Icarus 209 (2010) 390–404

Contents lists available at ScienceDirect

Icarus

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

Concentric crater fill in the northern mid-latitudes of Mars: Formationprocesses and relationships to similar landforms of glacial origin

Joseph Levy a,*, James W. Head a, David R. Marchant b

a Department of Geological Sciences, Brown University, Providence, RI 02912, United Statesb Department of Earth Science, Boston University, Boston, MA 02215, United States

a r t i c l e i n f o

Article history:Received 6 November 2009Revised 24 February 2010Accepted 31 March 2010Available online 4 April 2010

Keywords:MarsMars, ClimateMars, Surface

0019-1035/$ - see front matter � 2010 Elsevier Inc. Adoi:10.1016/j.icarus.2010.03.036

* Corresponding author. Present address: DepartmeUniversity, Portland, OR 97201, United States. Fax: +1

E-mail addresses: jlevy@pdx.edu, joseph_levy@bro

a b s t r a c t

Hypotheses accounting for the formation of concentric crater fill (CCF) on Mars range from ice-free pro-cesses (e.g., aeolian fill), to ice-assisted talus creep, to debris-covered glaciers. Based on analysis of newCTX and HiRISE data, we find that concentric crater fill (CCF) is a significant component of Amazonian-aged glacial landsystems on Mars. We present mapping results documenting the nature and extent ofCCF along the martian dichotomy boundary over �30 to 90�E latitude and 20–80�N longitude. On thebasis of morphological analysis we classify CCF landforms into ‘‘classic” CCF and ‘‘low-definition” CCF.Classic CCF is most typical in the middle latitudes of the analysis area (�30–50�N), while a range of deg-radation processes results in the presence of low-definition CCF landforms at higher and lower latitudes.We evaluate formation mechanisms for CCF on the basis of morphological and topographic analyses, andinterpret the landforms to be relict debris-covered glaciers, rather than ice-mobilized talus or aeolianunits. We examine filled crater depth–diameter ratios and conclude that in many locations, hundredsof meters of ice may still be present under desiccated surficial debris. This conclusion is consistent withthe abundance of ‘‘ring-mold craters” on CCF surfaces that suggest the presence of near-surface ice. Anal-ysis of breached craters and distal glacial deposits suggests that in some locations, CCF-related ice wasonce several hundred meters higher than its current level, and has sublimated significantly during themost recent Amazonian. Crater counts on ejecta blankets of filled and unfilled craters suggests thatCCF formed most recently between �60 and 300 Ma, consistent with the formation ages of other martiandebris-covered glacial landforms such as lineated valley fill (LVF) and lobate debris aprons (LDA). Mor-phological analysis of CCF in the vicinity of LVF and LDA suggests that CCF is a part of an integratedLVF/LDA/CCF glacial landsystem. Instances of morphological continuity between CCF, LVF, and LDA areabundant. The presence of formerly more abundant CCF ice, coupled with the integration of CCF intoLVF and LDA, suggests the possibility that CCF represents one component of the significant Amazonianmid-latitude glaciation(s) on Mars.

� 2010 Elsevier Inc. All rights reserved.

1. Introduction

Concentrically-lineated, crater-filling deposits (concentric cra-ter fill, or CCF, Fig. 1) have been described at middle-high latitudeson Mars since the Viking era, and have long been associated with asuite of features typical of fretted terrain (e.g., lobate debris aprons,LDA; and lineated valley fill, LVF) (Sharp, 1973; Squyres, 1978,1979; Malin and Edgett, 2001). Hypotheses explaining the forma-tion of these unusual landforms range from entirely water-freeprocesses to those requiring the deposition and maintenance ofhundreds of meters of nearly pure ice. For example, Zimbelmanet al. (1989) accounts for CCF by the emplacement and removal

ll rights reserved.

nt of Geology, Portland State401 863 3978.

wn.edu (J. Levy).

of aeolian layers. Invoking limited quantities of water, Sharp(1973) explains fretted terrain as debris derived from bedrockrecession along scarps due to sublimation of ground ice or theemergence of groundwater. Squyres (1978) suggests that lineatedvalley fill results from the mobilization and flow of talus by diffu-sionally-emplaced pore ice; while Squyres (1979) and Squyres andCarr (1986) extend this rock-glacier-like mechanism (e.g., Haeberliet al., 2006) to lobate debris aprons and concentric crater fill. Agreater role for internally-deforming ice in LVF processes was pro-posed by Lucchitta (1984), drawing explicit analogs between lin-eated valley fill and polythermal Antarctic glaciers. In a study oflarge craters in the Vastitas Borealis Formation, Kreslavsky andHead (2006) suggest a range of crater-filling mechanisms to ac-count for the filling of high-latitude craters on Mars, including deb-ris-covered glacial processes (see also Garvin et al., 2006); similarto the debris-covered glacier mechanism invoked by Levy et al.

Fig. 1. ‘‘Classic” and ‘‘low-definition” concentric crater fill (CCF) on the martian dichotomy boundary. (a) Classic CCF. Portion of P15_007028_2164. (b) Low-definition CCF.Portion of P16_007437_2090. Illumination from the left and north to image top in both images.

J. Levy et al. / Icarus 209 (2010) 390–404 391

(2009b) to account for CCF morphology in Utopia Planitia. The po-tential for the current presence of significant quantities of ice inCCF and LVF was suggested by the documentation of abundantexamples of integrated, glacier-like flow features in LVF and LDAacross the martian dichotomy boundary (e.g., Head et al.,2006a,b, 2010; Levy et al., 2007; Morgan et al., 2009), the presenceof ring-mold craters in LVF/LDA deposits (Kress and Head, 2008),and the discovery of significant volumes of nearly pure ice withinlobate debris aprons and lineated valley fill described by Holt et al.(2008) and Plaut et al. (2009) using SHARAD radar data.

Advances in understanding of CCF and related landforms havefollowed from the improving spatial resolution and coverage ofspacecraft data, including image and topographic datasets. Here,we use Context Camera (CTX) (Edgett et al., 2008) and High Reso-lution Imaging Science Experiment (HiRISE) (McEwen et al., 2009)image data, coupled with HRSC high-resolution stereo DEMs(Oberst et al., 2005; Jaumann et al., 2007) and supplemented byMOLA topographic data, to address the following outstandingquestions related to CCF: (1) What are the morphological charac-teristics of CCF at high resolution, and what does the morphologysuggest about CCF composition? (2) Are there morphological sub-types of CCF, and if so, what is their geographical distribution? (3)

If ice is involved in CCF processes, is any still present? (4) What isthe relationship between CCF and other crater-filling materials(e.g., dunes) on Mars? (5) What is the relationship between CCFand LVF/LDA? (6) What do relationships between LVF/LDA/CCFprocesses suggest about the origin and development of martianglacial landforms? We focus on CCF found along the martiandichotomy boundary (particularly within �30 to 90�E and 20–80�N), in the vicinity of a range of landforms that have been inter-preted as features of glacial origin.

2. Mapping mid-latitude crater-filling units

Concentric crater fill (CCF) has been classically described in re-gions containing abundant lineated valley fill (LVF) and/or lobatedebris aprons (LDA) as a crater-interior unit characterized by mul-tiple rings or lineations concentric with the crater rim (Squyres,1979) (Figs. 1a and 2). At Mars Orbiter Camera (MOC) resolution(1.5–20 m/pixel), the surface texture associated with LVF, LDA,and CCF lineations appears mounded and furrowed, a texture de-scribed as ‘‘brain or corn-like pit and mound textures” (Malinand Edgett, 2001), ‘‘pit-and-butte” texture (Mangold, 2003) or

Fig. 2. Comparisons between HiRISE and CTX image data detailing CCF structure at fine and coarse scales. (a) and (c) show the same patch of CCF surface, showing typical‘‘brain terrain” morphology. HiRISE reveals details of surface structure not visible in CTX images. OC-BT indicates open-cell ‘‘brain terrain” (arcuate and cuspate cells that aredelimited by a convex-up boundary ridge surrounding a flat-floored depression) and CC-BT indicates closed cell ‘‘brain terrain” (arcuate or domed mound features that arecommonly �10–20 m wide, and that may have surface grooves or furrows located near the centerline of the cell’s long axis). LDM Annulus indicates exposures of latitude-dependent mantle material ringing the interior of the crater. (b) and (d) show the same CCF-containing crater (areas shown in (a) and (c) indicated by white box). CTX clearlyresolves large-scale lineations in CCF caused by topographic undulation and by orientation of ‘‘brain terrain” cells; HiRISE viewing of the same features show similar structure,but is noisier due to the resolution of finer-scale features. All images excerpted from CTX image P03_002175_2211 and HiRISE image PSP_002175_2210. Illumination from theleft. North to image top.

392 J. Levy et al. / Icarus 209 (2010) 390–404

‘‘knobs—brain coral” (Williams, 2006). Detailed analysis of HiRISEimages has led to a description of CCF surface texture as ‘‘brain cor-al terrain” (Dobrea et al., 2007) or ‘‘brain terrain” (Levy et al.,2009b). For succinctness and consistency, we will refer to this sur-face texture throughout as ‘‘brain terrain.” HiRISE images revealtextural differences between ‘‘brain terrain” mounds or ‘‘cells”—particularly ‘‘closed cells” (arcuate or domed mound features thatare commonly �10–20 m wide, and that may have surface groovesor furrows located near the centerline of the cell’s long axis) and‘‘open cells” (arcuate and cuspate cells that are delimited by a con-vex-up boundary ridge surrounding a flat-floored depression)(Fig. 2) (Levy et al., 2009b).

Mapping of crater-filling units at martian mid-latitudes can beaccomplished using both HiRISE and CTX image data in concert(Fig. 3). Full-resolution CTX data are sufficient to resolve broad-scale ridges and troughs associated with CCF (Fig. 2), while detailedanalysis of relationships between ‘‘brain terrain” surface textures isfacilitated by observation at HiRISE resolution. Full-resolution CTXimages spanning Mars Reconnaissance Orbiter orbits 1331–7927from the study region (�30 to 90�E, 20–80�N) were inspected forthe presence of crater-filling units. Crater-fill units from the 1000images surveyed were mapped on a sinusoidal equal-area globalmap (Fig. 3).

For mapping morphological subtypes, we define ‘‘classic” CCF asa crater-interior unit showing lineations concentric with at leastsome portion of the crater rim and displaying ‘‘brain terrain” sur-face texture (Fig. 1a). It is the orientation of the long axis of ‘‘brain

terrain” cells that provides the fine structure of CCF lineations(Fig. 2). Topographic ridges and troughs �100 m wide define thebroader concentric lineations associated with CCF and can beclearly resolved in CTX image data (Fig. 2) (Levy et al., 2009b).The distribution of classic CCF along the martian dichotomyboundary is shown in Fig. 3. In our survey region, classic CCF iswidely distributed between �27�N and �50�N, with scatteredexamples at higher and lower latitudes (Fig. 3).

In some craters, CCF lacks clearly defined concentric lineationsproduced by topography or lacks clearly-defined ‘‘brain terrain”cells (Fig. 1b). We refer to this CCF as ‘‘low-definition” CCF. Low-definition CCF may show widened fractures or linear depressionsoriented both concentric-with and radial-to the crater rim(Fig. 1b) or may have a muted surface texture, showing only subtlestippling or shallow undulations in broad-scale topography. Thelatter surface texture is characteristic of higher-latitude low-defi-nition CCF, where continuous latitude-dependent mantle surfacesare thicker and more widespread (Head et al., 2003; Levy et al.,2009b). The distribution of low-definition CCF along the dichotomyboundary is mapped in Fig. 3.

3. CCF morphological relationships

Having briefly outlined the morphological characteristics ofindividual CCF deposits, the next step in assessing the origin,development, and modification history of CCF is to document therange of stratigraphic and morphological relationships observed

Fig. 3. (Above) Map showing the distribution of CCF landforms along the martian dichotomy boundary. CCF deposits larger than �3 km in diameter are mapped. Six hundredand fourteen classic CCF deposits are mapped, as well as 262 low-definition CCF deposits. White line indicates region of most concentrated ‘‘classic” CCF detection. Surveyedimage data include all CTX images from orbit groups P01–P17 located in the mapping region spanning �30 to 90�E and 20–80�N. Base map is MOLA topography. (Below)Histogram comparing the distribution of CCF occurrences (dark bars) and surveyed CTX images (light bars) by 5� latitude bins. CCF distribution appears to be stronglylatitude-dependent, while CTX image distribution is not. CCF occurrences can exceed the number of CTX images in a given latitude bin if multiple examples of CCF are presentin a given CTX image.

J. Levy et al. / Icarus 209 (2010) 390–404 393

between CCF and other martian landforms (e.g., Squyres, 1979;Squyres and Carr, 1986; Levy et al., 2009b). Here, we consider rela-tionships observed between both ice-related landforms (e.g., LVF,LDA) (Squyres, 1979; Lucchitta, 1984; Squyres and Carr, 1986;Head et al., 2006a,b, 2008, 2010; Holt et al., 2008; Levy et al.,2009b; Plaut et al., 2009) and non-ice-related landforms (such astalus, craters, and dunes).

3.1. CCF and host craters

Because CCF is a crater-interior unit, relationships between CCFand its host craters can be used to evaluate CCF-forming processes.Analysis of image and topography data indicates that CCF does notmerely line or coat the interior of craters; rather it consists of a flatand high (mesa-like) or convex-up surface that fills host craters

(Fig. 4). Fill thicknesses can be estimated by comparing predictedcrater depths from Garvin et al. (2002) with observed depths fromMOLA or HRSC topographic profiles (e.g., Levy et al., 2009b). Sixmeasured fill depths vary from �600 m to �1700 m, and com-monly constitute �75% of expected crater depth (Fig. 4).

Contacts between CCF material and crater walls have variablemorphology. In some instances, the upper surface of CCF materialsits well above the crater wall and is topographically separatedfrom the crater wall by a bounding scarp up to �100 m high(Fig. 4). In other cases, CCF surfaces grade from sloping crater wallsto the CCF surface—this latter relationship is most typical of cratersin which abundant latitude-dependent mantle material rings thecrater interior (Fig. 2) (Head et al., 2003; Levy et al., 2009b).

In some cases, crater-fill material appears to have cross-cut ortransgressed over crater rims (Fig. 5). For example, one crater in

Fig. 4. (a) Classic concentric crater fill that is clearly separated from the interior crater wall talus slopes. White line indicates track of topographic profile. Portion ofP14_006570_2241. Illumination from left. North to image top. (b) HRSC high-resolution DTM topographic profile across CCF-containing crater. Inset shows calculated depthexpected given crater diameter from Garvin et al. (2002) and an idealized bowl-shaped profile.

394 J. Levy et al. / Icarus 209 (2010) 390–404

Deuteronilus Mensae (Fig. 5) shows low-definition CCF within acrater that is connected to degraded ‘‘brain terrain” surfaces out-side the crater by a series of linear, positive-relief ridges. Similarridges are present to the east and to the west of the crater. Ridgesto the east of the crater extend from a nearby valley floor, upslopeand onto the outer walls of the crater rim. HRSC topographic pro-files of the deposit suggest that the low-definition CCF material is�300 m lower within the crater and outside the crater than it is on

the rim of the crater (Fig. 5). The ridges are morphologically similarto drop moraines typical of terrestrial glaciers (Marchant et al.,1993; Benn et al., 2003), as well as martian glaciers (Head and Mar-chant, 2003; Shean et al., 2005) that form as debris is shed at theglacier margins. If this interpretation is correct, it indicates thatpreviously crater-filling material was at least 300 m higher in thepast—sufficient to permit flow over the crater rim and onto the val-ley floor outside of the crater.

Fig. 5. (a) Low-definition CCF present in a Deuteronilus Mensae crater. ‘‘Brain terrain” surface textures extend out of the crater between several arcuate ridges interpreted tobe moraines. Box shows location of part c. Portion of P16_007385_2149. Illumination from left. North to image top. (b) HRSC high-resolution DTM profile across crater andnearby plains. (c) ‘‘Brain terrain” knobs present within the distal moraine-like features.

J. Levy et al. / Icarus 209 (2010) 390–404 395

3.2. CCF and dunes

Zimbelman et al. (1989) hypothesized that CCF is an aeolianfeature, primarily on the basis of a layered surface texture visiblein Viking images of ‘‘classic” CCF terrain; what do high-resolu-tion images suggest about crater-filling aeolian processes (e.g.,sediment and dune deposition)? At both high and low latitudes,dunes partially fill some craters (Fig. 6). Dunes are seldom pres-ent along the entire interior of a crater, and typically drape,

rather than fully fill, the crater interior (Fig. 6). Dune morphol-ogy can be readily distinguished from ‘‘brain terrain” texturesin HiRISE and CTX images, suggesting that CCF is not primarilycomposed of mobile aeolian material. Rather, higher-resolutionimage data reveal that the layered surface texture observed inViking image data (Zimbelman et al., 1989) is a visual effect pro-duced by closely-spaced lineations composed of ‘‘brain terrain”surface texture on topographically undulatory CCF ridges(Fig. 2) (Levy et al., 2009b).

Fig. 6. Dunes in craters are morphologically distinct from CCF. (a) Typical low-latitude dunes. Illumination from lower left. Portion of P08_004197_2007. Illumination fromleft. North to image top. (b) Typical high-latitude dunes. Illumination from lower left. Portion of P16_007187_2581. Illumination from lower left. North to image top.

396 J. Levy et al. / Icarus 209 (2010) 390–404

3.3. CCF and talus

In CTX and HiRISE images, contacts between CCF and talusslopes are clearly discernable (Fig. 7). Some examples of ‘‘low-def-inition” CCF (e.g., in Deuteronilus Mensae or Arabia Terra, Fig. 7)show that ‘‘brain terrain” patterned crater-fill material abuts talusslopes associated with crater rims. The contact between such talusslopes and the CCF is abrupt, with CCF material present down-slope along a steep scarp. Such relationships are most clearly visi-ble at low latitudes, where annuli of mantling material are notpresent at the crater/crater-fill contact (e.g., Fig. 2). Similar con-tacts between LVF and talus scarps in Nilosyrtis Mensae wereinterpreted by Levy et al. (2007) as evidence of the recession ofLVF surfaces from a previous high-stand. This model suggests thatthe talus slope was emplaced down to the level of the (previouslyhigher) LVF margin, and remains as a steep talus slope now thatthe LVF/CCF surface has vertically ablated. In other locations, CCFis topographically separated from the interior slopes of crater wallsby tens to hundreds of meters of relief, suggesting that CCF mate-rial is physically distinct from talus shed by crater walls (Fig. 4)(Russell et al., 2004; Russell and Head, 2005). Lastly, CCF is present

in craters with a range of rim morphologies, ranging from crisp andrelatively well-defined (Figs. 1, 2, and 8) to complexly eroded(Fig. 9), suggesting that CCF is not simply composed of mobilizedtalus shed from crater rims.

3.4. CCF and LVF

CCF and lineated valley fill (LVF) commonly are present in thesame geographical areas on Mars (Squyres, 1979; Squyres and Carr,1986; Head et al., 2006b, 2010). What do contacts between LVF andCCF show? In many locations (Figs. 10 and 11), contacts betweenCCF and LVF show surficial merging of CCF and LVF lineations.LVF-containing valleys commonly breach craters hosting CCF, bothfrom upslope (Fig. 10) and down-slope (Figs. 10 and 11). Mergingof surface lineations occurs in both cases (Head et al., 2006a). HiR-ISE analysis of ‘‘brain terrain” surface textures on both CCF and LVFshow similar groupings of open and closed cells forming character-istic lineations (Fig. 12). Topographic profiles across LVF/CCF con-tacts are gradational and smooth in HiRISE image data, and atboth MOLA resolution (�460 m/pixel) and HRSC-DTM resolution(�100 m/pixel) (Figs. 10 and 11).

Fig. 7. (a) ‘‘Brain terrain” patterned ‘‘low-definition” CCF surface in contact with a crater central peak talus slope. Crater-fill material is significantly lower than the steep talusslope. Inset shows higher-resolution view of talus–CCF contact. Portion of P17_007743_2169. Illumination from left. North to image top. A larger view of this crater is found inFig. 14. (b) ‘‘Brain terrain” patterned ‘‘low-definition” CCF surface in contact with a crater rim talus slope. Crater-fill material is significantly lower than the steep talus slope.Box shows location of part c. (c) Higher-resolution view of talus–CCF contact. For (b) and (c), illumination is from the left, and north is towards image top. Portion of HiRISEimage PSP_005752_2130 overlaid on CTX image P17_005752_2130.

J. Levy et al. / Icarus 209 (2010) 390–404 397

3.5. CCF and LDA

Along with lineated valley fill (LVF), CCF commonly occurs inthe same geographical areas as lobate debris aprons (LDA) (Squy-res, 1979; Squyres and Carr, 1986; Head et al., 2006b, 2010). Insome craters, ‘‘brain terrain” surfaced lobes extend from craterwalls, towards the crater interior, but do not entirely fill the crater(Fig. 13). Lineations on these lobes are commonly concentric withcrater rims. In other craters (Fig. 14), crater-filling material spillsfrom within the crater down to proximal, LVF-surfaced valleyfloors. Such fill is commonly lineated along strike, rather than con-centrically with the crater, similar to lineations in lobate debrisaprons (LDA) (Squyres, 1979; Squyres and Carr, 1986; Head et al.,2006a, 2010; Levy et al., 2007).

3.6. CCF and the LDM

What relationships are observed between CCF and the martianlatitude-dependent mantle (LDM) (Kreslavsky and Head, 1999,2000; Mustard et al., 2001; Head et al., 2003; Milliken et al.,2003; Levrard et al., 2004)? Relatively low-albedo mantling mate-rial typically overlies some portion of CCF surfaces along thedichotomy boundary (e.g., Fig. 2). This mantling material is com-monly polygonally patterned by thermal contraction cracks andranges in thickness from <1 m to several tens of meters (Levyet al., 2009b). This mantling material overlies CCF ‘‘brain terrain”and shows little evidence of deformation associated with CCF flow,consistent with a very youthful age of emplacement (�1–2 Ma), ascompared to the more ancient CCF (see below) (Levy et al., 2009b).

Fig. 8. CCF with a range of impact crater morphologies. Portion of P17_007585_2149. Illumination from upper left. North to image top. (b) Fresh, bowl-shaped impact crater.(c) Three ‘‘ring-mold” craters (Kress and Head, 2008).

Fig. 9. CCF in a crater with an extensively eroded rim. Contrast with CCF in craterswith intact rims (e.g., Figs. 1 and 5). Illumination from left. North to image top.Portion of P16_007385_2149.

398 J. Levy et al. / Icarus 209 (2010) 390–404

4. Age estimates

Crater counting on CCF surfaces poses many of the same chal-lenges as crater counting on LDA and LVF surfaces: small cratersare obscured by ‘‘brain terrain” cells (Levy et al., 2009b) while lar-ger craters may have undergone a range of modification processes(Mangold, 2003; Kress and Head, 2008). Two principle morpholo-gies of craters are observed on CCF surfaces: fresh, bowl-shapedcraters (Fig. 8a and b) and ‘‘ring-mold” craters (RMCs) (Fig. 8aand c) that display the same inner ring or inner plateau morphol-ogies described for RMCs on LVF and LDA by Mangold (2003) andKress and Head (2008). Both fresh and RMC craters are observedon CCF surfaces, suggesting a complex geological history for CCF.

In order to provide an age estimate for the formation of CCF, weavoid counting directly on CCF surfaces. Rather, we identify CTXimages in which both filled (CCF) and unfilled craters are present(Fig. 15, Table 1). By counting on the ejecta of these craters, it ispossible to bracket the formational period of CCF: the CCF musthave formed after the filled-crater ejecta was emplaced, but beforethe unfilled-crater ejecta was emplaced (Fig. 15). Crater countswere made on 10 impact crater ejecta deposits surrounding com-parably-sized craters with and without CCF deposits. The com-bined crater count results are summarized in Fig. 16. Thisbinned-count approach assumes that CCF deposits formed atapproximately the same time throughout the study region. Bothcounts show significant roll-off at small crater size—in part a reso-lution effect associated with CTX data and in part due to small-cra-ter resurfacing at high latitudes due to mantle emplacementprocesses (e.g., Head et al., 2003; Levy et al., 2009a; Kreslavsky,2009). Still, counts on unfilled-crater ejecta show a quantitativelyyounger age. Best-fit ages calculated using Hartmann (2005) isoch-rons and only craters larger than 100 m are �70 Ma for unfilledcraters and �320 Ma for CCF filled craters; similar to age estimatesfor LDA and LVF that span �0.1–1 Ga (Mangold, 2003; Head et al.,2006a,b, 2010; Levy et al., 2007; Kress and Head, 2008). Visualinspection of isochron plots suggest an age for unfilled craters of�60–80 Ma, with filled craters forming a population of surfacesthat may be up to �1 Ga in age. The kink in the filled crater curveat the �0.5 km diameter bin may be indicative of an older popula-tion of craters that has undergone a range of resurfacing processes(Neukum and Hiller, 1981; Werner, 2009), or the Amazonian fillingof considerably older craters. Both curves indicate, however, thatCCF formation is a process typical of the relatively recentAmazonian.

5. Discussion

On the basis of the observations presented above, we can eval-uate hypotheses regarding the origin and significance of concentriccrater fill (CCF). Given the strong morphological dissimilarity be-tween CCF and landforms identified as dunes, we suggest thatCCF does not represent a purely aeolian crater-filling process.However, atmospheric processes such as deposition of windblown

Fig. 10. (a) CCF integrated into lineated valley fill (LVF). Surface lineations characteristic of LVF grade into CCF from image right to center, and from CCF back to LVF fromimage center to left. Illumination from lower right. North to image left. Portion of P12_005870_2181. (b) MOLA gridded topography data extracted from profile from LVF toCCF to LVF. A nearly monotonic down-slope profile suggests integration of CCF into LVF. Regional flow patterns in this area were mapped by Head et al. (2006a).

J. Levy et al. / Icarus 209 (2010) 390–404 399

sediment (e.g., Zimbelman et al., 1989) and snow/ice (Head et al.,2006a,b) may be important in CCF formation (see below). Likewise,CCF may be modified by aeolian processes, such as dust-devilredistribution of surface sediment (e.g., van Gasselt et al., 2010,their Fig. 7). Nonetheless, clear morphological differences exist be-tween CCF surface textures and dunes typical of martian low andhigh latitudes (Fig. 6). Although CCF deposits bear a superficial

resemblance to terrestrial seif dunes, dune-fields typically lackthe concentric pattern typical of CCF, and do not display open-and closed-cell ‘‘brain terrain” morphology (Levy et al., 2009b).

The observation that CCF is commonly elevated above the inte-rior slopes of craters, and in places is markedly convex-up, sug-gests that CCF is not simply crater rim talus material that hasbeen transported down-slope by ice-assisted creep (e.g., Squyres,

Fig. 11. (a) CCF integrated into lineated valley fill (LVF). Surface lineations characteristic of CCF grade into LVF in a confined valley and then into a broader LVF trunk valley.Illumination from lower right. North to image left. Portion of P04_002653_2216. (b) High-resolution HRSC-DTM profile from CCF to LVF. A nearly monotonic down-slopeprofile suggests integration of CCF into LVF.

400 J. Levy et al. / Icarus 209 (2010) 390–404

1979; Squyres and Carr, 1986; van Gasselt et al., 2010). Indeed, CCFis found on craters with well-preserved rims, as well as craters thathave eroded rims (Figs. 1, 4 and 8–10 respectively). Two key mor-phological observations are difficult to reconcile with such a rock-glacier-like mechanism (Haeberli et al., 2006): (1) recession of CCFfrom high stands (implying the removal of a large volume of excess

ice) (Figs. 7 and 2) the presence of convex-up and high-standingCCF in crater interiors (implying preferential removal of CCF mate-rial around crater wall interiors) (e.g., Russell et al., 2004; Russelland Head, 2005) (Fig. 4). In both instances, internal rock–rock con-tact in the rock-glacier model for CCF would provide a uniform sur-face topography, even as interstitial ice was removed from the

Fig. 12. Comparison of similar ‘‘brain terrain” surface textures on CCF (b) and LVF (c). (a) Context image showing CCF integrated with LVF (see Fig. 11). Illumination from theleft. North to image top. (b) Typical CCF ‘‘brain terrain.” Illumination from image bottom. Portion of PSP_008296_2175. (c) Typical LVF ‘‘brain terrain.” Illumination fromimage bottom. Portion of PSP_009588_2175.

Fig. 13. Partial fill of a crater by lobe-like, rim-concentric, ‘‘brain terrain” surfaceddeposits. The partial fill material is similar in position and morphology to lobatedebris aprons observed elsewhere on Mars. Illumination from left. Portion ofP16_007242_2159.

Fig. 14. Lobate debris apron-like flow emerging from a crater and integrating intoLVF. This feature suggests a similar origin for CCF, LVF, and LDA. Illumination fromleft. North to image top. Portion of P04_002442_2221.

J. Levy et al. / Icarus 209 (2010) 390–404 401

deposit. Rather, these morphological observations suggest that sig-nificantly excess ice (ice volume exceeding pore space volume) isrequired to form CCF.

What does the presence of landforms that have been inter-preted as indicators of debris-covered glacial ice in or on CCFdeposits suggest about the origin of concentric crater fill? Ring-mold craters (RMCs) have been interpreted as interactionsbetween impactors and near-surface ice through spallation anddifferential sublimation (Mangold, 2003; Kress and Head, 2008).‘‘Brain terrain” has been interpreted as the surface expression ofa multi-stage sublimation process in the upper portions of CCF,LVF, and LDA surfaces, in which fracture and sediment infill drive

differential sublimation of buried ice (Levy et al., 2009b). Finally,tracing of surface lineations on LVF, LDA, and CCF shows continuityof flow and folding over multiple-km length-scales, strongly sug-gesting the internal deformation and flow of an ice-rich substrateforming these landforms (Head et al., 2006a,b, 2010; Levy et al.,2007; Dickson et al., 2008; Morgan et al., 2009). Indeed, the pres-ence of topographic ridges and troughs, as well as fractures (boththermal-contraction-driven and resulting from structural failureof deforming ice) are typical for terrestrial debris-covered glaciers,and may form parallel or orthogonal to the flow direction(Marchant et al., 2002; Head et al., 2006a,b; Levy et al., 2007).The combined observation of ring-mold craters, ‘‘brain terrain,”and flow-related lineations on CCF all suggest that CCF is composedlargely of debris-covered ice that underwent flow and has beenmodified by sublimation. We interpret CCF to represent crater-interior debris-covered glaciers (e.g., Garvin et al., 2006; Kreslavsky

Fig. 15. Filled and unfilled craters provide a means of estimating the age of CCFemplacement. The most recent CCF emplacement period must be older than the agedetermined by counts on unfilled-crater ejecta and younger than the agedetermined by counts on filled-crater ejecta. Illumination from upper left. Northto image upper right. Portion of P06_003272_2079.

Table 1Index of craters used for crater retention age dating.

Crater type Image number Latitude Longitude

Unfilled P03_003272_2079 28.05�N 34.91�EUnfilled P13_006147_2226 41.45�N 21.43�EUnfilled P15_006964_2274 48.28�N 33.13�EUnfilled P17_007872_2151 35.22�N 82.10�EFilled P04_002562_2203 40.05�N 23.01�WFilled P12_005659_2123 34.49�N 26.41�EFilled P15_006953_2253 44.26�N 26.28�WFilled P15_007081_2183 40.11�N 79.87�EFilled P17_007783_217 43.34�N 8.98�WFilled P17_007677_2287 47.26�N 4.82�E

Fig. 16. Crater counts for ejecta of CCF filled craters (white circles) and unfilledcraters (black circles). Best-fit ages are �60 Ma for unfilled craters and �300 Ma forfilled craters, suggesting that CCF formed during the recent Amazonian.

402 J. Levy et al. / Icarus 209 (2010) 390–404

and Head, 2006; Levy et al., 2009b) that initially flowed down-slope from protected crater-rim accumulation zones (e.g., Fig. 13)before merging and filling crater interiors (e.g., Fig. 1).

What is the relationship between CCF and other martian glaciallandforms, such as LVF and LDA? The morphological similarity be-tween ‘‘brain terrain” surfaces associated with LDA, CCF, and LVFsuggest that these units have undergone similar developmentalhistories. The confluence and merging of CCF, LVF, and LDA linea-tions suggests that they have similar rheological properties andwere active during the same period. The comparable crater reten-tion ages recorded by CCF, LVF, and LDA suggest that similar pro-cesses were involved in the formation and modification of theseunits, over the course of the recent Amazonian. A large body of evi-dence suggests that LVF and LDA formed as debris-covered glaciersas ice accumulated at high obliquity in protected alcoves alongdichotomy boundary slopes, flowed as cold-based glaciers, andeventually sublimated as climate conditions grew colder and/ordrier (Forget et al., 2006; Head et al., 2006a,b, 2010; Levy et al.,2007; Dickson et al., 2008; Madeleine et al., 2009). Morphologicalsimilarities between CCF and LVF/LDA suggest that similar mecha-nisms may have controlled CCF formation and development. Given(1) the detection of buried ice beneath martian LDA and LVF (Holtet al., 2008; Plaut et al., 2009), (2) the thinness of overlying subli-mation residue deposits suggested by the presence of ring-moldcraters (Mangold, 2003; Kress and Head, 2008), and (3) the hun-dreds to thousands of meters of fill material still present in CCFdeposits, we suggest that CCF may represent a significant reservoirof non-polar martian ice, extending over �38,000 km2 along themartian dichotomy boundary.

What does CCF morphology indicate about changes in CCF vol-ume over time? The presence of steep talus slopes above some CCFdeposits (e.g., Fig. 7) suggests that in places at least tens of metersof CCF material has been removed since the cessation of flow. At alarger scale, the presence of ‘‘uphill flowing” low-definition CCFdeposits (Fig. 5), similar to perched glacier deposits observed inColoe Fossae crater interiors (e.g., Dickson et al., 2008, 2009), sug-gests that in places at least 300 m of CCF material has been re-moved. The vertical drop-down of CCF surfaces is consistent withloss of relatively clean glacier ice (Marchant et al., 2002, 2007;Kowalewski et al., 2006; Marchant and Head, 2007), and is incon-sistent with loss of pore ice in grain-supported regolith (e.g., rock-glaciers would not result in a change in surface elevation). Thissuggests that CCF deposits were once considerably more volumi-nous than is indicated by their current extent. For many CCF depos-its, the presence of an additional 300 m of material would over-fillthe craters that the CCF currently occupy, suggesting along withthe presence of breached CCF craters (e.g., Figs. 5 and 10–12), theintriguing possibility that CCF deposits were once part of a larger,inter-crater glacial landsystem (Marchant and Head, 2008, 2009),remnants of which are preserved as CCF, LVF, and LDA deposits,along with ice buried beneath mid-latitude pedestal craters (Kad-ish et al., 2008). If this were the case, then the current extent ofCCF represents only the fraction of the maximum extent of ice thathas been preserved beneath a debris cover as the glaciation waned.

Finally, what is the relationship between CCF and the martianlatitude-dependent mantle (LDM) (Mustard et al., 2001; Headet al., 2003; Milliken et al., 2003; Levrard et al., 2004)? The differ-ence in age between LVF/LDA/CCF and the LDM by one to two or-ders of magnitude suggests that they represent distinct periods ofice deposition on Mars that have left a record of unique character-istic landforms. Young, meters-thick LDM deposits drape CCF, LVF,and LDA (Levy et al., 2009b)—in places showing limited evidence ofrecent deformation as viscous flow features (Milliken et al., 2003).In contrast, �100 Ma old LVF, LDA, and CCF show evidence of inte-grated glacier-like flow over tens of kilometers, and to depths ofhundreds to thousands of meters (Forget et al., 2006; Head et al.,2006a,b, 2010; Levy et al., 2007; Dickson et al., 2008; Madeleineet al., 2009). Accordingly, the transition over the most recentAmazonian from widespread deposition of CCF, LVF, and LDA to

J. Levy et al. / Icarus 209 (2010) 390–404 403

latitude-dependent LDM emplacement may represent a transitionfrom more active glacial periods to a more quiescent ‘‘ice age” per-iod of martian climate evolution (e.g., Head et al., 2003; Levy et al.,2009b).

6. Conclusions

On the basis of morphological analyses of concentric crater fill(CCF) along the martian dichotomy boundary, we conclude that adebris-covered glacier-like formation mechanism is the most con-sistent explanation for the current distribution and characteristicsof CCF, rather than ice-mobilized talus creep or purely aeolian pro-cesses. Given strong morphological similarities between surfacetextures associated with CCF, LVF, and LDA, we suggest that similardebris-covered glacier processes account for the formation of allthree landforms. Filled crater depth–diameter analysis indicatesthat in many locations, hundreds of meters of ice may still bepresent under desiccated surficial debris. Analysis of breachedCCF-containing craters suggests that in at least several locations,CCF-related ice was once several hundred meters higher than itscurrent level, and has sublimated significantly during the mostrecent Amazonian (between �60 and 300 Ma). The presence offormerly more abundant CCF ice, coupled with the integration ofCCF into LVF and LDA glacial landsystems suggests the possibilitythat CCF represents one component of a significant, Amazonian-age glaciation of mid-latitude Mars.

Acknowledgments

We gratefully acknowledge support from the Mars ExplorationProgram Mars Express HRSC Grant JPL 1237163, and the Mars DataAnalysis Program Grants NNG05GQ46G and NNX07AN95G toJ.W.H. Special thanks to Caleb Fassett for assistance in processingand interpreting crater count information and for image process-ing; to James Dickson for coordination of image processing and dis-tribution. The thoughtful comments of two anonymous reviewersgreatly improved this manuscript.

References

Benn, D., Kirkbride, M.P., Owen, L.A., Brazier, V., 2003. Glaciated valley landsystems.In: Evans, D.J.A. (Ed.), Glacial Landsystems. Edward Arnold, London, pp. 372–406.

Dickson, J.L., Head, J.W., Marchant, D.R., 2008. Late Amazonian glaciation at thedichotomy boundary on Mars: Evidence for glacial thickness maxima andmultiple glacial phases. Geology 36, 411–414.

Dickson, J.L., Head, J.W., Marchant, D.R., 2009. Kilometer-thick ice accumulation inthe northern mid-latitudes of Mars: Evidence for crater-filling events in the lateAmazonian at the Phlegra Montes. Earth Planet. Sci. Lett.. doi:10.1016/j.epsl.2009.08.031.

Dobrea, E.Z.N., Asphaug, E., Grant, J.A., Kessler, M.A., Mellon, M.T., 2007. Patternedground as an alternative explanation for the formation of brain coral textures inthe mid latitudes of Mars: HiRISE observations of lineated valley fill textures. In:7th International Conference on Mars, Pasadena, CA (Abstract #3358).

Edgett, K.S., Malin, M.C., and t.C.S.O. Teams, 2008. MRO context camera (CTX)investigation primary mission results. In: American Geophysical Union FallMeet., San Francisco, CA (Abstract #P31D-04).

Forget, F., Haberle, R.M., Montmessin, F., Levrard, B., Head, J.W., 2006. Formation ofglaciers on Mars by atmospheric precipitation at high obliquity. Science 311(5759), 368–371.

Garvin, J.B., Sakimoto, S.E.H., Frawley, J.J., Schnetzler, C., 2002. Global geometricproperties of martian impact craters. Lunar Planet. Sci. 33, League City, TX.Abstract #1255.

Garvin, J.B., Head, J.W., Marchant, D.R., Kreslavsky, M.A., 2006. High-latitude cold-based glacial deposits on Mars: Multiple superposed drop moraines in a craterinterior at 70�N latitude. Meteorit. Planet. Sci., 1659–1674.

Haeberli, W., and 10 colleagues, 2006. Permafrost creep and rock glacier dynamics.Permafrost Periglacial Processes 17, 189–214.

Hartmann, W.K., 2005. Martian cratering 8: Isochron refinement and thechronology of Mars. Icarus 174, 294–320.

Head, J.W., Marchant, D.R., 2003. Cold-based mountain glaciers on Mars: WesternArsia Mons. Geology 31, 641–644.

Head, J.W., Mustard, J.F., Kreslavsky, M.A., Milliken, R.E., Marchant, D.R., 2003.Recent ice ages on Mars. Nature 426, 797–802.

Head, J.W., Marchant, D.R., Agnew, M.C., Fassett, C.I., Kreslavsky, M.A., 2006a.Extensive valley glacier deposits in the northern mid-latitudes of Mars:Evidence for late Amazonian obliquity-driven climate change. Earth Planet.Sci. Lett. 241, 663–671.

Head, J.W., Nahm, A.L., Marchant, D.R., Neukum, G., 2006b. Modification of thedichotomy boundary on Mars by Amazonian mid-latitude regional glaciation.Geophys. Res. Lett. 33. doi:10.1029/2005GL024360.

Head, J.W., Marchant, D.R., Kreslavsky, M.A., 2008. Formation of gullies on Mars:Link to recent climate history and insolation microenvironments implicatesurface water flow origin. Proc. Natl. Acad. Sci. 105, 13258–13263.

Head, J.W., Marchant, D.R., Dickson, J.L., Kress, A.M., Baker, D.M., 2010. Northernmid-latitude glaciation in the late Amazonian period of Mars: Criteria for therecognition of debris-covered glacier and valley glacier landsystem deposits.Earth Planet. Sci. Lett., in press. doi:10.1016/j.epsl.2009.06.041.

Holt, J.W., and 11 colleagues, 2008. Radar sounding evidence for buried glaciers inthe southern mid-latitudes of Mars. Science 322. doi:10.1126/science.1164246.

Jaumann, R., 24 colleagues, and the HRSC Camera Instrument Team, 2007. The high-resolution stereo camera (HRSC) experiment on Mars Express: Instrumentaspects and experiment conduct from interplanetary cruise through thenominal mission. Planet. Space Sci. 55, 928–952.

Kadish, S.J., Head, J.W., Barlow, N.G., Marchant, D.R., 2008. Martian pedestal craters:Marginal sublimation pits implicate a climate-related formation mechanism.Geophys. Res. Lett. 35. doi:10.1029/2008GL034990.

Kowalewski, D.E., Marchant, D.R., Levy, J.S., Head, J.W., 2006. Quantifying low ratesof summer-time sublimation for buried glacier ice in Beacon Valley, Antarctica.Antart. Sci. 18, 421–428.

Kreslavsky, M.A., 2009. Dynamic landscapes at high latitudes on Mars: Constraintsfrom populations of small craters. Lunar Planet. Sci., The Woodlands, TX, March23–27. Abstract #2311.

Kreslavsky, M., Head, J., 1999. Kilometer-scale slopes on Mars and their correlationwith geologic units: Initial results from Mars Orbiter Laser Altimeter (MOLA)data. J. Geophys. Res. 104, 21911–21924.

Kreslavsky, M.A., Head, J.W., 2000. Kilometer-scale roughness on Mars: Results fromMOLA data analysis. J. Geophys. Res. 105, 26695–26712.

Kreslavsky, M.A., Head, J.W., 2006. Modification of impact craters in the northernplains of Mars: Implications for the Amazonian climate history. Meteorit.Planet. Sci. 41, 1633–1646.

Kress, A.M., Head, J.W., 2008. Ring-mold craters in lineated valley fill and lobatedebris aprons on Mars: Evidence for subsurface glacial ice. Geophys. Res. Lett.35. doi:10.1029/2008GL035501.

Levrard, B., Forget, F., Montmessin, F., Laskar, J., 2004. Recent ice-rich depositsformed at high latitudes on Mars by sublimation of unstable equatorial iceduring low obliquity. Nature 431, 1072–1075.

Levy, J.S., Head, J.W., Marchant, D.R., 2007. Lineated valley fill and lobate debrisapron stratigraphy in Nilosyrtis Mensae, Mars: Evidence for phases of glacialmodification of the dichotomy boundary. J. Geophys. Res. 112. 1029/2006JE002852.

Levy, J.S., Head, J.W., Marchant, D.R., 2009a. Thermal contraction crack polygons onMars: Classification, distribution, and climate implications from HiRISEobservations. J. Geophys. Res. 114. doi:10.1029/2008JE003273.

Levy, J.S., Head, J.W., Marchant, D.R., 2009b. Concentric crater fill in Utopia Planitia:Timing and transitions between glacial ‘‘brain terrain” and periglacialprocesses. Icarus 202. doi:10.1016/j.icarus.2009.1002.1018.

Lucchitta, B., 1984. Ice and debris in the fretted terrain, Mars. J. Geophys. Res. 89,B409–B418.

Madeleine, J.-B., Forget, F., Head, J.W., Levrard, B., Montmessin, F., Millour, E., 2009.Amazonian northern mid-latitude glaciation on Mars: A proposed climatescenario. Icarus. doi:10.1016/j.icarus.2009.1004.1037.

Malin, M.C., Edgett, K.S., 2001. Mars Global Surveyor Mars Orbiter Camera:Interplanetary cruise through primary mission. J. Geophys. Res. 106, 23429–23570.

Mangold, N., 2003. Geomorphic analysis of lobate debris aprons on Mars at MarsOrbiter Camera scale: Evidence for ice sublimation initiated by fractures. J.Geophys. Res. 108, 8021–8033.

Marchant, D.R., Head, J.W., 2007. Antarctic Dry Valleys: Microclimate zonation,variable geomorphic processes, and implications for assessing climate changeon Mars. Icarus 192 (1), 187–222. doi:10.1016/j.icarus.2007.06.018.

Marchant, D.R., Head, J.W., 2008. Kilometer-thick ice-sheet in the northern mid-latitudes in the Amazonian: Analogs from the East Antarctic Ice Sheet and thedry valleys. Lunar Planet. Sci. 39, League City, TX. Abstract #2097.

Marchant, D.R., Head, J.W., 2009. The glacial deposits of the northern mid-latitudes:Remnants of large-scale plateau glaciation. Lunar Planet. Sci. 40, TheWoodlands, TX. Abstract #2355.

Marchant, D.R., Denton, G.H., Swisher, C.C.I., 1993. Miocene–Pliocene–Pleistoceneglacial history of Arena Valley, Quartermain Mountains, Antarctica. Geogr. Ann.75A, 269–302.

Marchant, D.R., Lewis, A.R., Phillips, W.M., Moore, E.J., Souchez, R.A., Denton, G.H.,Sugden, D.E., Potter, N.J., Landis, G.P., 2002. Formation of patterned ground andsublimation till over Miocene glacier ice in Beacon Valley, southern VictoriaLand, Antarctica. Geol. Soc. Am. Bull. 114, 718–730.

Marchant, D.R., Phillips, W.M., Schaefer, J.M., Winckler, G., Fastook, J.L., Shean, D.E.,Kowalewski, D.E., Head, J.W., Lewis, A.R., 2007. Establishing a chronology for theworld’s oldest glacier ice. Paper presented at 10th International Symposium onAntarctic Earth Sciences, Santa Barbara, CA.

404 J. Levy et al. / Icarus 209 (2010) 390–404

McEwen, A.S., and 69 colleagues, 2009. The High Resolution Imaging ScienceExperiment (HiRISE) during MRO’s primary science phase. Icarus. doi:10.1016/j.icarus.2009.1004.1023

Milliken, R.E., Mustard, J.F., Goldsby, D.L., 2003. Viscous flow features on the surfaceof Mars: Observations from high-resolution Mars Orbiter Camera (MOC)images. J. Geophys. Res. 108. doi:10.1029/2002JE002005.

Morgan, G.A., Head, J.W., Marchant, D.R., 2009. Lineated valley fill (LVF) and lobatedebris aprons (LDA) in the Deuteronilus Mensae northern dichotomy boundaryregion, Mars: Constraints on the extent, age, and episodicity of Amazonianglacial events. Icarus 202, 22–38.

Mustard, J.F., Cooper, C.D., Rifkin, M.K., 2001. Evidence for recent climate change onMars from the identification of youthful near-surface ground ice. Nature 412,411–414.

Neukum, G., Hiller, K., 1981. Martian ages. J. Geophys. Res. 86, 3097–3121.Oberst, J., Scholten, F., Gwinner, K., Giese, B., Waehlisch, M., Roatsch, T., Neukum, G.,

2005. The topographic mapping performance of the HRSC (High ResolutionStereo Camera) on Mars Express. In: American Geophysical Union Fall Meeting,San Francisco, CA (Abstract #G53C-03).

Plaut, J.J., Safaeinili, A., Holt, J.W., Phillips, R.J., Head, J.W.I., Seu, R., Putzig, N.E.,Frigeri, A., 2009. Radar evidence for ice in lobate debris aprons in the mid-northern latitudes of Mars. Geophys. Res. Lett. 36. doi:10.1029/2008GL036379.

Russell, P.S., Head, J.W., 2005. Circumpolar craters with interior deposits on Mars:Polar region geologic volatile and climate history with implications for groundice signature in Arabia Terra. Lunar Planet. Sci. 36, League City, TX. Abstract#1541.

Russell, P.S., Head, J.W., Hecht, M.H., 2004. Evolution of ice deposits in the localenvironment of martian circum-polar craters and implications for polar caphistory. Lunar Planet. Sci. 35, League City, TX. Abstract 2007.

Sharp, R.P., 1973. Mars: Fretted and chaotic terrains. J. Geophys. Res. 78, 4073–4083.

Shean, D.E., Head, J.W., Marchant, D.R., 2005. Origin and evolution of a cold-basedtropical mountain glacier on Mars: The Pavonis Mons fan-shaped deposit. J.Geophys. Res. 110. doi:10.1029/2004JE002360.

Squyres, S.W., 1978. Martian fretted terrain: Flow of erosional debris. Icarus 34,600–613.

Squyres, S.W., 1979. The distribution of lobate debris aprons and similar flows onMars. J. Geophys. Res. 84, 8087–8096.

Squyres, S.W., Carr, M.H., 1986. Geomorphic evidence for the distribution of groundice on Mars. Science 231, 249–252.

van Gasselt, S., Hauber, E., Neukum, G., 2010. Lineated valley fill at the martiandichotomy boundary: Nature and history of degradation. J. Geophys. Res., inpress. doi:10.1029/2009JE003336.

Werner, S.C., 2009. The global martian volcanic evolutionary history. Icarus 201,44–68.

Williams, R.M.E., 2006. Latitude dependence of meter-scale surface textures inDeuteronilus Mensae, Mars. Lunar Planet. Sci. 37, League City, TX. Abstract#1445.

Zimbelman, J.R., Clifford, S.M., Williams, S.H., 1989. Concentric crater fill on Mars:An aeolian alternative to ice-rich mass wasting. Paper presented at 19th Lunarand Planetary Science Conference, Houston, TX.