Transverse Aeolian Ridges on Mars: First results from HiRISE images

Geomorphology 121 (2010) 22–29

Contents lists available at ScienceDirect

Geomorphology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / g e o m o r p h

Transverse Aeolian Ridges on Mars: First results from HiRISE images James R. Zimbelman ⁎ Center for Earth and Planetary Studies, MRC 315, National Air and Space Museum, Smithsonian Institution, Washington, D.C. 20013-7012, United States

a r t i c l e

i n f o

Article history: Received 11 July 2008 Received in revised form 23 February 2009 Accepted 26 May 2009 Available online 2 June 2009 Keywords: Sand Dune Ripple Granule ripple Topography Transverse dune

a b s t r a c t Three images obtained by the High Resolution Imaging Science Experiment (HiRISE) were analyzed for the information they could provide regarding Transverse Aeolian Ridges (TARs) on Mars. TARs from five locations in a HiRISE image of the floor of Ius Chasma show remarkably symmetric (cross-sectional) profiles, with average slopes for the entire feature of ~15°; these results apply to TARs that span an order of magnitude in wavelength and a factor of 6 in height. A HiRISE image of Gamboa impact crater in the northern lowlands shows low albedo sand patches b 2 m high that are covered with sand ripples, surrounded by larger TAR-like ripples that are very similar in profile to surveyed granule ripples on Earth. TARs in a HiRISE image from Terra Sirenum, in the cratered southern highlands, are comparable in height to those in Ius Chasma, but many have tapered extensions that are more consistent with them being erosional remnants rather than the result of extension of the TAR by deposition from the tapered end. The new observations generally support a reversing transverse dune origin for TARs with heights ≥ 1 m, and a granule ripple origin for TAR-like ripples with heights ≤ 0.5 m. Published by Elsevier B.V.

1. Introduction “Transverse Aeolian Ridge” (TAR) is a general term proposed for linear to curvilinear aeolian features that could be the result of either dune or ripple processes (Bourke et al., 2003). These features are nearly always oriented with crests perpendicular to wind directions that could result from topographic confinement, as on the floor of a valley or linear depression (Fig. 1). As summarized below, TARs were first described from Mars Orbiter Camera (MOC) images, but since November of 2006, aeolian features of various kinds are abundant in the remarkable images being returned by the High Resolution Imaging Science Experiment (HiRISE) camera (e.g., Bridges et al., 2007). This report presents observations of TARs derived from three HiRISE images that improve on our knowledge about these distinctive windrelated features, as first reported in a presentation at the workshop on “Planetary dunes: A record of climate change”, which was held in Alamogordo, New Mexico (Zimbelman, 2008). 2. Background 2.1. Sand ripples and granule ripples Bagnold (1941) described a “characteristic path” for well sorted, wind-blown sand grains, which he argued was the result of momentum removed from the wind at the height reached by most

⁎ Tel.: +1 202 633 2471; fax: +1 202 786 2566. E-mail address: [email protected]. 0169-555X/$ – see front matter. Published by Elsevier B.V. doi:10.1016/j.geomorph.2009.05.012

of the saltating grains (p. 62). He further argued that the characteristic path represented a length scale determined by the velocity gained from the wind, from which he inferred that ripple wavelength must be the physical manifestation of the characteristic path of wind-driven sand (Bagnold, 1941, p. 150). Sharp (1963) argued that because ripples begin as small-amplitude, short-wavelength forms, and change wavelength as they evolve to steady state, Bagnold's characteristic path concept was suspect. Through geometric arguments, Sharp (1963) proposed that ripple wavelength depends on ripple amplitude and the angle at which saltating grains approach the bed, controlled by the length of the zone shadowed from significant impact and by wind velocity. Anderson (1987) developed an analytical model for the initiation of sand ripples produced by perturbations on a flat sand surface. Using the “splash function” concept derived from experiments by Unger and Haff (1987), Anderson (1987) modeled the relatively low-velocity ejection of grains caused by the impact of faster-moving saltating grains in a process termed ‘reptation’, from the Latin for “to crawl”, to distinguish this mode of sand transport from saltation, suspension, and impact creep. Numerical modeling including saltation and reptation (Anderson and Haff, 1988) showed that small, fast-moving ripples overtake and merge with larger, slower forms, resulting in the growth of the mean wavelength and a decline in the dispersion of wavelengths. Reptation and saltation are now incorporated into continuum analytical models developed for study of aeolian features on Earth and Mars (e.g., Momiji et al., 2000; Yizhaq et al., 2004; Yizhaq, 2005). When aeolian sand is poorly sorted, the coarse grains become concentrated at the crests of features Bagnold (1941, p. 153–6) termed

J.R. Zimbelman / Geomorphology 121 (2010) 22–29

23

2.3. Dunes and TARs on Mars

Fig. 1. Transverse Aeolian Ridges (TARs) on the floor of Nirgal Vallis, Mars. TARs are oriented with crest axes perpendicular to trend of the valley floor, with wavelengths of from 30 to 100 m. Portion of MOC image E02-02651, 27.8° S, 316.7° E, 2.8 m/pixel, NASA/ JPL/MSSS.

‘ridges’ rather than ripples, arguing that the coarse grains are moved by the impact of saltating sand (also called ‘impact creep’, Greeley and Iversen,1985, p.17). Sharp (1963) described ‘granule ripples’ consisting of a surface concentration of grains N1 mm, roughly equivalent to Bagnold's ‘ridges’; Sharp concurred with Bagnold that impact creep, along with significant deflation, was the dominant process of motion in granule ripples. Granule ripples can grow to great size, attaining wavelengths exceeding 20 m and heights of over 60 cm (Bagnold, 1941, p. 155), although they are more commonly observed in the wavelength range of ~ 3 to 10 m with heights of ~25 cm (Sharp, 1963; Zimbelman and Williams, 2005). The granule coating tends to be concentrated at the surface, with the interior of the bedform comprised primarily of sand with only occasional granules; the granules are most thickly accumulated at the crest of the ripple (Sharp, 1963). 2.2. Ripples versus dunes Bagnold (1941) described important distinctions between ripple and dune processes; dunes usually display a slip face dominated by avalanche processes (p. 189) while ripples have the coarsest material collected at the crests with the finest material in the troughs, whereas the opposite is true for dunes (p. 145). Normal sand ripples (as described in Section 2.1) are the smallest member in a hierarchy of aeolian landforms. Wilson (1972a,b) documented three distinct scales for aeolian landforms; ripples (wavelength 0.01 to 10 m), dunes (wavelength ~ 10 to 500 m), and draas (wavelength ~ 0.7 to 5.5 km). Note that Wilson's work includes both (small) sand ripples and (large) granule ripples in the features he called ‘ripples’. Regions of overlap exist between small dunes and large ripples, but the sediment size of a 10 m dune is typically b0.2 mm while surface sediments on a ripple of 10 m wavelength are N1 mm, and no observed transitional features occur between these two groups (Wilson, 1972a). Therefore, particle size is an important determinant of which class a 10-m-scale aeolian landform belongs to; this 10 m size is comparable to the smallest features initially identified as TARs on Mars (Wilson and Zimbelman, 2004). Features at all three scales can be present at one location, but they are built by differing wind intensities and durations (Wilson, 1972b), consistent with the ideas of Sharp (1963), who stated that each form reaches its own quasiequilibrium state; i.e., ripples do not grow to become dunes, nor do dunes grow into draas.

Dunes were first clearly identified on Mars in Mariner 9 images through comparison with terrestrial dunes (McCauley et al., 1972; Cutts and Smith, 1973). Viking Orbiter images led to many reports about Martian dunes, but primarily for the barchans and barchoid ridges in the extensive north polar erg (e.g., Cutts et al., 1976; Breed et al., 1979; Tsoar et al., 1979; Lancaster and Greeley, 1990). Most Martian dunes appear to be transverse in orientation, with crescentic barchans the dominant dune type in the north polar erg (Greeley et al., 1992). Viking resolution was insufficient to indicate slip faces except for the largest dunes of the north polar erg (Tsoar et al. 1979). Zimbelman (1987) examined all Viking Orbiter images with a resolution b10 m/ pixel; one fourth of these highest resolution images included patches of aeolian lineations, essentially independent of the image location on Mars. The Mars Orbiter Camera (MOC) greatly improved on the Viking resolution, providing images of up to 1.5 m/pixel resolution (Malin and Edgett, 2001). MOC revealed a wealth of aeolian features across the planet, and its images clearly resolved slip faces on some dunes N150 m in width. The term ‘Transverse Aeolian Ridge’ (TAR) was first proposed by Bourke et al. (2003) for aeolian features on Mars where slip faces are not observed directly (e.g., Fig. 1), to preserve both ripple and dune options for their origin. A systematic examination of MOC images within a longitudinal sector (~ 240° to 180°E) from the south to the north poles documented the latitudinal distribution of TARs (Wilson and Zimbelman, 2004); equatorward of ~ 60° latitude, TARs are relatively abundant, most with wavelengths of 20 to 60 m, whereas TARs are much less frequent in the polar regions. Balme et al. (2008) and Shockey and Zimbelman (2007) extended the survey of MOC images that show TARs to longitudinal sectors ~180° from that of the Wilson and Zimbelman (2004) study, obtaining results for TAR distributions that are comparable to those from the earlier study. TARs therefore are relatively ubiquitous throughout the equatorial and midlatitudes of Mars. The Mars Exploration Rovers (MERs) Spirit (Squyres et al., 2004a) and Opportunity (Squyres et al., 2004b) have provided a wealth of new information about aeolian bedforms and sediments. The MERs have documented the presence of granule ripples at both sites, as well as broad deposits comprised primarily of basaltic sand (Greeley et al., 2004, 2006; Sullivan et al., 2005, 2008). The High Resolution Imaging Science Experiment (HiRISE) now provides the capability to obtain orbital images of Mars with a spatial resolution of as good as 25 cm/ pixel (McEwen et al., 2007). Initial results from HiRISE provided important new insights into aeolian deposition and erosion on Mars, such as multiple length scales observed for aeolian features at many locations (Bridges et al., 2007). HiRISE images will not resolve individual particles that comprise aeolian landforms on Mars. The obscuring effect of the pervasive Martian dust makes it difficult to derive particle size information for individual landforms through infrared remote sensing, except at locations where dust deposition is inhibited by the activity of large sand deposits (e.g., Fenton, 2005; Fenton and Mellon, 2006). The only confirmation of particle size for Martian aeolian features currently comes from rovers that can place a Microscopic Imager directly onto the feature (e.g., Greeley et al., 2004; Herkenhoff et al., 2004). 3. Ius Chasma The first publicly released full resolution HiRISE image (TRA_ 000823_1720, centered at 7.7°S, 279.5°E, in aerocentric coordinates) demonstrated that 25 cm/pixel (map-projected) resolution was obtainable from the mapping orbit. This first image was of the floor of Ius Chasma, part of the extensive Valles Marinaris canyon system (Fig. 2). The image revealed intricate layering exposed in the deposits of the canyon floor, along with several prominent patches of TARs. Nearly all of

24

J.R. Zimbelman / Geomorphology 121 (2010) 22–29

Fig. 2. First publicly released full-resolution HiRISE image of Mars, which includes numerous TARs scattered across the floor of Ius Chasma. A through E indicate the locations of TARs where profiles were obtained from simplified photoclinometry (see Table 1 and Fig. 3). Browse version of HiRISE image TRA_000823_1720, 7.7° S, 279.5° E, 25 cm/pixel mapprojected resolution, taken 9/29/06, NASA/JPL/U of A.

the TARs in the Ius Chasma image have a ‘simple’ crest-ridge plan view morphology, following the classification scheme of Balme et al. (2008). TARs were measured at five locations in the Ius Chasma image (letters in Fig. 2) by collecting the pixel values along lines perpendicular to the local crest orientation of the TARs. The measured profile lines typically were within ~30° of the solar illumination azimuth, and the solar incidence angle was 30° from the horizon. The pixel values along each line were used to generate a photoclinometric profile through two simplifying assumptions: 1) the albedo (visual reflectivity) of the materials is assumed to be constant along the line (that is, the line does not cross albedo boundaries), and 2) the photometric properties of the surface materials are considered to be Lambertian, where reflected brightness is solely dependent upon the solar incidence angle to the surface (Zimbelman and Williams, 2008). Non-Lambertian aspects of Martian materials (i.e., Minneart photometry) can cause profiles derived using Lambertian photometry to produce slopes that are slightly (less than a few percent change) steeper than reality (R. Kirk, pers. comm.), but still the overall profile shape is correctly reproduced in a Lambertian profile. Plans are in place to compare Lambertian profiles to the more accurate relief obtained from stereo HiRISE images (e.g., Kirk et al., 2007), but results from such a comparison are not available at present. In the meantime, slopes on profiles presented here should be considered to represent upper limits. Remarkably symmetric profiles were obtained for TARs from the five locations examined within the Ius Chasma image. The profile for location C (Fig. 3) is illustrative of TAR profiles from all five locations (data for the C profile appear in Fig. 6b of Balme et al., 2008). When the two sides of the TAR profile are folded back on itself at the crest, the upper slopes on either side of the crest are essentially identical.

The slope within 2 m of either side of the crest in Fig. 3 is 24°; even considered as an upper limit, this value is less the angle of repose expected for a slip face. The TAR profiles are more symmetric than profiles measured across several terrestrial aeolian features ranging from normal sand ripples through granule ripples to stabilized and

Fig. 3. Profile across TAR at location C in Fig. 2. The profile shape is remarkably symmetric, similar to symmetric reversing transverse dunes on Earth (see Fig. 4). Inset shows profile location. ‘v’s indicate breaks in slope at the bases of both sides of the ridge, which is defined here to represent the width of the feature. Vertical exaggeration of the profile is 6.1×. Data from HiRISE image TRA_000823_1720 (Fig. 2).

J.R. Zimbelman / Geomorphology 121 (2010) 22–29

active transverse dunes, covering more than three orders of magnitude variation in feature size (Zimbelman and Williams, 2005). The one terrestrial feature that can approach the symmetry displayed by the TAR profiles is a reversing sand dune, where bimodal winds pile up the sand along a ridge that is perpendicular to both wind directions (e.g., Burkinshaw et al., 1993; Frank and Kocurek, 1996; McKenna Neuman et al., 1997; Walker, 1999; Pye and Tsoar, 2009, pp. 270–272). While active reversing dunes may not often achieve the striking profile symmetry of the TARs, one reversing dune at Great Sand Dunes National Park and Preserve (GSDNPP), in south-central Colorado, produced a measured profile similar to that of TARs like the one shown in Fig. 3 (Zimbelman, 2008). A tool that holds promise for assessing the similarity of dune and ripple shapes to that of TARs is the comparison of similarly scaled terrestrial and Martian profiles. We experimented with scaling a TAR profile from Ius Chasma and surveyed profiles of several terrestrial aeolian landforms, with mixed results. Use of wavelength as the scaling factor suffered from the great degree of variability in wavelength present in aeolian environments on Earth and Mars; an ‘average’ wavelength showed such variability, even within the same dune or ripple field, that use of this parameter as a scaling factor proved to not be very satisfactory (Zimbelman and Williams, 2007a). Next, scaling of the profiles by the height of the feature was explored (Zimbelman and Williams, 2007b), but height is the most difficult dimension to measure remotely (Bourke et al., 2006), so that uncertainty and variability was not improved over scaling by wavelength. Scaling of both the horizontal and vertical dimension of profiles by the width of the feature, defined to be the distance between the basal break in slope on either side of the crest (‘v’s in Fig. 3), has proven to be the best scaling approach examined to date, which improved the reproducibility of profiles over both the variability of the wavelength scaling and the uncertainty of the height scaling (Zimbelman and Williams, 2007c, 2008). To illustrate this technique, the TAR profile from Fig. 3 was scaled by the feature width (determined from the breaks in slope in the profile) and compared to similarly scaled tape-inclinometer profiles of reversing transverse dunes at Coral Pink Sand Dunes State Park in southern Utah (Fig. 4); the profile similarities become readily apparent in the scaled plots. The advantage of using scaled profiles is that details of the profile shape are preserved (compare TAR profiles in Figs. 3 and 4) while eliminating the size differences between the features being compared.

Fig. 4. Comparison of profiles across reversing transverse dunes at Coral Pink Sand Dunes State Park in southern Utah (Earth1 and Earth2) and the Ius Chasma TAR shown in Fig. 3 (Mars). Both height and distance have been scaled by the width of each feature, which is shown in parentheses (in meters) in the symbol key. The scaled heights of both the terrestrial transverse dunes and the TAR are nearly twice the scaled height of granule ripples and Gamboa crater ripples (see Fig. 8). Vertical exaggeration of the profiles is 5.0×.

25

Table 1 Dimensions of individual TARs in Ius Chasma and Terra Sirenum. Loca

Wavelengthb (m)

Widthc (m)

Heightd (m)

Ave. slopee (°)

A B C D E F G H J K L M N

7 24 38 70 21 41 41 41 40 40 40 40 40

7.0 18.0 45.3 29.0 13.5 20.5 11.5 10.5 34.0 22.5 24.5 16.5 25.5

1.0 1.7 5.8 5.8 1.7 4.7 1.6 1.5 8.0 5.7 5.4 2.3 6.7

16 11 14 22 14 25 16 16 25 27 24 16 28

a

See Figs. 2, 9, and 10 for locations. A–E, average of crest distance to either side of measured TAR crest; F–H and J–N, average crest-to-crest distance across Figs. 9 and 10, respectively. c A–E, distance between basal breaks in slope on either side of crest, from TAR profile; F–N, twice the length of the sunlit face of TAR, perpendicular to crest direction. d A–E, height difference from crest to basal break in slope, from TAR profile; F–N, height from shadow length. e Average slope is the arc-tangent of height divided by half the width. b

Table 1 provides dimensions obtained from profiles across individual TARs at the five locations indicated in Fig. 2, where wavelength is the average of the distance between the crests to either side of the measured crest (following the profile orientation), width is the horizontal distance between the basal breaks in slope on both sides of the crest, and height is the elevation difference between the basal breaks in slope and the crest. If height is divided by half of the feature width, the average slope values are near 15°, with one outlier (D in Table 1). Even as upper limits, these average slopes are indicative of landforms that are intermediate between that of granule ripples (~ 8°; Zimbelman et al., 2009) and sand at the angle of repose (~ 33°). 4. Gamboa crater Gamboa impact crater, at 40.5°N, 315.5°E (THEMIS, Thermal Emission Imaging Spectrometer; Christensen et al., 2003, image V12571006), was the target of repeated imaging by MOC to monitor gullies present on the interior crater walls. Some of the MOC images also crossed the central peak on the crater floor (e.g., MOC image R1600011), around which are located several dune patches with a lower albedo than the surroundings, similar to the “Large Dark Dunes" discussed by Balme et al. (2008). The extensive image archive for this crater from earlier missions caused it to be targeted early by HiRISE; image PSP_002721_2210 (25 cm/pixel map-projected resolution) provides new insight into dunes and ripples present around the central peak. The solar incidence angle in this HiRISE image is 24° above the horizon, which makes even subtle features clearly visible. The dark dune patches seen in the MOC images are now resolved into rounded to elongate sand masses surrounded by a complex array of TAR-like ripples (Fig. 5). No slip faces are evident on any of the sand patches near the center of Gamboa crater, but the patches are all thoroughly covered by ripples of wavelength ~ 1 m. The overall appearance of the sand patches (Fig. 5) is quite different from that of TARs (inset in Fig. 3), where a sharp crest is a dominant attribute, and the sand patches do not fit into any of the TAR classification types of Balme et al. (2008). The sand patch dimensions are greater than ~40 m, which is consistent with the concept of Bagnold (1941, p. 183) regarding the minimum size of a sand patch that can be considered a dune, where saltation across the patch alters the wind sufficiently to induce sand deposition on (rather than erosion of) the sand patch. The ripples surrounding the sand patches display crests that are oriented radial to the sand mass (Fig. 5), which suggests that the local wind

26

J.R. Zimbelman / Geomorphology 121 (2010) 22–29

Fig. 5. Broad sand patches, surrounded by ripples, on the floor of Gamboa crater south of the central peak. The sand patches lack any evidence of a slip face, and are covered by ripples with wavelength ~ 1 m. Lines A–A′ and B–B′ show the location of profiles in Figs. 6 and 7, respectively. C is a boulder whose shadow length indicates it is 0.8 m high. Portion of HiRISE image PSP_002721_2210, 40.8° N, 315.7° E, 25 m/pixel map-projected resolution, taken 2/24/07, NASA/JPL/U of A.

Fig. 7. Profile across two TAR-like ripples adjacent to a broad sand patch (B–B′ in Fig. 5). Inset shows the location of the profile. The ripples have dimensions and shape very similar to that of granule ripples (see Fig. 8). Vertical exaggeration of the profile is 21.9×. Data from HiRISE image PSP_002721_2210 (Fig. 5).

flow was influenced by the presence of the sand patches; some of the radially oriented ripples may be superposed on the lower margins of the dune, although it is also possible that the sand patches may bury some preexisting ripples. The photoclinometry profiling technique described above was applied to a broad sand patch and nearby ripples along a single line, taking care to restrict the profile line to a region that appears to have a relatively uniform albedo. The sand patch has limited relief, b2 m in height, and shallow slopes across the entire feature (Fig. 6); similar heights are expected anywhere along the sand patch. The sand patch relief (Fig. 6) is less than half the relief of a TAR of comparable width (Fig. 3), and the sand patch also lacks even a hint of the sharp crest observed on all TARs. Ripples immediately adjacent to the sand patch have relief of 0.2 to 0.3 m, while ripples on the surface of the sand patch have relief ≤0.05 m; these heights suggest that ripples off the

sand patches may be granule ripples whereas those on the sand patches are more likely to be normal sand ripples (Zimbelman and Williams, 2005, 2006). The granule ripple hypothesis can be evaluated by considering some of the large ripples well removed from the sand patch (Fig. 7). When the two Martian ripple profiles are individually scaled by the feature width, as described earlier, the Martian profiles are seen to be very comparable to similarly scaled profiles of granule ripples at GSDNPP (Zimbelman and Williams, 2006) and in the Coachella Valley of southern California (Sharp, 1963; Zimbelman and Williams, 2006) (Fig. 8). The scaled height of the terrestrial granule ripples and the Gamboa ripples (Fig. 8) are much smaller than the scaled heights of either the Ius Chasma TAR or the terrestrial reversing dunes (Fig. 4), providing a good method to distinguish between dunes and granule ripples in HiRISE images. The Gamboa image

Fig. 6. Profile across a broad sand patch on the floor of Gamboa crater (A–A′ in Fig. 5). Inset shows the location of the profile. The sand patch is b2 m tall and shows no evidence of a slip face, but it is still sufficiently broad to be considered a minimal dune (see text). Small ripples are superposed on the dune; they are most discernable along the profile on the sunlit side of the dune. Ripples immediately adjacent to the margin of the dune are 4 times the height of ripples on the sand patch surface, and are likely small examples of ripples like those in Fig. 7. Vertical exaggeration of the profile is 16.9×. Data from HiRISE image PSP_002721_2210 (Fig. 5).

Fig. 8. Comparison of profiles across granule ripples at Great Sand Dunes National Park and Preserve in south-central Colorado (Earth1), in the Coachella Valley of southern California (Earth2), and two Martian TAR-like ripples (Mars1 and Mars2, left and right in Fig. 7, respectively). Both height and distance have been scaled by the width of each feature, which is shown in parentheses (in meters) in the symbol key. The scaled heights of both the terrestrial and Martian ripples are about half the scaled height of terrestrial transverse dunes and a TAR (see Fig. 4). Vertical exaggeration of the profiles is 10.0×.

J.R. Zimbelman / Geomorphology 121 (2010) 22–29

27

provides examples of features that can be confidently categorized as small dunes (the broad sand patches lacking slip faces), normal sand ripples (the small ripples on the sand patches), and granule ripples (the large TAR-like ripples surrounding the sand patches), all of which are very distinct from the Ius Chasma TARs (Fig. 3). 5. Terra Sirenum A HiRISE image from the cratered highlands in the southern midlatitudes was examined to compare attributes of mid-southernlatitude TARs to those seen in the HiRISE image of equatorial Ius Chasma. Crater walls in the Terra Sirenum region of Mars were imaged by HiRISE to gain insight into the numerous gullies present in this area. HiRISE image PSP_001684_1410, centered at 38.9°S, 196.0°E, has 25 cm/pixel map-projected resolution, and while gullies are common throughout the image, so are many interesting TARs. The solar incidence angle in this image is only 16° above the horizon, an angle comparable to the average slope of the TARs in Ius Chasma, so that distinct shadows are cast by individual TARs (Figs. 9 and 10). Shadow lengths were measured along the solar incidence azimuth (clearly indicated by the shadows of individual boulders; see Fig. 8), and then converted into the height of the TAR crests. The numbers in Figs. 9 and 10 give the TAR crest height in meters, next to the line where the shadow length was measured. The strong shadows preclude applying the photoclinometric technique to generate a complete profile across the features, but obtaining a height from a shadow length is generally less dependent on assumptions than is the photoclinometry procedure. However, the shadow lengths can be affected by subtle irregularities on the surface over which the shadow is cast, which might cause some shadows to appear longer than they would on a flat surface. TAR heights (obtained from the shadow lengths) and widths (obtained from doubling the sunlit width perpendicular to the crest at the point of the shadow measurement) at Terra Sirenum are generally consistent with those obtained from profiles across the Ius Chasma TARs (Table 1). The average slopes of the Terra Sirenum TARs fall into two distinct groups, much like the Ius Chasma TARs did. It is instructive to note that the one average slope outlier from Ius Chasma (D in Table 1) is

Fig. 9. TARs in Terra Sirenum. TARs taper to very thin linear ridges extending from their northern ends (three small arrows, upper right), some of which separate into isolated linear segments following the trend of the nearby TAR (four small arrows, upper left). The narrow extensions suggest erosion of inactive TARs rather than extension of active TARs along such narrow bands. Subtle aeolian bedforms are visible both between and on the TARs (large arrows). Shadow lengths provide TAR heights (in meters) at three locations (lines indicate where shadow length was measured). Letters indicate measurement locations for Table 1. Portion of HiRISE image PSP_001684_1410, 38.9° S, 196.0° E, 25 cm/pixel map-projected resolution, taken 12/5/06, NASA/JPL/U of A.

Fig. 10. Forked terminations of Terra Sirenum TAR crests, along with evidence of eroded layers exposed both in the interdune regions and even on the slopes of some TARs. The low solar incidence angle (16°) casts long shadows that provide heights (in meters) for several TARs (white lines show where shadow length was measured) and one boulder. Letters indicate measurement locations for Table 1. Portion of HiRISE image PSP_001684_1410 (as in Fig. 9).

comparable to the majority of the average slopes for the Terra Sirenum TARs, suggesting a possible bimodal distribution of average slope may exist for TARs in general. Many TARs in Terra Sirenum show a pronounced ‘tail’ from the northeast end of their crest, some of which can be followed down to widths of only 1 m, while others break into meter-wide linear features aligned with the tapered end of a nearby TAR (small arrows in Fig. 9). For a tapered end 1 m wide, if two of the four pixels are considered to be in shadow, then the crest at this location is b14 cm high. Such narrow tails are exceedingly difficult to explain as original depositional features associated with the emplacement of the main TAR. Predictions that sand saltation path lengths might be one or more meters under current conditions on Mars (White, 1979; Almedia et al., 2008) make it difficult to envision how blowing sand could be deposited while confined onto such narrow features, particularly for winds transverse to the crest orientation. If the last winds were parallel to the crest orientation, perhaps the feature could have extended along the crest direction. Terrestrial longitudinal dunes are thought to involve winds at slight angles to either side of the crest orientation (Tsoar, 1983; Tsoar et al., 2004), which might be expected to preserve some visible sinuosity of the crest line or regular variation in the crest height, neither of which are apparent in the HiRISE image. The discreet linear segments aligned with the tapered end of one TAR (upper left of Fig. 9) represent an even tighter constraint against the dominant wind being along the crestline orientation. If sand was being shed from the tapered tip, sand accumulations downwind should develop into barchanoid shapes pointing toward the tip rather than the observed linear axis parallel to the crest at the tip. All of the above considerations lead to a favored interpretation that the tapered extensions of the TARs represent erosional remnants of once more extensive TARs, rather than depositional features resulting from winds parallel to the TAR crest. The textures on and around the Terra Sirenum TARs, revealed by the shallow illumination conditions, suggest the presence of second-order ripples (e.g., Bridges et al., 2007). In particular, features between the

28

J.R. Zimbelman / Geomorphology 121 (2010) 22–29

TARs (large arrows in Fig. 9) display linear trends traceable for tens of meters, but without the gentle crenulations prevalent on the sand ripples superposed on the Gamboa sand patches (Fig. 5). Some of the textural patterns continue from the inter-TAR plains onto the sunlit sides of some TARs (Fig.10). The Terra Sirenum textures seem more consistent with second-order ripples (Bridges et al., 2007) than to possible exposures of eroded layers. Terra Sirenum is within the latitudinal band of a ‘mid-latitude mantle’ identified in MOC images, interpreted to incorporate volatiles within the mantling unit (Mustard et al., 2001; Kreslavsky and Head, 2002). Perhaps the Terra Sirenum image may reveal layering that is part of the composite mantling layer observed in MOC images. However, it is important to note that no layer-like texture is evident in the Gamboa crater image, which is from a similar latitude in the northern hemisphere. Study of many HiRISE images from both hemispheres is needed to investigate further the possible causes of the subtle textures seen in Figs. 9 and 10. Whatever their cause, the textures are clearly superposed on the TARs and thus do not appear to be intimately involved with the original emplacement of the TAR materials. 6. Discussion HiRISE images are unable to resolve individual particles out of which TARs are comprised, but they are of sufficient resolution to improve an evaluation of hypotheses of formation (e.g., Zimbelman, 2001) for these intriguing features. The primary hypotheses of origin proposed to explain TARs are either as small sand dunes or as large ripples (Bourke et al., 2003; Wilson and Zimbelman, 2004; Balme et al., 2008). Results from the three HiRISE images presented here are supportive of both origins, to differing degrees. A dune origin is most consistent with the remarkable symmetry of TAR profiles (Fig. 3), and also for TARs of height ≥1 m (all entries in Table 1); a comparison of TARs with reversing transverse dunes is particularly striking (Fig. 4). Slip faces have not yet been observed on any TARs, consistent with an upper limit of ~24° for the near-crest TAR slopes (Fig. 3), and with the overall average slope ≤28° (Table 1). Arguing against a dune origin, classic sand ripples like those that are ubiquitous on the sand patches in Gamboa crater (Fig. 5) are not observed on TARs. However, if the Terra Sirenum textures (Figs. 9 and 10) are indeed second-order ripples, they may be long linear sand ripples perhaps similar to those studied at El Dorado by the Spirit rover (Sullivan et al., 2008). A granule ripple origin is most consistent with TAR-like features of a size represented by heights ≤0.5 m, where their profiles (Fig. 7) are very similar to both the shape and the aspect ratio of granule ripples in the western United States (Fig. 8; Zimbelman and Williams, 2005, 2006). An inferred surface coating of granules is also consistent with MER observations of granules on most ripples larger than ~ 10 to 20 cm in height (Greeley et al., 2004; Sullivan et al., 2005, 2008). Ripples with heights b10 cm are most likely normal sand ripples (i.e., the ripples on the sand patch in Fig. 6), although small ripples comprised of coarse (relative to normal sand) particles cannot be ruled out. Transverse bedforms with heights of ~0.5 m to ~1 m still may fall into either small dune or large granule ripple categories; this height range for transverse features should be explored more fully during future studies. 7. Conclusions 1) TARs from five locations in a HiRISE image of the floor of Ius Chasma show remarkably symmetric profiles, with average slopes for the entire TAR feature of ~15°. 2) A HiRISE image of Gamboa crater shows low albedo sand patches b2 m high and N40 m wide that are covered with normal sand ripples (≤5 cm height), surrounded by larger ripples (0.2 to 0.5 m height) that are very similar in cross-sectional profile to granule ripples on Earth.

3) TARs in a HiRISE image from Terra Sirenum are comparable to those in Ius Chasma, but with average slopes grouped around ~ 16° and ~ 25°. Many of the Terra Sirenum TARs have tapered extensions to the northeast that are more consistent with being erosional remnants of formerly larger TARs rather than the result of extension of the feature by deposition from the tapered end. 4) The new observations generally support a reversing transverse dune origin for TARs with heights ≥1 m, and a granule ripple origin for TAR-like features with heights ≤0.5 m. Acknowledgements Comments by two anonymous reviewers were very helpful during the revision of the manuscript. This work was supported by grants NNG04GN88G and NNX08AK90G to JRZ from the Mars Data Analysis Program of NASA. References Almedia, M.P., Parteli, E.J.R., Andrade, J.S., Hermann, H.J., 2008. Giant saltation on Mars. Proc. Nat. Acad. Sci. 105 (17), 6222–6226. doi:10.1073/pnas.0800202105. Anderson, R.S., 1987. A theoretical model for aeolian impact ripples. Sed. 34, 943–956. Anderson, R.S., Haff, P.K., 1988. Simulation of eolian saltation. Science 241, 820–823. Bagnold, R.A., 1941. The Physics of Blown Sand and Desert Dunes. Chapman and Hall, London. Balme, M., Berman, D.C., Bourke, M.C., Zimbelman, J., 2008. Transverse aeolian ridges (TARs) on Mars. Geomorph. 101, 703–720. doi:10.1016/j.geomorph.2008.03.011. Bourke, M.C., Wilson, S.A., Zimbelman, J.R., 2003. The variability of transverse aeolian ridges in troughs on Mars. Lunar and Planetary Science XXXIV, Lunar and Planetary Institute, Houston (CD-ROM), Abstract 2090. Bourke, M.C., Balme, M., Beyer, R.A., Williams, K.K., Zimbelman, J., 2006. A comparison of methods used to constrain the height of sand dunes on Mars. Geomorph 81 (3–4), 440–452. doi:10.1016/j.geomorph.2006.04.023. Bridges, N.T., Geissler, P.E., McEwen, A.S., Thomson, B.J., Chuang, F.C., Herkenhoff, K.E., Keszthelyi, L.P., Martinez-Alonso, S., 2007. Windy Mars: a dynamic planet as seen by the HiRISE camera. Geophys. Res. Lett. 34, L23205. doi:10.1029/ 2007GL031445. Breed, C.S., Grolier, M.J., McCauley, J.F., 1979. Morphology and distribution of common ‘sand’ dunes on Mars: comparison with the Earth. J. Geophys. Res. 84, 8183–8204. Burkinshaw, J.R., Illenberger, W.K., Rust, I.C., 1993. Wind-speed profiles over a reversing transverse dune. Geol. Soc. London Sp. Pub. 72, 25–36. Christensen, P.R., et al., 2003. Morphology and composition of the surface of Mars: Mars Odyssey THEMIS results. Science 300, 2056–2061. Cutts, J.A., Smith, R.S.U., 1973. Aeolian deposits and dunes on Mars. J. Geophys. Res. 78, 4139–4154. Cutts, J.A., Blasius, K.R., Briggs, G.A., Carr, M.H., Greeley, R., Masursky, H., 1976. North polar region of Mars: imaging results from Viking 2. Science 194, 1329–1337. Fenton, L.K., 2005. Potential sand sources for the dune fields in Noachis Terra. Mars. J. Geophys. Res. 110, E11004. doi:10.1029/2005JE002436. Fenton, L.K., Mellon, M.T., 2006. Thermal properties of sand from Thermal Emission Spectrometer (TES) and Thermal Emission Imaging System (THEMIS): spatial variations within the Proctor crater dune field on Mars. J. Geophys. Res. 111, E06014. doi:10.1029/2004JE002363. Frank, A.J., Kocurek, G., 1996. Airflow up the stoss slope of sand dunes: limitations of current understanding. Geomorph. 17 (1–3), 47–54. Greeley, R., Iversen, J.D., 1985. Wind as a Geological Process on Earth, Mars, Venus and Titan. Cambridge Univ. Press, New York. Greeley, R., Lancaster, N., Lee, S., Thomas, P., 1992. Martian aeolian processes, sediments, and features. In: Kieffer, H.H., Jakosky, B.M., Snyder, C.W., Matthews, M.S. (Eds.), Mars. Univ. of Arizona Press, Tucson, pp. 730–766. Greeley, R., et al., 2004. Wind-related processes detected by the Spirit Rover at Gusev crater, Mars. Science 305, 810–821. Greeley, R., et al., 2006. Gusev crater: wind-related features and processes observed by the Mars Exploration Rover Spirit. J. Geophys. Res. 111, E02S09. doi:10.1029/ 2005JE002491. Herkenhoff, K.E., et al., 2004. Textures of the soils and rocks at Gusev crater from Spirit's Microscopic Imager. Science 305, 824–826. Kirk, R.L., Howington-Kraus, E., Rosiek, M.R., Cook, D., Anderson, J., Becker, K., Archinal, B.A., Keszthelyi, L., King, R., McEwen, A.S., HiRISE Team, 2007. Ultrahigh resolution topographic mapping of Mars with HiRISE stereo images: methods and first results. Lunar and Planetary Science XXXVIII, Lunar and Planetary Institute, Houston (CD-ROM), Abstract 1428. Kreslavsky, M.A., Head, J.W., 2002. Mars: nature and evolution of young latitude-dependent water-ice-rich mantle. Geophys. Res. Lett. 29 (15). doi:10.1029/2002GL015392. Lancaster, N., Greeley, R., 1990. Sediment volume in the north polar sand seas of Mars. J. Geophys. Res. 95, 10921–10927. Malin, M.C., Edgett, K.S., 2001. Mars Global Surveyor Mars Orbiter Camera: interplanetary cruise through primary mission. J. Geophys. Res. 106, 23429–23570. McCauley, J.F., Carr, M.H., Cutts, J.A., Hartmann, W.K., Masursky, H., Milton, D.J., Sharp, R.P., Wilhelms, D.E., 1972. Preliminary Mariner 9 report on the geology of Mars. Icarus 17, 289–327.

J.R. Zimbelman / Geomorphology 121 (2010) 22–29 McEwen, A.S., et al., 2007. Mars Reconnaissance Orbiter's High Resolution Imaging Science Experiment (HiRISE). J. Geophys. Res. 112, E05S02. doi:10.1029/2005JE002605. McKenna Neuman, C., Lancaster, N., Nickling, W.G., 1997. Relations between dune morphology, air flow, and sediment flux on reversing dunes, Silver Peak, Nevada. Sediment. 44 (6), 1103–1111. Momiji, H., Carretero-Gonzalez, R., Bishop, S.R., Warren, A., 2000. Simulation of the effect of wind speedup in the formation of transverse dune fields. Earth Surf. Process. Landforms 25, 905–918. 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. Pye, K., Tsoar, H., 2009. Aeolian Sand and Sand Dunes. Springer, New York. Sharp, R.P., 1963. Wind ripples. J. Geology 71, 617–636. Shockey, K.M., Zimbelman, J.R., 2007. Sand dunes across the dichotomy boundary of Mars. Lunar and Planetary Science XXXVIII, Lunar and Planetary Institute, Houston (CD-ROM), Abstract 1328. Squyres, S.W., et al., 2004a. The Spirit Rover's Athena science investigations at Gusev crater. Mars. Science 305, 794–799. Squyres, S.W., et al., 2004b. The Opportunity Rover's Athena science investigation at Meridiani Planum, Mars. Science 306, 1698–1703. Sullivan, R., et al., 2005. Aeolian processes at the Mars Exploration Rover Meridiani Planum landing site. Nature 436. doi:10.1038/nature03641. Sullivan, R., Arvidson, R., Bell, J.F., Gellert, R., Golombek, M., Guinness, E., Greeley, R., Herkenhoff, K., Johnson, J., Squyres, S., Thompson, S., Whelley, P., Wray, J., 2008. Wind-driven particle mobility on Mars: insights form MER observations at “El Dorado” and surroundings at Gusev crater. J. Geophys. Res. (Planets) 113, E06S07. doi:10.1029/2008JE003101. Tsoar, H., 1983. Dynamic processes acting on a longitudinal (seif) sand dune. Sed. 30, 567–578. Tsoar, H., Greeley, R., Peterfreund, A.R., 1979. Mars: the north polar sand sea and related wind patterns. J. Geophys. Res. 84, 8167–8180. Tsoar, H., Blumberg, D.G., Stoler, Y., 2004. Elongation and migration of sand dunes. Geomorph 57, 293–302. doi:10.1016/SO169-555X(03)00161-2. Unger, J., Haff, P.K., 1987. Steady state saltation in air. Sediment. 34, 289–299. Walker, I.J., 1999. Secondary airflow and sediment transport in the lee of a reversing dune. Earth Surf. Proc. Landforms 24 (5), 437–448. White, B., 1979. Soil transport by winds on Mars. J. Geophys. Res. 84 (B9), 4643–4651.

29

Wilson, I., 1972a. Sand waves. New Scientist 53, 634–637. Wilson, I.G., 1972b. Aeolian bedforms: their development and origins. Sed. 19, 173–210. Wilson, S.A., Zimbelman, J.R., 2004. The latitude-dependent nature and physical characteristics of transverse aeolian ridges on Mars. J. Geophys. Res. 109, E10003. doi:10.1029/2004JE002247. Yizhaq, H., 2005. A mathematical model for aeolian megaripples on Mars. Physica A, 357, 57–63. Yizhaq, H., Balmforth, N.J., Provenzale, A., 2004. Blown by the wind: nonlinear dynamics of aeolian sand ripples. Physica D, 195, 207–228. Zimbelman, J.R., 1987. Spatial resolution and the geologic interpretation of Martian morphology: implications for subsurface volatiles. Icarus 71, 257–267. Zimbelman, J.R., 2001. Image resolution and evaluation of genetic hypotheses for planetary landscapes. Geomorph. 37, 179–199. Zimbelman, J.R., 2008. Transverse aeolian ridges on Mars: results obtained from analysis of HiRISE images. Planetary Dunes Workshop: A Record of Climate Change. LPI Contribution No. 1403. Lunar and Planetary Institute, Houston, pp. 83–84. Zimbelman, J.R., Williams, S.H., 2005. From ripples to dunes: transitions in bedform profile across aeolian deposits, with implications for Mars. Trans. Am. Geophys. Union 86 (18) Spring Meeting Suppl., Abstract P32A-04. Zimbelman, J.R., Williams, S.H., 2006. Aeolian ripples on Earth and Mars: scale diversity and implications for modes of particulate transport. Lunar and Planetary Science XXXVII, Lunar and Planetary Institute, Houston (CD-ROM), Abstract 2047. Zimbelman, J.R., Williams, S.H., 2007a. Dunes versus ripples: topographic profiling across terrestrial examples, with application to the interpretation of features on Mars. Trans. Am. Geophys. Union 88 (23) Joint Assembly Suppl., Abstract P34A-07. Zimbelman, J.R., Williams, S.H., 2007b. An evaluation of formation processes for transverse aeolian ridges on Mars. Seventh International Conference on Mars, Lunar and Planetary Institute, Houston (CD-ROM), Abstract 3047. Zimbelman, J.R., Williams, S.H., 2007c. Comparison of transverse aeolian ridges on Mars to ripples and dunes on Earth. Geol. Soc. Am. Abs. Prog. 39 (6), 218-5 Abstract. Zimbelman, J.R., Williams, S.H., 2008. Inferences about sand dunes on Mars derived from the analysis of two HiRISE images. Lunar and Planetary Science XXXIX, Lunar and Planetary Institute, Houston (CD-ROM), Abstract 1699. Zimbelman, J.R., Irwin, R.P., Williams, S.H., Bunch, F., Valdez, A., Stevens, S., 2009. The rate of granule ripple movement on Earth and Mars. Icarus 203, 71–76.