of the highland surface. The Nepenthes Mensae unit consists of knobs and mesas of highland rock and interposed slope and plains material, which forms much of the highland–lowland margin. Locally, relatively thick lava plains marked by wrinkle ridges make up the Lunae Planum and Tempe Terra units. The highland units are represented by a number of type areas for which crater counts were performed (Appendix A, Fig. A1). Their distribution and locations are shown in Fig. 3 and the ages and dimensions are summarized in Table 1. These older units generally range in age from 4.04 to 3.6 Ga. Geologically, the Noachis Terra unit (the oldest unit counted in this study in areas 20, 37, and 47) is interpreted to reflect a long history of resurfacing due to secondary erosional and depositional processes as well as superposed tectonism. These regions are characterized by high crater densities, valley networks, isolated depressions (which may be subdued, eroded impact craters), as well as ridges, scarps, and troughs. Although resurfacing events in these units are recognized in the crater size-frequency distribution, gradual deviation from the expected crater-production function suggests that such events are generally not limited to single episodes, but rather continual processes of surface degradation and alteration. Werner (2008) argued that there is no mapped unit anywhere on Mars older than about 4.1 Ga, which is close to the crater-retention age found for Noachis Terra.

3.1. The Chryse region The physiographic setting is dominated by Chryse Planitia, a topographic depression about 1500 km in diameter proposed to be an impact basin (Schultz et al., 1982). The largest and most prominent outflow channels on Mars (Kasei, Maja, Shalbatana, Simud, Tiu, Ares, and Mawrth Valles) emanate from chaotic terrains and drain into the roughly circular Chryse Planitia, which opens northward into the lowland plains. Except for the arcuate Xanthe Montes, which forms a possible basin rim on the western margin of Chryse Planitia, clear evidence for a Chryse impact basin cannot be confidently identified, perhaps because of the removal of rim structures by earlier fluvial processes. The Chryse basin does not show a distinctive gravity anomaly that would be indicative of a large impact (Yuan et al., 2001). The circum-Chryse outflow channels were mainly active during the Hesperian (Tanaka et al., 1992; Rotto and Tanaka, 1995), postdating the development of highland valley networks. Baker (1982) argued that the Chryse channels are large-scale complexes of fluid-eroded troughs that emanated from collapse zones in regional chasmata and chaos regions. Channel relics, floodplains along channel courses, depositional streamlined bars, and erosional islands led to the interpretation that channel-systemforming floods originated from aquifer outbreaks (Carr, 1979). Although catastrophic flooding is the most commonly accepted interpretation for the origin of the circum-Chryse outflow channels, morphologic features indicate that these have been additionally sculpted by diverse erosional forces including winds, debris or mud flows, glaciers, and lavas (e.g., Chapman et al., 2010a,b). However, these interpretations are consistent with the hypothesis that the Vastitas Borealis units represent ocean sediments derived from material discharge from the Chryse channels (e.g., Parker et al., 1993; Baker et al., 1991). Some channel systems show no obvious source regions, but possibly emanate at fracture and fissure zones, including those associated with impact craters. Most source-region situations are believed to be causatively related to volcanic and tectonic processes due to their occurrence within or near volcanic materials and/or zones of deformation, including Valles Marineris and the Tharsis and Elysium rises (Basilevsky et al., 2009).

1148

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

Elevation (km) <-8

0

>21

Fig. 3. The northern hemisphere of Mars with geological units as mapped by Tanaka et al. (2005) shown with the MOLA derived elevation model (A) and color-coded geological units (B). The outline of the numbered crater-counted areas are given, for detailed information such as unit name, stratigraphic classification, average (median) elevation, area size, crater statistics, crater density and model age information see Table 1.

Table 1 Listed absolute ages and other statistical parameters of the crater size-frequency statistics measured for the patches shown in Fig. 3. Relative and absolute ages were derived first for crater size-frequency distributions measured on Viking image data and refined through crater size-frequency distributions measured on THEMIS daytime IR. Viking-based ages in parentheses are considered inaccurate and are rejected for the discussion. Ncum(1 km) values represent the best fit to the measured crater size-frequency distributions at the reference diameter Ncum ¼ Dð Z 1 kmÞ. They are translated to absolute ages by applying a cratering chronology model as given in Section 2, Eq. (4). Ages annotated by are resurfacing ages and are derived as described in Section 2. The elevation value is the mean elevation and its standard deviation given for each patch. %

Results of the crater size-frequency measurements # Geological unit Stratigraphy (following Tanaka et al., 2005)

Chryse Planitia region—zone 1 47 Noachis Terra

Nn

Longitude

6.73

Latitude

39.74

Mean elevation (m)

 3083.04 (423.32)

Area (km2)

78565.80

47

Age (Ga)

4.84e  2

4.04

3.29e  3 Noachis Terra

Nn

 44.23

9.39

 2030.61 (353.55)

29343.16

Themis IR image-based

Ncum (1 km); (km  2)

1.43e  2

47 37

Viking image-based

%

%

2.86e  2

3.85 3.55 3.96

þ 0:07 0:19 þ 0:03 0:04 þ 0:04 0:06 þ 0:04 0:06

Ncum (1 km); (km  2)

Age (Ga)

4.84e  2

4.04

1.43e  2 3.29e  3

1.63e  3 HNTl

 62.43

34.00

 1181.81 (129.97)

28183.27

2.89e  3

33

Chryse Planitia 1

HNCc1

 26.05

18.89

 3064.58 (154.12)

43710.23

(1.27e  2)

33 32

4.12e  3 Chryse Planitia 1

HNCc1

 28.05

21.00

 3340.09 (163.03)

51913.23

9.51e  3

39

Chryse Planitia 1

HNCc1

 45.25

16.97

 3316.9 (174.65)

19709.69

5.12e  3

41

Chryse Planitia 1

HNCc1

 52.07

19.39

 2608.92 (287.89)

49003.48

4.12e  3

32

4.48e  3

þ 0:05 0:08 þ 0:04 (3.83 0:06 ) þ 0:04 3.61 0:07 þ 0:08 3.78 0:29 þ 0:03 3.63 0:04 þ 0:05 3.66 0:09 þ 0:05 3.61 0:08

3.51 %

%

41  45.80

14.40

 3002.49 (66.18)

15150.80

8.50e  3

3.76

þ 0:05 0:09

8.50e  3

Chryse Planitia 2

HCc2

 47.68

40.73

 3711.45 (78.50)

36851.95

5.91e  3

3.69

þ 0:06 0:13

5.91e  3

Chryse Planitia 2

HCc2

 24.71

28.74

 3886.19 (86.64)

57768.01

(2.46e  3)

(3.45

Chryse Planitia 3

HCc3

 45.57

27.83

 3758.09 (91.50)

57338.82

3.29e  3

3.55

þ 0:06 0:12

3.29e  3

Chryse Planitia 4

HCc4

 39.46

23.70

 3870.77 (72.35)

45956.70

3.53e  3

3.57

3.53e  3

Simud Vallis

Hcs

 39.12

12.99

 3791.84 (251.22)

48466.58

3.53e  3

3.57

þ 0:10 1:00 þ 0:13 0:38 þ 0:07 0:21

36

Vastitas Borealis interior

ABVi

 36.42

57.76

 4897.44 (84.97)

75354.35

3.66e  3

3.58

þ 0:05 0:09

3.66e  3

36 46

Vastitas Borealis interior

ABVi

 29.32

36.89

 4286.79 (112.33)

122717.67

2.65e  3

3.48

þ 0:06 0:12 þ 0:27 0:28 þ 0:09 0:33

8.92e  4

30

2.06e  3

30 35

9.95e  4 þ 0:10 0:38 )

35 40 40 31

5.21e  4

31 28

1.78e  3

%

3.25

28

8.92e  4 Vastitas Borealis marginal

ABVi

 38.13

30.06

 3934.39 (64.76)

53615.86

%

2.98e  3

1.83 3.52

29

Nn

113.95

5.12

30.1408 (245.26)

21992.89

4.84e  2

4.04

þ 0:04 0:06

20 22

113.75

9.20

 1266.73 (448.68)

25101.61

1.13e  2

Nepenthes Mensae

HNn

67.83

36.20

 2047.79 (451.97)

54652.38

4.89e  3

22 08 08

2.65e  3

3.81 %

3.48 3.65

þ 0:05 0:10 þ 0:09 0:34 þ 0:05 0:08

%

3.36 3.69

%

2.04 3.63

%

2.91 3.55

%

1.07 3.57

%

3.25 3.57

%

2.06

%

3.52 %

4.89e  3

1.76

4.04 %

%

1.13e  2 2.65e  3

1.83

3.85 3.47 3.81

%

3.48

3.65 Resurfacing, compare Fig. 1

þ 0:07 0:19 þ 0:03 0:04 þ 0:04 0:06 þ 0:04 0:06 þ 0:03 0:04 þ 0:13 0:28 þ 0:05 0:08 þ 0:03 0:04 þ 0:04 0:07 þ 0:08 0:29 þ 0:03 0:04 þ 0:05 0:09 þ 0:05 0:08 þ 0:22 0:22 þ 0:05 0:09 þ 0:09 0:25 þ 0:06 0:13 þ 0:24 0:26 þ 0:06 0:13 þ 0:13 0:25 þ 0:06 0:12 þ 0:17 0:18 þ 0:10 1:00 þ 0:13 0:38 þ 0:07 0:21 þ 0:31 0:32 þ 0:05 0:09 þ 0:06 0:12 þ 0:27 0:28 þ 0:09 0:33 þ 0:22 0:23

þ 0:04 0:06 þ 0:03 0:04 þ 0:05 0:09 þ 0:05 0:10 þ 0:09 0:34 þ 0:05 0:08

1149

HNn

1.76 3.76

4.84e  2 2.59e  3

Nepenthes Mensae

3.61 %

2.98e  3

1.43e  2

20

3.63 3.66

3.58 Resurfacing, compare Fig. 1 2.65e  3 3.48

8.58e  4

Utopia Planitia region—zone 2 20 Noachis Terra

%

3.53e  3 1.00e  3

46 29

1.78e  3

3.61 3.78

4.48e  3 1.44e  3

3.15 3.89

%

4.12e  3

HCc2

3.74 3.51

5.12e  3 8.58e  4

38

%

9.51e  3

Chryse Planitia 2

38

%

1.62e  2

4.48e  3

3.55 3.96

2.89e  3 4.12e  3

3.85

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

7.63e  3

37 Lunae Planum

%

2.86e  2

37 34

%

1150

Table 1 (continued ) Results of the crater size-frequency measurements # Geological unit Stratigraphy (following Tanaka et al., 2005)

Longitude

Latitude

Mean elevation (m)

Area (km2)

Viking image-based Ncum (1 km); (km

Themis IR image-based

2

)

Age (Ga)

26

Syrtis Major Planum

HIs

77.80

9.79

 1538.17 (1191.75)

82849.72

3.40e  3

3.56

26 21

Utopia Planitia 1

HBu1

113.55

14.83

 2610.9 (237.81)

29623.39

(8.50e  3)

(3.76

þ 0:03 0:04 þ 0:04 0:06 )

21 Utopia Planitia 1

HBu1

73.33

39.46

 2917.93 (202.27)

50116.06

4.89e  3

09 27

Utopia Planitia 2

HBu2

94.41

20.51

 3505.56 (58.19)

76588.43

4.12e  3

27

9.85e  4 Utopia Planitia 2

HBu2

113.40

19.29

 3521.39 (144.19)

28897.43

3.65

2.46e  3 1.34e  3

%

AHEe

112.83

33.44

 4727.12 (55.12)

31919.11

2.98e  3

Elysium rise

AHEe

127.31

26.61

 4161.61 (44.28)

37639.36

2.73e  3

Tinjar Valles a

AEta

132.13

41.07

 4598.41 (77.14)

82297.54

(3.40e  3)

Tinjar Valles a

AEta

127.53

32.76

 4334.67 (79.34)

56956.72

2.65e  3

Tinjar Valles a

AEta

110.66

46.37

 4937.72 (30.21)

134931.48

1.82e  3

13

Isidis Planitia

AIi

86.95

6.42

 3827.67 (50.01)

24735.53

2.40e  3

13 12

Astapus Colles

ABa

82.65

44.22

 3787.93 (120.63)

88187.19

4.29e  3

14

1.56e  3

15 16

8.09e  4

þ 0:05 0:09

þ 0:06 0:12 þ 0:51 2.02 0:52 þ 0:10 3.45 0:40 þ 0:40 2.73 0:55 þ 0:13 3.52 1:80 þ 0:21 3.08 0:50 þ 0:09 3.49 0:38 þ 0:38 1.66 0:39 þ 0:06 (3.56 0:14 )

3.61 %

Elysium rise

15

%

%

16 17

2.78e  4

%

þ 0:04 0:07 þ 0:06 0:13

 3596.83 (44.07)

72416.07

3.29e  3

3.55

Vastitas Borealis interior

ABVi

78.10

58.64

 3815.02 (67.27)

60256.33

(4.11e  3)

(3.61

4.63e  4

3.08 3.49

%

1.66 3.48

%

1.24 3.48

%

1.30 3.27

%

0.95

2.40e  3

3.44 Resurfacing, compare Fig. 1 4.29e  3 3.62

þ 0:03 0:04 þ 0:03 0:05 þ 0:04 0:07 þ 0:05 0:09 þ 0:06 0:12 þ 0:51 0:52 þ 0:10 0:40 þ 0:40 0:55 þ 0:13 1:80 þ 0:21 0:50 þ 0:09 0:38 þ 0:38 0:39 þ 0:07 0:14 þ 0:21 0:22 þ 0:08 0:20 þ 0:25 0:27 þ 0:17 0:83 þ 0:17 0:19 þ 0:09 0:28 þ 0:04 0:07

Resurfacing, compare Fig. 1 þ 0:06 3.29e  3 3.55 0:13 þ 0:21 4.36e  4 0.894 0:23 %

þ 0:05 0:10 )

2.52e  3 4.12e  4

Vastitas Borealis interior

ABVi

95.55

28.07

 3817.26 (92.05)

26260.76

3.96e  3 2.13e  3

Vastitas Borealis interior

ABVi

112.70

26.18

 4304.55 (112.52)

40096.88

Vastitas Borealis interior

ABVi

111.50

64.81

 4207.89 (83.87)

238625.69

3.60 %

1.93e  3

3.32

1.28e  3

2.62

136.39

4.00

 2661.78 (77.57)

63195.11

5.12e  3 1.13e  3

Arcadia Planitia

HAa

170.32

36.56

 3802.92 (97.84)

130296.54

þ 0:08 0:29 þ 0:08 0:18 þ 0:20 1:90 þ 0:42 0:56

3.96e  3 2.13e  3 1.28e  3

%

3.53e  3

þ 0:52 0:81

1.29e  3

3.66

þ 0:04 0:07 þ 0:45 0:47 þ 0:05 0:09

5.12e  3

2.32 3.57

1.13e  3

Elysium rise

AHEe

161.07

22.58

 1911.31 (292.88)

135372.49

4.89e  3

Elysium rise

AHEe

139.33

16.32

 1029.86 (757.63)

121332.66

2.80e  3

1.32e  3

3.65 %

2.69 3.50

þ 0:04 0:06 þ 0:29 0:35 þ 0:05 0:08

3.66 %

3.53e  3 1.17e  3

02

%

1.93e  3

2.65

2.80e  3

2.32 3.57

%

4.89e  3 1.32e  3

þ 0:08 0:20 þ 0:16 0.845 0:17 þ 0:08 3.60 0:29 þ 0:08 3.38 0:18 þ 0:20 3.32 1:90 þ 0:42 2.62 0:56 þ 0:07 (3.43 0:12 ) þ 0:52 2.65 0:81

3.46 %

(2.35e  3) 1.29e  3

HBu2

3.38

01

03

%

2.65e  3

3.62

53.27

04

03

8.09e  4

2.73 3.52

2.73e  3

3.44

67.01

Amazonis Planitia region—zone 3 04 Utopia Planitia 2

02

1.56e  3

3.27 0.57

3.45 %

2.98e  3

1.82e  3

ABVm

19

01

1.34e  3

þ 0:17 0:83 þ 0:27 0:28 þ 0:09 0:28

Vastitas Borealis marginal

24 19

2.46e  3

6.34e  4

25 24

%

2.65e  3

11 25

3.65 Resurfacing, compare Fig. 1 4.12e  3 3.61 9.85e  4 2.02

þ 0:08 0:20

10 11

4.89e  3

6.05e  4

18

12 10

3.56 Resurfacing, compare Fig. 1 8.50e  3 3.80 2.52e  3 3.46

3.48

17 18

3.40e  3

2.40 3.65

%

2.69

3.50 Resurfacing, compare Fig. 1

þ 0:04 0:07 þ 0:45 0:47 þ 0:05 0:09 þ 0:20 0:21 þ 0:04 0:06 þ 0:29 0:35 þ 0:05 0:08

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

23 14

Age (Ga)

%

09

23

Ncum (1 km); (km  2)

þ 0:06 0:14 þ 0:29 0:32

0.948 %

4.80e  4

3.56 3.40e  3 87061.21  4488.57 (177.25) 72.44  128.10 ABVi

(3.64 (4.68e  3) 44483.76  3934.83 (115.81) ABVm

 122.83

67.63

 3449.75 (198.79) 63.52  119.24 HTa Scandia region

þ 0:06 0:14

Vastitas Borealis interior 44 45

45

1.04

3.33 Resurfacing, compare Fig. 1 3.40e  3 3.56

%

5.07e  4

1.96e  3

þ 0:07 0:20 )

Vastitas Borealis marginal 44

(3.53 (3.08e  3) 140044.89

128222.94  2618.85 (281.98) 58.21

Amazonis Planitia 2 N

AAa2n

 116.84

 161.67

29.85

 3908.2 (25.74)

 4047.82 (39.02) 47.48  172.10 AAa1n Amazonis Planitia 1 N

43

1.00

2.06e  3 þ 0:07 0:13 )

43

þ 0:09 0:28

3.37 2.09e  3

6.78e  4

180451.26

231954.99

1.82e  3 07

 3879.03 (51.99) 36.90  146.71 AAa1n Amazonis Planitia 1 N

3.36

þ 0:09 0:28 þ 0:13 0:13 þ 0:09 0:21 þ 0:12 0:13 þ 0:12 0:56

3.37

%

4.88e  4

2.09e  3

43

Between Alba Mons and the North Pole—zone 4 42 Alba Mons HTa

%

2.73e  5

6.78e  4 þ 0:65 0:77

1.39

(0.965

%

1.95e  4

(4.70e  4) 112428.17

06

06

05

05

%

1.90e  5

%

1.46e  4

07

1.82e  3

07

1.95e  4

þ 0:58 0:68 ) þ 0:08 0.399 0:09 þ 0:18 3.27 1:10

8.68e  4

%

1.78

þ 0:69 0:81 þ 0:08 0.399 0:09 þ 0:18 3.27 1:10 þ 0:18 0.299 0:20 þ 0:02 0.039 0:03 þ 0:65 1.39 0:77 þ 0:01 0.056 0:01

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

1151

The Ares and Simud Valles units represent materials covering parts of the outflow channel floors, whereas the Chryse Planitia 1–4 units represent materials emplaced by outflow channel dissection as well as by mass wasting of the highland margin (Tanaka et al., 2005). Scoured channels extend across the Chryse region but are buried by the Vastitas Borealis interior and marginal units (Tanaka et al., 2005, and references therein), which themselves appear to be resurfaced vestiges of a precursor unit. The landforms, broadly characterizing the Vastitas Borealis interior and marginal units, are indicative of periglacial-like resurfacing, localized sedimentary diapirism and mud volcanism, and cyclic deposition and erosion of decameter-scale mantle deposits. In total, 16 type areas were used to describe the timing of the evolution in this region (Fig. 3). They were selected to be representative for each geological unit in this region. Most of the counted areas represent units belonging to the Chryse Planitia’s inner slope region. Ages determined for the Chryse type areas (the Chryse Planitia 1–4 units) range from 3.83 to 3.25 Ga. The variations in age are reflected by changes in morphology and support spatially linked areas that show erosional activity and depositional units. Many depositional units correlate with younger ages as they are found at lower elevation. For example, we sampled the Chryse Planitia 1 unit in four type areas (numbers þ 0:03 32, 33, 39, and 41), and the ages range from 3.89 0:04 to 3.61 þ 0:05 0:08 Ga. The crater size-frequency distributions measured in these areas of the Chryse Planitia 1 unit dominantly result in the formation and resurfacing surface ages around 3.63 Ga (averaged), but crater counts for two of these areas (numbers 32 and 33) reveal an older age (about 3.85 Ga, averaged). In particular, the surface in the southern part of area 33 resembles degraded highland terrain, while other parts that resemble smooth plains with wrinkle ridges and fluvial overprint are visible along with many knobs (Appendix A, Fig. A1). Areas representing the Chryse Planitia 2 (30, 35, 38) and 3 (40) units generally fall into the same age range (3.67 Ga, averaged), but all of these surfaces show resurfacing events constrained by their crater size-frequency distributions. The outflow channel activity is illustrated by type areas from the Simud Vallis (28) and Chryse Planitia 4 (31) units þ 0:07 with an age of 3.57 0:21 Ga. In the Chryse Planitia 4 unit, a distinct þ 0:13 resurfacing event is indicated at about 3.25 0:38 Ga ago, while the resurfacing event observed in the crater size-frequency distribution measured in the Simud Vallis unit ended more recently þ 0:31 (2.06 0:32 Ga). Surface morphology and crater counts collectively suggest a resurfacing that is continuously erosional rather than due to depositional activity. The surface downslope of Simud Vallis appears to be resurfaced in a more homogeneous manner, perhaps by fluvial deposition, as suggested by surface morphology. The Vastitas Borealis (29, 36, 46) units show slightly younger formation ages of around 3.52 Ga with resurfacing around 1.8 Ga. Hesperian ridged plains materials such as those mapped in Lunae Planum are interpreted to be the bulk of materials scoured by the Chryse outflow channels. Ages of Lunae Planum materials indicate a presumptive upper-time limit for the start of the erosive phase of outflow channel activity. However, Lunae Planum unit type area (in this case from Tempe Terra (34), Fig. 3) has an þ 0:05 age of 3.51 0:08 Ga, contradicting the view that these materials entirely pre-date circum-Chryse erosion. Ages for the nearby outflow channel units are found to be older, which contrasts with geologic mapping results. The favored interpretation, then, might be that the process(es) that emplaced (or modified) Lunae Planum ended after the chief formational phase of the Kasei channel system. Alternatively, this might be a classic example of craterbased superposition details that are insufficiently observed in the Viking-MDIM 2.1 global dataset. Detailed studies of Kasei Valles (Tanaka and Chapman, 1992; Chapman et al., 2010a) suggest that the Lunae Planum unit was resurfaced during the Kasei flooding.

1152

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

Major parts of the channels were excavated before 3.5 Ga ago, while any subsequent resurfacing (fluvial and/or volcanic) only marginally shaped the inner landforms and surfaces of the channels (Chapman et al., 2010b). Similar findings were made at the confluence of Tiu and Ares Valles (Marchenko et al., 1998).

3.2. The Utopia and Isidis basins and vicinity The Utopia basin has also been considered to be of impact origin and appears relatively circular, about 3200 km in diameter and 1–3 km deep (Thomson and Head, 2001). No impact basin rim features are preserved, though their existence has been postulated by the occurrence and distribution of surficial landforms including knob fields, subtle topographic depressions, and cone and vent features (McGill, 1989; Skinner and Tanaka, 2007). Overlapping Utopia, the Isidis basin rim is preserved on its southern flank as the uplifted and eroded massifs of Libya Montes. The Nili Fossae are arcuate grabens that partly demarcate the northwestern outline of the impact structure. Both the Utopia and Isidis basins display high-amplitude gravity anomalies indicating buried mascons (e.g., Mohit and Phillips, 2007; Muller and Sjogren, 1968). Utopia basin as mapped by Tanaka et al. (2005) is surrounded by highland materials made up of the Libya Montes, Noachis Terra, Nepenthes Mensae, and Syrtis Major units. The Utopia basin margin forms undulating plains that slope down from the highland margin and are made up of the Utopia Planitia 1 and 2 units at progressively greater distance from the highland– lowland boundary margin. Sheet lavas and volcanic debris flows emanate from the western flank of the Elysium rise and extend westward more than 2000 km into the Utopia basin, covering much of its floor and burying Early Amazonian Vastitas Borealis materials. These units include the Elysium rise and Tinjar Valles (a and b) units (Fig. 3). The Tinjar Valles units source from northwest-trending fissures and troughs of Elysium Fossae and generally bury lava flows of the Elysium rise unit. The Tinjar Valles b unit delineates materials emplaced through confined flow to form the Granicus, Tinjar, Apsus and Hrad Valles, while the Tinjar Valles (a) unit represents the complex and rugged overbank and sheet flow units that generally define the boundary of the Utopia channeled lobe. On the northwest margin of Utopia basin, the Astapus Colles unit thinly buries other lowland units and is interpreted to represent a degraded ice-rich mantle deposit (e.g., Mustard et al., 2001). Isidis basin is also surrounded by ancient highland materials and is bounded to the west by the volcanic flows of the Syrtis Major Planum unit and abuts Utopia basin along a low topographic saddle covered by the Utopia Planitia 2 unit. The Isidis Planitia unit makes up almost the entire Isidis basin floor. Flow tongues are visible in the southeast, originating from Amenthes Planum. We crater counted 19 areas representing nearly all the aforementioned geological units (see Fig. 3 and Table 1; for the crater size-frequency distributions and image data, Appendix A). In this region, the best-fit ages range from the formation ages of 4.04 þ 0:04 þ 0:16 0:06 (Early Noachian) to resurfacing ages of 0.845 0:17 Ga (Late Amazonian). The oldest surface of this region occurs within the Noachis Terra unit (20), likewise as for the circum-Chryse region. Roughly along the 1201E-meridian, several type units provide a surface-age sequence across the Utopia basin (areas 14, 18–24). The resulting surface ages gradually decrease with decreasing elevation and toward the center of the Utopia basin. All units involved show at least one resurfacing event, and all units were resurfaced (20–22) or formed (14, 18, 19, 23, 24) at about 3.48 Ga. The stratigraphic profile line is apparently interrupted by areas representing the Elysium rise and Tinjar Valles units. The Elysium

rise (14, 15) unit in Utopia basin has a model age of about 3.5 Ga, þ 0:21 but for both units resurfacing occurred that ended at 3.08 0:50 Ga þ 0:38 Ga (15). The crater size-frequency distribution (14) and 1.66 0:39 measured for the Tinjar Valles units (16–18) indicates the formation ages of 3.48 Ga, followed by more or less continuous resurfacing ending more than 1 Ga ago. The surface formation age of the Vastitas Borealis interior (19) unit seems to be the þ 0:5 youngest, about 2.65 0:8 Ga. However, the crater size-frequency distribution measured on the THEMIS mosaic appears to have an excess of craters for diameters between 2 and 5 km, and can be þ 0:07 fitted by an isochron of 3.43 0:12 . This apparent modification of large-diameter craters was first described by McGill (1986) and will be further discussed in Section 4. The ages of Utopia Planitia 1 unit were compared between the þ 0:03 þ 0:05 eastern flank (3.80 0:05 Ga; 21) and western flank (3.65 0:09 Ga; 9). Similarly, the Utopia Planitia 2 unit, which was mapped based on broad morphologic uniformity, shows different surface ages deterþ 0:06 þ 0:1 mined for areas 27 (3.61 0:12 Ga) and 23 (3.45 0:4 Ga); these ages appear to agree with local stratigraphic sequences of progressively younger surfaces with decreasing elevation but also might indicate poor regional correlation in the ages. This results in some type areas (8, 9, 12) that represent stratigraphically and morphologically different units to have similar crater ages around 3.6 Ga. In other words, even though the morphologic stratigraphy is consistent from region to region, the unit surface ages are time transgressive at a regional scale. Alternatively, geologic contacts may be misplaced in some regions, particularly where their lateral continuity is uncertain (as identified by approximate contacts). Crater counts for area 13 on the Isidis Planitia floor provide an þ 0:09 age of about 3.44 0:28 Ga, while the basin itself originated at about 3.96 70.01 Ga (Werner, 2008). Nili and Meroe Paterae, the probable source for Syrtis Major volcanic flows, show an age of about 3.73 Ga with resurfacing ending at 2.5 Ga (Werner, 2009) in agreement with Hiesinger and Head (2004). Here, the crater þ 0:03 counts for this unit (area 26) indicate an age of 3.56 0:04 Ga, which is close to the latest-stage activity of other paterae (Werner, 2009).

3.3. The Elysium volcanic province and Amazonis Planitia The third zone of interest includes the volcanic province Elysium and surrounding plains units (Fig. 3). Three volcanoes, Elysium Mons, Hecates Tholus and Albor Tholus, as well as other local fissure vents on the southern flank of the Elysium rise and in Amazonis and Arcadia Planitiae, constitute major volcanic vents in this region. From these, several volcanic lobate flow units were sourced that cover a rather large area of the northern lowlands (Tanaka et al., 2005). Amazonis Planitia was the original type region for the youngest of the Martian stratigraphic periods (Scott and Carr, 1978). The Arcadia Planitia unit represented by area 1 has an age of þ 0:05 3.57 0:09 Ga. Areas 2 and 3 sample widely separated surfaces of the Elysium rise unit, yet both yield crater ages near 3.60

þ 0:05 0:1

Ga.

þ 0:04 0:07

Ga) is found for area 4 (Utopia Planitia A similar age (3.66 2 unit) SSW of Elysium Mons. In other areas surrounding the Elysium rise, such as around Athabasca Valles, the same age of 3.66 Ga has been found in earlier investigations (Werner et al., 2003; Murray et al., 2005). Two areas representing units in northern Amazonis Planitia west of Olympus Mons indicate ages of Middle to Late Amazonian þ 0:65 þ 0:69 (1.39 0:77 Ga, area 5; 1.78 0:81 Ga, area 6 with a resurfacing þ 0:08 Ga, which is reflected in surface expression event ending 0.4 0:08 by mantling). Although areas 6 and 7 belong to the same unit

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

(Amazonis Planitia 1 north unit), area 7 appears to represent a þ 0:18 much older surface with an age of about 3.27 1:1 Ga. Most of the large craters in area 7 are embayed or partly covered by younger deposits and the shape of the crater distribution indicates continuous resurfacing at craters smaller than 3–4 km in diameter. Similar measurements across the entire Amazonis Planitia demonstrated a strong spatial variation in crater densities (Werner, 2005). Although the eastern part of Amazonis Planitia (close to the Olympus Mons aureole) is characterized homogeneously by Amazonian plains-forming lava flow units, their emplacement may have began in the Hesperian (as the number of buried craters suggest) and may be not as thick as in other areas of Amazonis Planitia. THEMIS daytime-infrared images reveal multiple flow features, and suggest a temporal subdivision, although the process forming the plains stayed the same. 3.4. Alba Mons to the north pole This region includes the broad volcanic edifice of Alba Mons, marking the northernmost extent of the Tharsis rise, and its adjacent plains. From north to south, this surface is composed of the Scandia region, Vastitas Borealis marginal and interior units, and the Alba Mons unit (Tanaka et al., this issue). We produced crater counts for each of these units along a profile roughly following the 2401E-meridian. The selected areas follow a gentle slope down into the Borealis basin. Ages based on crater sizefrequency distributions for areas 42, 43, 44, and 45 indicate a surface age ranging from 3.56 to 3.37 Ga. All four units show resurfacing ending at about 1 Ga. Differences between crater counting ages performed on Viking and THEMIS image bases show the most dramatic differences for surfaces in this region. Not only do slight age variations of 100–200 Ma result, but an erroneous age sequence would be derived if only based on Viking image data (surface ages are overestimated for the units represented by the patches 43 and 44, see Table 1). In these cases, clouds and haze in the Viking images limited crater detectability, so that some albedo variations were erroneously interpreted as craters. 3.5. The highland–lowland dichotomy boundary The varying geologic, morphologic, and stratigraphic characteristics of the highland–lowland boundary are perhaps best seen between Chryse and Isidis Planitiae. Topographically, the highland–lowland boundary changes from a gently sloping surface in Arabia Terra to a steep escarpment bordering Terrae Sabaea and Cimmeria. The transition longitude is at about 3301W, and is marked coincidentally but not causatively by the most prominent fresh impact structure of the northern lowlands, the 221.2-km-diameter, double-ringed Lyot crater (centered at 501N, 3311W). The location of this crater marks also a change in Bouguer gravity anomalies and calculated crustal thickness (Janle, 1983; Neumann et al., 2004). The heavily cratered highlands crustal thickness is estimated to be nearly everywhere thicker than 60 km, while the lowland crust ranges from 20 to 40 km in thickness. Although Arabia Terra displays heavily cratered highland morphology, its relatively low elevation as it slopes into the lowlands results in a thinning crustal thickness profile and smaller lowland crustal thicknesses (Zuber et al., 2000). Using preliminary northern plains map units and boundaries of Tanaka et al. (2003), Werner (2005) investigated the crater-size frequencies in order to unravel the differences in topography and crustal structure between regions east and west of Lyot crater, which also correspond to correlating or different episodes of unit development. Ages in this area range from about 4 to 3.1 Ga

1153

(Werner, 2005). The are plotted in Fig. 4. These ages decrease in age from old to young across the dichotomy and correlate with the south-to-north decreasing elevation from the highlands to the lowlands. Within given geological units, younger ages are found in western versus eastern outcrops. The highland surface ages range between 3.98 and 3.89 Ga. These Early to Middle Noachian ages are representative (see above) and are in good agreement with an almost homogeneously formed highland plateau, where large-scale surface modifying processes are dominated by impact cratering. The crater size-frequency distribution measured close to the Isidis basin reveals the oldest surface age and resurfacing reflecting an eventful history since impact basin formation (Frey et al., 1988). The eastern highland–lowland boundary is expressed by an escarpment that originated through dissection of highland plateau material. Here, a broad age range for surfaces at the base of the escarpment between 3.83 and 3.67 Ga (Late Noachian) is observed. Areas measured closer to the escarpment appear older than those more distant from the escarpment. Larger craters are still visible on the remaining blocks and mesas of the dissected highland plateau, while farther away smooth lowland plains dominate the terrain. As a general trend, younger surface ages are found towards topographic lows in the lowlands. Plains units east of Lyot formed during the Late Hesperian. Compared to the eastern part of the investigation area, only a single transition unit is present. The ages for this unit vary between 3.74 and 3.68 Ga, or Late Noachian. In the Deuteronilus region, we obtain Noachian ages if measuring all craters on mesas and the lower units together, but if just the lowland parts are considered, the age is as young as the lowland units west of Lyot (about 3.15 Ga, Early Amazonian). The impact structure Lyot formed about 3.4 70.05 Ga ago (Werner, 2008). The temporal separation between the eastern and western part of the investigated dichotomy boundary area, indicated in morphology, topography and gravity anomalies, has been confirmed by crater counts (Werner, 2005). 3.6. Chronostratigraphical inventory of the northern lowlands In the Chryse and Utopia–Isidis basin regions, the Noachis Terra unit surfaces formed between 4.04 and 3.96 Ga. This unit generally forms the lowland-bounding cratered highlands, where the longevity of cratering processes appears to have resulted in resurfacing of highland rocks (based on multiple intersections of isochrons). In Utopia Planitia, the Nepenthes Mensae unit ages reflect the morphologic transition between highlands and lowlands, an equivalent interpretation to map-based stratigraphies. Similarly, in the circumChryse region, the Chryse Planitia 1–4 units that make up the region’s southern slope have ages from 3.89 to 3.6 Ga. These units generally tend to become gradually younger with decreasing elevation and increased latitude. Many mapped units (although discriminated primarily by superposition relationships) are interpreted to overlap in both age and morphologic expression. Outflow channel floors mostly formed prior to 3.55 Ga; those adjacent to floodplains appear slightly younger. The ages of Vastitas Borealis units in the Chryse region are around 3.4570.05 Ga, similar to their ages in the Scandia region between the north pole and Alba Mons (3.33–3.56 Ga). The ages of Vastitas Borealis units in the Utopia basin and its vicinity are similarly distributed, whereas the Utopia basin-marginal plains ages are about 3.60 Ga, with higher elevations showing older ages (3.85 Ga). Especially in the Utopia basin, a gradual decrease in ages towards topographic lows is observed, with a minimum age of about 3.32 Ga for the lowest lying parts of the Vastitas Borealis interior unit (24). Lava flows, lahars, and channels (Elysium rise and Tinjar Valles units) occurred less than 3.5 Ga ago, followed by later phases of resurfacing between 1.5 and 1 Ga ago.

1154

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

10−5

Highlands

Volcanic deposits/plains

0.1

LA 10−4

0.2 0.3 0.4 0.5 1.0

MA 10−3

1.5 2.0 2.5 3.0

EA LH

10−2

3.5 3.6

EH LN

3.7 Themis IR day

formation age resurfacing age

MN

Viking MDIM 2.1

formation age resurfacing age formation age

EN

Viking MDIM 1

formation age

Amazonis Planitia 1−2 unit

3.8 3.9 4.0 4.1

Lyot

Lunae Planum

Alba Patera

Syrtis Major Planum

Elysium rise/Tinjar Valles

Isidis Planitia

Arcadia Planitia unit

Nepenthes Mensae unit

Noachis Terra unit

Astapus Colles unit

Dichotomy

Vastitas Borealis interior Vastitas Borealis marginal

Utopia Planitia 1−4 unit

Chryse Planitia 1−4 unit

Ares/Tiu Valles(#)

Kasei Valles (*)

10−1

Cratering ModelAge[Ga]

Cum.CraterFrequency[km-2]atD=1km

0.05

Fig. 4. Summary of crater frequencies Ncum (1 km) (left scale) and model ages derived applying the cratering chronology model by Hartmann and Neukum (2001) (right scale) for outflow channels, lowland units, highland areas, and volcanic deposits and plains units. These measurements have been performed on Viking MDIM2.1 mosaics (squares and dashed empty circles, the latter symbolize resurfacing events) and Themis IR daytime moasaiced images (circles, empty circles symbolize resurfacing events). These datasets are supplemented by studies around the highland–lowland dichotomy boundary (black squares, compare Section 3.5), in Ares and Tiu Valles (black stars, Marchenko et al., 1998), and Kasei Valles (black stars, Lanz, 2004), all measured on Viking image data. Horizontal lines show the epoch boundaries as published in Hartmann and Neukum (2001), they vary only for the Amazonian boundaries (dashed are Hartmann’s favored boundaries, solid the ones favored by Neukum.)

Most volcanic activity in and around the northern lowlands may have occurred within a few hundred million years. The Lunae Planum unit appears to have been emplaced up to 3.51 Ga. The Syrtis Major Planum unit shows a similar age (3.56 Ga), though the surface of Isidis Planitia is younger (3.44 Ga). The Elysium flank and adjacent plains flows of Arcadia Planitia were emplaced between 3.57 and 3.48 Ga. Our crater counts in the extensive Amazonis Planitia 1 north unit suggest multiple events. Although these multiple events are not recognized everywhere, morphologically we interpret these plains to have formed by the emplacement of sheet lavas, and the time frame is very extended. Determined ages range from 3.27 to 0.04 Ga. The construction of northern Alba Mons concluded around 3.37 Ga, coinciding with the formation of the characteristic landforms observed in the Scandia region unit (3.36 Ga) situated north of the Alba Mons. However, the emplacement and/or major resurfacing of most sedimentary units located within the Martian lowlands appears to be broadly post-date the formation of major volcanic constructs. Major plains-related volcanic flows (e.g., in Utopia and Amazonis Planitiae) overlap with the emplacement and/or resurfacing of lowland units. Although most constructs were in place by the Early Amazonian, it appears that volcanic and sedimentary units may complexly interfinger along the margins of the Martian lowlands. When translating the absolute surface ages (Table 1) to geological epochs (Fig. 4; see also Section 5), most lowland geologic units appear to have been emplaced (or underwent major resurfacing) during the Hesperian, with the exception of the highland units (Noachian age). Only a few regions, such as the plains northwest of Olympus Mons, are Amazonian in age. However, there is morphologic and stratigraphic evidence that major resurfacing was common in regions of the lowlands.

4. Giant polygonal trough units in the northern lowlands Landforms such as polygonal patterned terrain so-called giant polygons, are result of major resurfacing events. Utopia, Elysium, and Acidalia Planitiae are occupied by extensive areas of these giant polygons. Originally, this polygonally fractured terrain was observed on the Mariner 9 B-frames (Mutch et al., 1976) and revealed, accompanied by other morphologies, that large portions of the Martian surface likely was modified by the action of water. Later, high-resolution Viking photographs resolved in greater detail the locations and morphology of the giant polygons. In general, they consist of 200–800-m-wide steep-walled and flat-floored troughs, some tens of meters deep, the intersections of which form polygons that are 5–30 km in diameter (Pechmann, 1980). These features occur in southeastern Acidalia Planitia, northwestern Elysium Planitia, and the low elevation margins of Utopia Planitia, mostly in the Martian northern mid-latitudes. These giant polygons are much larger than small-scale polygons that are interpreted to form by periglacial processes at higher latitudes (Rossbacher and Judson, 1981). In previous mapping approaches (Scott and Tanaka, 1986; Greeley and Guest, 1987; Tanaka and Scott, 1987), the giant polygonal terrain occupying regions were mapped as the grooved member (Hvg), one of four members of the Vastitas Borealis Formation. Textural and albedo differences as well as morphologic characteristics have been used to outline these different members. Pechmann (1980) concluded that the polygonal trough units intersect units of all other Vastitas Borealis members. McGill (1986) investigated crater distributions of these members, which revealed apparent variability in the distribution shapes. A discussion about the origin of the giant polygons and crater distribution variability is inherently linked to the question: Did

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

large standing bodies of water fed by catastrophic flood events exist in the past? We find that our crater count and map-based geologic and stratigraphic assessments can provide some information for addressing this outstanding question.

4.1. Crater distributions in polygonal troughed terrain (Acidalia and Utopia Planitiae) Various hypotheses for the formation of giant polygons have been proposed, including thermal cooling and contraction in permafrost (Carr and Schaber, 1977; Carr et al., 1976), desiccation of water-saturated sediments (Morris and Underwood, 1978), cooling of lava (Morris and Underwood, 1978; Masursky and Crabill, 1976; Carr et al., 1976), or tectonic deformation (Mutch et al., 1976; Carr et al., 1976). Pechmann (1980) showed that no terrestrial analogs satisfactorily describes the scales observed or the mechanisms involved. McGill and Hills (1992) explained the observed large size of the polygons and accounted for the stresses responsible for their formation by plate-bending and finiteelement models. These assessments suggested the polygons formed by thermal contraction and vertical compaction, perhaps during the cooling of water-saturated sediments or volcanic rocks accompanied by differential compaction over buried topography. The giant polygonal pattern is accompanied by single or double ring-like graben structures, which are assumed to occur above buried craters. These have been used to estimate the thickness of the deforming overburden material (McGill, 1986; McGill and Hills, 1992). Lucchitta et al. (1986) suggested that the troughed material is of sedimentary origin and was deposited in a standing body of water by emphasizing the spatial correlation of giant polygons, outflow channels mouths, and topographic lows within the lowlands. Another explanation for the origin of undulating surfaces and polygonal crack patterns of this size is based on the formation through Rayleigh convection in rapidly emplaced water-rich sediments (Lane and Christensen, 2000). Elastic rebound has also been proposed for the formation of giant polygons, a process which implies uplift of the northern Martian lowlands crust following the evaporation or sublimation of a water or ice bodies (Hiesinger and Head, 2000). As shown in Fig. 6, even though craters that superpose fields of giant polygons may appear rather pristine, they are embayed and/or subdued by later deposits. These deposits appear to drape the original relief and form circular graben structures within the polygonal troughed terrain. These circular grabens have diameters between 8 and 30 km, and are believed to be fracture patterns tracing buried craters (Carr, 1981; McGill, 1986), yielding a regional cover thickness of about 600 m (McGill and Hills, 1992). These single- and double-ringed grabens are associated with slight topographic depressions seen in the MOLA topography (Buczkowski and McGill, 2002). Based on the modeled graben spacing of doubleringed troughs, Buczkowski and Cooke (2004) suggested a cover thickness of 1–2 km. Furthermore, MOLA topographic data also revealed the presence of roughly circular basins with diameters of several tens of kilometers in both the Martian highlands and lowlands, which are usually not associated with any structural feature in image data. These were called quasi-circular depressions and are potential candidates for large, buried or deeply eroded impact structures (e.g., Frey et al., 2002. Buried basins of diameter ranging between 20 and 70 km in diameter would refer to a cover thickness between 1 and 2 km, respectively, (Garvin et al., 2003). We measured the crater size-frequency distribution specifically for the polygonal-troughed terrain region in the Utopia and Acidalia Planitiae (Fig. 5A and D) and recorded an unusual deficiency of large craters (Fig. 5C and F). Initially, crater distributions of clearly visible craters were measured and resulted in

1155

þ 0:02 a surface age of 3.52 0:03 Ga for craters larger than 1.4 km in diameter for both locations. Additionally, the distribution of the ringed-graben structures was measured assuming that they are the imprinting crater population of the underlying surface or basement. Locations of both ghost and visible craters are shown in Fig. 5B and E. The distribution of visible craters was stacked with the population of ghost craters (Cruikshank et al., 1973) for both regions in Utopia and Acidalia Planitiae (Fig. 5C, F). The combined visible and ghost crater populations yield an age of þ 0:02 3.77 0:02 Ga for Utopia Planitia for craters larger than 6 km in þ 0:02 Ga for craters diameter and a slightly younger age of 3.68 0:03 larger than 3 km in diameter for Acidalia Planitia’s basement. This implies that a substantial amount of deposition took place in parts of Acidalia and Utopia Planitiae. Using estimates of Garvin et al. (2003), in order for buried craters to form circular grabens with diameters about 30 km, a layer of compactable (sedimentary) materials must be 4 350 m thick.

4.2. Cratering variations in Utopia Planitia Crater morphologies observed in Utopia Planitia seem to be characteristically of Mars. However, craters as large as about 14 km in diameter are bowl-shaped, like simple craters on the Moon or Mercury. A similar size-range for the transition from simple-to-complex interior crater morphology in the Martian lowland plains has been noted by Wood et al. (1978). In principle, central peak morphology is expected for craters larger than 5 km (Pike, 1980), but has been observed for craters as small as 1.5 km in diameter in heavily cratered highland units (Mouginis-Mark, 1979). Pristine craters larger than 15 km are not present in this part of Utopia Planitia. Many craters show ejecta morphologies which are single- or double-lobed ramparts, so-called fluidized ejecta blankets. They are believed to indicate the presence of volatiles (e.g., water ice or water in the subsurface, Carr et al., 1977). On Earth, craters which were formed in shallow marine (sedimentary) environment commonly show a diminished relief compared to their on-shore counterparts (e.g., Shuvalov et al., 2002). A number of craters show ballistically emplaced ejecta, spreading outward over distances of less than one crater diameter. For both crater ejecta types, no clear diameter dependence is found. For a few of the larger craters, no ejecta are visible at the given image resolution, suggesting that they were formed before the upper part of the surrounding surface unit was emplaced or that the surface was degraded after the crater was emplaced. They appear embayed or draped by later layers of volcanic or sedimentary deposits. Based on the Viking-MDIM-2.1 image data, new crater sizefrequency measurements were performed and ages were determined in selected areas of the Utopia region, which cover polygonal terrain and surrounding units (Fig. 7a–c). Additionally, a mosaic of 28 high-resolution Viking frames (orbit 430B) with a pixel resolution of 15 m was prepared (Fig. 6D), and, to cover the full range of crater sizes (in this case from 50 km down to 10 m), small-scale information was retrieved from a MOC-NA image M08/03489 at a resolution of 3.03 m (Fig. 7e). The location of the Viking 430B frames and the MOC-NA image is shown in Fig. 7a. All crater size-frequency distributions of the selected units (Fig. 7a–c) converge in the smaller crater diameter size range and give an age of 3.4 Ga (Fig. 8). This plains age is observed in many measurements in other lowland regions (see above) and is also confirmed by the measurements of the crater size-frequency distribution in mosaicked image data at a pixel resolution of 15 m observed during Viking orbit 430B. A later resurfacing event is observed, which is also supported by the crater counts on

1156

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

120°0'0"E

30°0'0"W

120°0'0"E

30°0'0"W

120°0'0"E

30°0'0"W

30°0'0"N

30°0'0"N

+ 0.02 - 0.03Ga

3.77

+ 0.02 - 0.02Ga

ghost + pristine craters ghost craters pristine craters

Crater diameter D [km]

3.52 Cum.Crater Frequency Ncum [km-2]

Cum.Crater Frequency Ncum [km-2]

3.52

+ 0.02 - 0.03Ga

3.68

+ 0.02 - 0.03Ga

ghost + pristine craters ghost craters pristine craters

Crater diameter D [km]

Fig. 5. In Utopia Planitia (A) and in the Acidalia Planitia (D) giant polygonal terrains are shown (hatched) with respect to the map units of the northern lowlands (Tanaka et al., 2005). For these areas, crater counts of pristine and so-called ghost craters were performed. The position of the visible (pristine) craters are indicated in black, and the so-called ghost craters in red (for Utopia Planitia, B, for Acidalia Planitia, E). The resulting crater size-frequency distributions for the visible (pristine) crater distribution (squares) show a clear deficiency of large craters. The distribution of ghost craters (empty circles), all larger than 2 km, is recognized. The summed distribution of visible (pristine) craters and the buried crater population (ghost craters) is shown (filled circles). Both areas have been resurfaced about 3.52 Ga ago, but the detectability of ghost craters varies.

the MOC-NA image (Fig. 8). For the larger-crater diametersize range (larger than 3 km), a variety of shapes for the crater size-frequency distribution is observed for the different areas of Utopia Planitia, confirming earlier observations by McGill (1986). These variations correlate with the presence of giant polygons and

other surface structures, but mostly with the presence of either subdued (Fig. 7b) or embayed (Fig. 7c) large craters, indicating resurfacing processes have been acting in a different manner. The diversity or deviation from the expected crater-production function (Ivanov, 2001) for the larger-crater diameter-size range

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

1157

Fig. 6. In the central part of the Utopia Planitia a diversity of crater ejecta morphologies are found: (A) crater with fluidized double lobate ejecta blanket, (B) similarly sized crater with single ejecta blanket (C) rather ballistically emplaced than fluidized ejecta (D) pancake ejecta (E) double-ring graben suggesting buried crater (F) single-ring graben, and superposed by smaller craters. The areas of giant polygons are characterized by ring- or double-ringed graben features interpreted as buried impact structures.

(larger than 3 km) most likely depends on different target properties or variations in geological evolution. However, it is unclear why there is a lack of large craters in certain regions, yet in neighboring regions it appears that the large crater population of the lowland basement is exposed. In units (Fig. 7c) that show an apparent excess of larger craters in the measured crater size-frequency distribution (Fig. 8), the largest craters apparently formed on the surface of 3.75 Ga and survived embayment by younger volcanic or sedimentary deposits at about 3.4 Ga (Fig. 7b). Crater counts on highresolution images constrain the resurfacing at about 3.4 Ga and suggest that subsequently only minor resurfacing such as mantling affected craters at diameters smaller than 300 m. Irrespective of the cause of the diversity or obscuration of the measured crater size-frequency distribution in the larger-size range compared to the expected crater-production function, resurfacing occurred between 3.4 and 3.75 Ga in areas where polygonal terrain is observed. The earlier crater population may have been buried by a 1–2-km-thick layer as previously proposed (Buczkowski and Cooke, 2004). In other areas of Utopia Planitia, ring-graben structures and a lack of large craters are not observed. These areas appear to be at the outer edge of the Utopia basin, and this conforms with sedimentation concentrated in the Vastitas Borealis interior unit as might have occurred in a palaeoocean in the northern lowlands. Therefore, the interpretation that the polygons emerged through desiccation and differential compaction of sediment over buried topography is strongly supported, although not necessarily a unique interpretation. It also implies the slow (or frequently episodic) emplacement (sedimentation over roughly 350 Myrs) of sediments within these regions (isochron-cutting crater distribution, compare Fig. 1C), so that the formation of the giant polygons is unlikely to have formed through Rayleigh convection during the deposit emplacement as suggested by Lane and Christensen (2000). The elastic rebound of the crust due to the removal of water, a driving force suggested by Hiesinger and Head (2000), is possible and could aid the pattern formation after the disappearance of a possible ocean. Considering the analysis of the northern-lowland plains units, we observed a correlation between younger ages and topographic lows. These ages were found to cover an age range between 3.6 and 2.6 Ga,

which would imply a very slow water regression, if indeed this variation resulted from a Hesperian ocean.

5. Implications for the evolutionary history of the highland–lowland boundary, and the northern lowlands Summarizing the results of our investigation, we constrained that the highland–lowland dichotomy boundary developed during the pre-Noachian and was later modified. The Noachis Terra unit formed during a period between 4.04 and 3.96 Ga, whereas the slightly younger Nepenthes Mensae unit formed between 3.81 and 3.65 Ga. The Nepenthes Mensae unit, which marks the edge of the highland plateau, was resurfaced by inter-crater plains formation (sedimentary and volcanic) and erosion by surface water runoff (valley networks) during the Hesperian. Examining the dichotomy between Chryse and Isidis Planitiae, a morphologic difference of the regions eastward and westward of the impact basin Lyot is found. This separation is seen in crater counts, morphology, topography and gravity anomalies. However, it is unclear why these boundary parts developed differing topographic expressions. The original morphology of the dichotomy boundary has been partly obscured by volcanic, fluvial, and cratering processes, which is reflected in ages measured in this study. Fluvial activity, most dramatically displayed in the Chryse region, was less effective in obscuring the dichotomy boundary than volcanic accumulations (e.g., as at Alba Mons and Olympus Mons). Four volcanic provinces superpose the lowlands (Elysium rise and Apollinaris Mons), overlap the dichotomy (Tharsis rise), or cover adjacent highlands (Syrtis Major Planum). Therefore, information for the time frame for the existence of a Martian ocean may be indicated in the formation age of these volcanic flanks, and in the flank base morphology of the volcanoes along the highland–lowland boundary. Alba Mons and the Elysium rise show smooth slopes, characterized by superposing lava flows or similar volcanic morphologies, formed during the Late Hesperian. On the other hand, the Apollinaris Mons flank base forms a steep escarpment towards the lowlands. While the undisturbed flank formation of Alba Mons and the Elysium rise occurred only in the

1158

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

110°0'0"E

100°0'0"E

35°0'0"N

120°0'0"E

(a)

35°0'0"N (b) Orbit 430B (d)

30°0'0"N

M0803489 (e)

(c)

100°0'0"E

110°0'0"E

30°0'0"N

120°0'0"E Orbit 430B 10 km

M0803489 1 km (e)

(d)

Fig. 7. The central part of the Utopia Planitia (265–2401W and 30–371N) is shown here (Viking MDIM2.1) to illustrate the selected areas, which were used for crater counting (a–e). Additionally, the location of the Viking high resolution image mosaic (15 m/pxl taken during orbit 430B) and the location of the MOC (M0803489) image are plotted. The image data are shown as well. The subunits were selected according to their diverse crater morphology. The counts resulting crater size-frequency distributions are shown in Fig. 6.

0

200 Ma 3.40 Ga 3.75 Ga

−1 Elevation [km]

Cum. Crater Frequency N [km-2]

Utopia region formation resurfacing Chryse region formation resurfacing

−2 −3 −4 −5 0

1

2

3

4

Age [Ga] Fig. 9. The relation between age and elevation for the Utopia and the Chryse region.

15m VIKING MOC MDIM2 .1(all) MDIM2.1 (many visible large craters) MDIM2.1 (no visible large craters)

Crater Diameter D [km] Fig. 8. The crater size-frequency distributions are measured in units shown in Fig. 7. We counted on Viking MDIM2.1 and 15 m/pxl resolution Viking image data, as well as on a MOC image. The subunits b and c were selected according to the presence or deficiency of large craters (diameters around 15 km). The crater sizefrequency distributions of these selected areas confirm a diversity at the large-size range. They converge for craters smaller than about 2 km in diameter. Crater counts at higher resolution reveal one resurfacing event which is pronounced at about 300 m (compare Fig. 1A) followed by a more continuous resurfacing pattern as illustrated in Fig. 1C.

Late Hesperian, the lowland-facing cliff-like flank of Apollinaris Mons formed already before the Early Hesperian. However, Apollinaris Mons’ last volcanic episode is manifested in a southward erupted lava fan at about 3.71 Ga (Werner, 2009) at the transition between Noachian and Hesperian. This difference in morphology and timing between the main edifice of Apollinaris Mons and Alba Mons may imply that early stages of Apollinaris Mons formed in the presence of a standing body of water, while Alba Mons and the youngest fan of Apollinaris Mons did not. A general water cycle, including precipitation and surface runoff (valley networks, Fassett and Head, 2008), may have persisted until then (3.71 Ga ago). The formation of the most extensive surfaces in the lowlands, as defined by densities of craters generally 43 km in diameter, ended between 3.7 and 3.4 Ga ago. All ages measured in the closest vicinity of the steep dichotomy escarpment are around 3.7 Ga or older. By trend, neighboring northern lowland units show gradual changes to younger formation and resurfacing ages continuing down-slope toward topographic lows (Fig. 9). These ages cover an age range between 3.6 and 2.6 Ga. This could imply a gradual water regression (or other resurfacing processes).

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

1159

Fig. A1. Inventory of the clippings of the Viking MDIM2.1 mosaics on which the measurements have been performed. They represent exactly the area which was counted. The resulting crater size-frequency distributions for each patch are shown together with the best-fit model ages (thin line), derived by fitting crater production function by Ivanov (2001) and applying the cratering chronology model by Hartmann and Neukum (2001). The fit range is given by diameter values for which the curve is plotted (thick line). Statistics for these measurements are given in Table 1. Resurfacing events are plotted corrected already, see Section 2 for details.

1160

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

Fig. A1. (continued)

The youngest units in the northern lowlands are considered to be lavas or polar ice (Tanaka et al., 2005), arguing against the ocean theory during the Amazonian Period (younger than about 3.1 Ga).

If a (transient) Amazonian ocean existed, it did not result in any significant deposition or resurfacing, which we consider unlikely.

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

1161

Fig. A1. (continued)

Detailed studies of the giant polygonal terrain observed in the Utopia and Acidalia Planitiae indicate that the distribution of visible craters and the ring-graben structures (ghost craters) yield the same

basement ages (up to 3.77 Ga) as defined by entirely exposed craters closer to the highland–lowland dichotomy boundary (e.g., Utopia Planitia 1 and Chryse Planitia 1 units), where a much thinner deposit

1162

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

Fig. A1. (continued)

post-dating the boundary retreat is expected. The lowland Vastitas Borealis interior unit with an age of about 3.4 Ga mostly appears to constitute deposits at least 1–2 km thick. Results found in relation to the polygonal terrain temporally and morphologically conform with

the existence of a proposed depocenter (in a water or mud ocean) shortly after the crustal formation of the northern lowlands. These ages are consistent with the last time period expected for the existence of any kind of Martian ocean. A general water

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

1163

Fig. A1. (continued)

cycle, including precipitation and surface runoff (valley networks, Fassett and Head, 2008), may have persisted only until about 3.7 Ga ago.

S. C. Werner is supported by the Norwegian Research Council through a Centre of Excellence Grant to PGP.

Appendix A. Image data Acknowledgments This paper benefited greatly from thorough reviews by Brian Hynek and an anonymous reviewer. We appreciated comments by Rosalyn K. Hayward and Corey Fortezzo during internal USGS reviews. We are thanking Ursula Wolf (Freie Universit¨at Berlin) for her help counting on Viking image data. An earlier stage of this study was funded by the Deutsche Forschungsgemeinschaft (DFG, through G. Neukum); also funding was provided by the NASA Planetary Geology and Geophysics Program for K. L. Tanaka and J. A. Skinner.

Inventory of the clippings of the Viking MDIM2.1 mosaics and the resulting crater size-frequency distributions for each patch are shown in Fig. A1. References Andrews-Hanna, J.C., Zuber, M.T., Banerdt, W.B., 2008. The Borealis basin and the origin of the martian crustal dichotomy. Nature 453, 1212–1215. doi:10.1038/ nature07011. Archinal, B.A., Kirk, R.L., Howington-Kraus, E., Rosiek, M.R., Soderblom, L.A., Lee, E.M., 2003. Mapping Mars at global to human scales. In: IAU Special Session, 1.

1164

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

Arvidson, R.E., et al., 1979. Standard techniques for presentation and analysis of crater size-frequency data. Icarus 37, 467–474. Baker, V.R., 1982. The Channels of Mars. University of Texas Press, 204 pp. Baker, V.R., Strom, R.G., Gulick, V.C., Kargel, J.S., Komatsu, G., 1991. Ancient oceans, ice sheets and the hydrological cycle on Mars. Nature 352, 589–594. doi:10.1038/352589a0. Barlow, N.G., 1988. Crater size-frequency distributions and a revised Martian relative chronology. Icarus 75, 285–305. Barlow, N.G., 2001. Status of the ‘‘Catalog of Large Martian Impact Craters’’ revision. Bulletin of the American Astronomical Society 33, 1105. Basilevsky, A.T., et al., 2009. Episodes of floods in Mangala Valles, Mars, from the analysis of HRSC, MOC and THEMIS images. Planetary and Space Science 57, 917–943. doi:10.1016/j.pss.2008.07.023. Buczkowski, D.L., Cooke, M.L., 2004. Formation of double-ring circular grabens due to volumetric compaction over buried impact craters: implications for thickness and nature of cover material in Utopia Planitia, Mars. Journal of Geophysical Research, 2006. Buczkowski, D.L., McGill, G.E., 2002. Topography within circular grabens: implications for polygon origin, Utopia Planitia, Mars. Geophysical Research Letters 29, 59–61. Carr, M.H., 1979. Formation of Martian flood features by release of water from confined aquifers. Journal of Geophysical Research 84 (13), 2995–3007. Carr, M.H., 1981. The Surface of Mars. Yale University Press. Carr, M.H., Schaber, G.G., 1977. Martian permafrost features. Journal of Geophysical Research 82 (11), 4039–4054. Carr, M.H., et al., 1976. Preliminary results from the Viking orbiter imaging experiment. Science 193, 766–776. Carr, M.H., Crumpler, L.S., Cutts, J.A., Greeley, R., Guest, J.E., Masursky, H., 1977. Martian impact craters and emplacement of ejecta by surface flow. Journal of Geophysical Research 82 (11), 4055–4065. Chapman, M.G., et al., 2010a. Noachian–Hesperian geologic history of the Echus Chasma and Kasei Valles system on Mars: new data and interpretations. Earth and Planetary Science Letters 294, 256–271. doi:10.1016/j.epsl.2009.11.032. Chapman, M.G., Neukum, G., Dumke, A., Michael, G., van Gasselt, S., Kneissl, T., Zuschneid, W., Hauber, E., Mangold, N., 2010b. Amazonian geologic history of the Echus Chasma and Kasei Valles system on Mars: new data and interpretations. Earth and Planetary Science Letters 294, 238–255. doi:10.1016/ j.epsl.2009.11.034. ˜ a, M.H., Wasilewski, P.J., Kletetschka, G., Ness, N.F., Re me, Connerney, J.E.P., Acun H., Lin, R.P., Mitchell, D.L., 2001. The global magnetic field of Mars and implications for crustal evolution. Geophysical Research Letters 28, 4015–4018. doi:10.1029/2001GL013619. Cruikshank, D.P., Hartmann, W.K., Wood, C.A., 1973. Moon: ‘ghost’ craters formed during mare filling. Moon 7, 440–452. doi:10.1007/BF00564645. Fassett, C.I., Head, J.W., 2008. The timing of martian valley network activity: constraints from buffered crater counting. Icarus 195, 61–89. doi:10.1016/ j.icarus.2007.12.009. Frey, H., Schultz, R.A., 1988. Large impact basins and the mega-impact origin for the crustal dichotomy on Mars. Geophysical Research Letters 15, 229–232. Frey, H., Semeniuk, A.M., Semeniuk, J.A., Tokarcik, S., 1988. A widespread common age resurfacing event in the highland-lowland transition zone in eastern Mars. In: Lunar and Planetary Science Conference Proceedings, Lunar and Planetary Science Conference Proceedings, vol. 12, pp. 679–699. Frey, H.V., Roark, J.H., Shockey, K.M., Frey, E.L., Sakimoto, S.E.H., 2002. Ancient lowlands on Mars. Geophysical Research Letters 29, 22–31. Garvin, J.B., Sakimoto, S.E.H., Frawley, J.J., 2003. Craters on Mars: global geometric properties from gridded MOLA topography. In: Sixth International Conference on Mars, p. 3277. Greeley, R., Guest, J.E., 1987. Geological Map of the Eastern Equatorial Region of Mars (1:15,000,000), USGS. Hansen, V.L., 2000. Geologic mapping of tectonic planets. Earth and Planetary Science Letters 176, 527–542. doi:10.1016/S0012-821X(00)00017-0. Hartmann, W.K., 1966. Early lunar cratering. Icarus 5, 406–418. Hartmann, W.K., 1973. Martian Cratering 4. Mariner 9 initial analysis of cratering chronology. Journal of Geophysical Research 78, 4096–4116. Hartmann, W.K., 1977. Relative crater production rates on planets. Icarus 31, 260–276. Hartmann, W.K., 1978. Martian cratering V: toward an empirical Martian chronology. Geophysical Research Letters 5, 450–452. Hartmann, W.K., Neukum, G., 2001. Cratering chronology and the evolution of Mars. Space Science Reviews 96, 165–194. Hiesinger, H., Head, J.W., 2000. Characteristics and origin of polygonal terrain in southern Utopia Planitia, Mars: results from Mars Orbiter Laser Altimeter and Mars Orbiter Camera data. Journal of Geophysical Research 105, 11999–12022. Hiesinger, H., Head, J.W., 2004. The Syrtis Major volcanic province, Mars: synthesis from Mars Global Surveyor data. Journal of Geophysical Research (Planets) 109, 1004. doi:10.1029/2003JE002143. Ivanov, B.A., 2001. Mars/Moon cratering rate ratio estimates. Space Science Reviews 96, 87–104. Janle, P., 1983. Bouguer gravity profiles across the highland–lowland escarpment on Mars. Moon and Planets 28, 55–67. Kreslavsky, M.A., Head, J.W., 2002. Mars: nature and evolution of young latitudedependent water-ice-rich mantle. Geophysical Research Letters 29 (15), 150000–150001. doi:10.1029/2002GL015392. Lane, M.D., Christensen, P.R., 2000. Convection in a catastrophic flood deposit as the mechanism for the giant polygons on Mars. Journal of Geophysical Research 105, 17617–17628.

Lanz, J., 2004. Geometrische, morphologische und stratigraphische Untersuchun¨ ¨ gen ausgewahlter Outflow Channel (Ausflusstaler) der Circum-Chryse-region, Mars, mit Methoden der Fernerkundung. Ph.D.Thesis. Levenberg, K., 1944. A method for the solution of certain problems in least squares. Quarterly of Applied Mathematics 2, 164–168. Lucchitta, B.K., Ferguson, H.M., Summers, C., 1986. Sedimentary deposits in the northern lowland plains, Mars. Journal of Geophysical Research 91, 166. Marchenko, A.G., Basilevsky, A.T., Hoffmann, H., Hauber, E., Cook, A.C., Neukum, G., 1998. Geology of the common mouth of the Ares and Tiu Valles, Mars. Astronomicheskii Vestnik 32, 425. Marinova, M.M., Aharonson, O., Asphaug, E., 2008. Mega-impact formation of the Mars hemispheric dichotomy. Nature 453, 1216–1219. doi:10.1038/ nature07070. Marquardt, D., 1963. An algorithm for least-squares estimation of nonlinear parameters. SIAM Journal of Applied Mathematics 11, 431–441. Masursky, H., Crabill, N.L., 1976. The Viking landing sites—selection and certification. Science 193, 809–812. McGill, G.E., 1986. The giant polygons of Utopia, Northern Martian Plains. Geophysical Research Letters 13, 705–708. McGill, G.E., 1989. Buried topography of Utopia, Mars—persistence of a giant impact depression. Journal of Geophysical Research 94 (13), 2753–2759. McGill, G.E., Dimitriou, A.M., 1990. Origin of the Martian global dichotomy by crustal thinning in the late Noachian or early Hesperian. Journal of Geophysical Research 95 (14), 12595–12605. McGill, G.E., Hills, L.S., 1992. Origin of giant Martian polygons. Journal of Geophysical Research 97 (16), 2633–2647. Michael, G.G., Neukum, G., 2010. Planetary surface dating from crater sizefrequency distribution measurements: partial resurfacing events and statistical age uncertainty. Earth and Planetary Science Letters 294, 223–229. doi:10.1016/j.epsl.2009.12.041. Mohit, P.S., Phillips, R.J., 2007. Viscous relaxation on early Mars: a study of ancient impact basins. Geophysical Research Letters 34, 21204. doi:10.1029/ 2007GL031252. Morris, E.C., Underwood, J.R., 1978. Polygonal fractures of the Martian plains. Technical Report. Mouginis-Mark, P., 1979. Martian fluidized crater morphology—variations with crater size, latitude, altitude, and target material. Journal of Geophysical Research 84 (13), 8011–8022. Muller, P.M., Sjogren, W.L., 1968. Mascons: lunar mass concentrations. Science 161, 680–684. Murray, J.B., et al., 2005. Evidence from the Mars express high resolution stereo camera for a frozen sea close to Mars’ equator. Nature 434, 352–356. Mustard, J.F., Cooper, C.D., Rifkin, M.K., 2001. Evidence for recent climate change on Mars from the identification of youthful near-surface ground ice. Nature 412, 411–414. Mutch, T.A., Arvidson, R.E., Head, J.W., Jones, K.L., Saunders, R.S., 1976. The Geology of Mars. Princeton University Press, Princeton, N.J., 409 pp. Neukum, G., Hiller, K., 1981. Martian ages. Journal of Geophysical Research 86 (15), 3097–3121. Neukum, G., Ivanov, B.A., 1994. Crater size distributions and impact probabilities on Earth from Lunar, terrestrial-planet, and Asteroid cratering data. In: Hazards Due to Comets and Asteroids, p. 359. Neukum, G., Wise, D.U., 1976. Mars—a standard crater curve and possible new time scale. Science 194, 1381–1387. ¨ Neukum, G., Konig, B., Arkani-Hamed, J., 1975. A study of lunar impact crater sizedistributions. Moon 12, 201–229. 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. Journal of Geophysical Research 109, 8002. Nimmo, F., Hart, S.D., Korycansky, D.G., Agnor, C.B., 2008. Implications of an impact origin for the martian hemispheric dichotomy. Nature 453, 1220–1223. doi:10.1038/nature07025. Parker, T.J., Gorsline, D.S., Saunders, R.S., Pieri, D.C., Schneeberger, D.M., 1993. Coastal geomorphology of the Martian northern plains. Journal of Geophysical Research 98, 11061. doi:10.1029/93JE00618. Pechmann, J.C., 1980. The origin of polygonal troughs on the northern plains of Mars. Icarus 42, 185–210. Pike, R.J., 1980. Control of crater morphology by gravity and target type—Mars, Earth, Moon. In: Lunar and Planetary Science Conference, vol. 1, pp. 2159– 2189. Richardson, J.E., 2009. Cratering saturation and equilibrium: a new model looks at an old problem. Icarus 204, 697–715. doi:10.1016/j.icarus.2009.07.029. Rossbacher, L.A., Judson, S., 1981. Ground ice on Mars—inventory, distribution, and resulting landforms. Icarus 45, 39–59. Rotto, S., Tanaka, K.L., 1995. Geologic/geomorphologic map of the Chryse Planitia region of Mars (1:5,000,000) I-2441, USGS. Schultz, P.H., Schultz, R.A., Rogers, J., 1982. The structure and evolution of ancient impact basins on Mars. Journal of Geophysical Research 87, 9803–9820. doi:10.1029/JB087iB12p09803. Scott, D.H., Carr, M.H., 1978. The new geologic map of Mars (1:25 Million Scale). Technical Report. Scott, D.H., Tanaka, K.L., 1986. Geological map of the western equatorial region of Mars (1:15,000,000), USGS. Shoemaker, E., 1965. Preliminary analysis of the fine structure of the lunar surface in Mare Cognitum. In: Ranger VII, Part 2, Technical Report, Jet Propulsion Laboratory, California, pp. 75–134.

S.C. Werner et al. / Planetary and Space Science 59 (2011) 1143–1165

Shuvalov, V., Dypvik, H., Tsikalas, F., 2002. Numerical simulations of the Mjølnir marine impact crater. Journal of Geophysical Research (Planets) 107, 5047. doi:10.1029/2001JE001698. Skinner, J.A., Tanaka, K.L., 2007. Evidence for and implications of sedimentary diapirism and mud volcanism in the southern Utopia highland lowland boundary plain, Mars. Icarus 186, 41–59. doi:10.1016/j.icarus.2006.08.013. Sleep, N.H., 1994. Martian plate tectonics. Journal of Geophysical Research 99 (18), 5639. Smith, D.E., et al., 1998. Topography of the northern hemisphere of Mars from the Mars Orbiter Laser Altimeter. Science 279, 1686. Smith, D.E., Sjogren, W.L., Tyler, G.L., Balmino, G., Lemoine, F.G., Konopliv, A.S., 1999. The gravity field of Mars: results from Mars Global Surveyor. Science 286, 94–97. doi:10.1126/science.286.5437.94. Tanaka, K.L., 1986. The stratigraphy of Mars. Journal of Geophysical Research 91, 139–158. Tanaka, K.L., Chapman, M.G., 1992. Kasei Valles, Mars—interpretation of canyon materials and flood sources. In: Ryder, G., Sharpton, V.L. (Eds.), Lunar and Planetary Science Conference Proceedings, Lunar and Planetary Science Conference Proceedings, vol. 22; 1992, pp. 73–83. Tanaka, K.L., Scott, D.H., 1987. Geological map of the polar regions of Mars (1:15,000,000), USGS. Tanaka, K.L., Scott, D.H., Greeley, R., 1992. Global stratigraphy. In: Mars, pp. 345–382. Tanaka, K.L., Skinner, J.A., Hare, T.M., Joyal, T., Wenker, A., 2003. Resurfacing history of the northern plains of Mars based on geologic mapping of Mars Global Surveyor data. Journal of Geophysical Research 108, 24. Tanaka, K.L., Skinner, J.A., Hare, T.M., 2005. Geological map of the northern plains of Mars (1:15,000,000), USGS, /http://pubs.usgs.gov/sim/2005/2888/S. Tanaka, K.L., Fortezzo, C.M., Hayward, R., Rodriguez, J.A.P., Skinner, J.J.A. History of plains resurfacing in the Scandia region of Mars. Planetary and Space Science, this issue, doi:10.1016/j.pss.2010.11.004. Thomson, B.J., Head, J.W., 2001. Utopia Basin, Mars: characterization of topography and morphology and assessment of the origin and evolution of

1165

basin internal structure. Journal of Geophysical Research (Planets) 106, 23209–23230. doi:10.1126/science.286.5437.94. Watters, T.R., 2003. Lithospheric fexure and the origin of the dichotomy boundary on Mars. Geology 31, 271–274. Werner, S.C., 2005. Major aspects of the chronostratigraphy and geologic evolutionary history of Mars, Ph.D. Thesis, Freie Universitaet Berlin. Werner, S.C., 2008. The early Martian evolution constraints from basin formation ages. Icarus 195, 45–60. doi:10.1016/j.icarus.2007.12.008. Werner, S.C., 2009. The global Martian volcanic evolutionary history. Icarus 201, 44–68. doi:10.1016/j.icarus.2008.12.019. Werner, S.C., van Gasselt, S., Neukum, G., 2003. Continual geological activity in Athabasca Valles, Mars. Journal of Geophysical Research 108, 22–31. Werner, S.C., Ivanov, B.A., Neukum, G., 2009. Theoretical analysis of secondary cratering on Mars and an image-based study on the Cerberus Plains. Icarus 200, 406–417. doi:10.1016/j.icarus.2008.10.011. Wilhelms, D.E., Squyres, S.W., 1984. The martian hemispheric dichotomy may be due to a giant impact. Nature 309, 138–140. Wise, D.U., Golombek, M.P., McGill, G.E., 1979. Tectonic evolution of Mars. Journal of Geophysical Research 84 (13), 7934–7939. Wood, C.A., Head, J.W., Cintala, M.J., 1978. Interior morphology of fresh Martian craters–the effects of target characteristics. In: Lunar and Planetary Science Conference, pp. 3691–3709. Woronow, A., 1977. Crater saturation and equilibrium—a Monte Carlo simulation. Journal of Geophysical Research 82, 2447–2456. doi:10.1029/JB082i017 p02447. Yuan, D., Sjogren, W.L., Konopliv, A.S., Kucinskas, A.B., 2001. Gravity field of Mars: a 75th degree and order model. Journal of Geophysical Research 106 (15), 23377–23402. Zhong, S., Zuber, M.T., 2001. Degree-1 mantle convection and the crustal dichotomy on Mars. Earth and Planetary Science Letters 189, 75–84. Zuber, M.T., 2001. The crust and mantle of Mars. Nature 42, 220–227. Zuber, M.T., et al., 2000. Internal structure and early thermal evolution of Mars from Mars Global Surveyor topography and gravity. Science 287, 1788–1793.