du iti elm ted S , TX s 8238 Revised 10 June 2009 Accepted 16 June 2009 Available online 7 July 2009 regions of Mars. Ubiquitous yardangs are clearly the product of strong unidirectional winds acting over source or signi?cance (Bradley and Sakimoto, 2001; Zimbelman and Patel, 1998). As part of a comprehensive survey of 713 images of the MFF from the Mars Orbiter Camera (MOC) of the Mars Global Surveyor (MGS), we have inferred material properties that suggest that the lithology of the MFF is largely that of an ignimbrite (Mandt et al., 2008). Herein we focus on the development and evolution of yard- 2. The MFF Many geologic formations on Mars have origins that have yet to be agreed upon. One of the most prominent of these is the MFF, a deposit located along the equator stretching between 240 and 170E Longitude. It is commonly described as enigmatic because its origin has been the subject of debate for decades. The MFF is lo- cated in the Amazonis Planitia region and lies between two major volcanic centers: Tharsis and Elysium (Fig. 1). In all places where they are in contact, the southern portion of the MFF overlies the * Corresponding author. Address: Space Science and Engineering Department, Southwest Research Institute, San Antonio, TX 78238, United States. Icarus 204 (2009) 471?477 Contents lists availab ru .e lE-mail address: kathymandt@yahoo.com (K. Mandt).Erosional remnants on the surfaces of planetary bodies are valu- able as a record of past and current climatic and geologic processes on the planet. Among the most obvious features on the surface of Mars are erosional remnants like yardangs and less-than-kilome- ter-scale mesas. They provide insight into erosional processes and material properties of the formations on which they are found, such as the abundant yardangs in the Medusae Fossae Formation (MFF) that have been attributed to the mildly indurated nature of the MFF lithology and the presence of strong unidirectional winds (Schultz and Lutz, 1988; Scott and Tanaka, 1982; Mandt et al., 2008). To date, less-than-kilometer-scale mesas have been noted by very few authors, with little insight provided as to their produced by a sequence of erosional stages that suggest differing processes for formation that are distinct from each other. The mechanism for forming yardangs is well understood (e.g. Breed et al., 1979) and their development within the MFF provides indi- cations of local climatic conditions and material properties within the MFF. The mechanism for forming the mesas is a strong indica- tor of material properties and shows a lack of the strong unidirec- tional winds that formed the yardangs. Both forms require a lithology that is indurated in its upper parts and more friable in its lower portions. The separation of areas dominated by these respective forms is on the scale of 50?350 km, suggesting contrast- ing local environmental conditions at this spatial scale.Keywords: Mars, Surface Mars, Climate 1. Introduction0019-1035/$ - see front matter  2009 Elsevier Inc. A doi:10.1016/j.icarus.2009.06.031time on variably indurated deposits. Yardang orientation is used as a proxy to map regional and local wind direction at meso-scale resolution. In other, more limited areas not subjected to strong unidirec- tional winds, randomly oriented kilometer-scale mesas and buttes are found to be remnants of progres- sive cliff recession through mass wasting as support is lost from within the MFF lithology at its margins. The differing processes that dominate the formation of the distinctive landforms have implications for meso-scale variations in climate that remain unresolved by current modeling efforts. Additionally, the variability of erosional forms within the deposit emphasizes the overall complexity of this extensive formation.  2009 Elsevier Inc. All rights reserved. angs and less-than-kilometer-scale mesas and show that both areArticle history: Received 17 January 2009 The form of erosional remnants of the Medusae Fossae Formation (MFF) on Mars provide evidence of their development progression and implicate two spatially distinct environments in the equatorialDistinct erosional progressions in the Me indicate contrasting environmental cond Kathleen Mandt a,b,*, Shanaka de Silva c, James Zimb aUniversity of North Dakota, Department of Space Studies, Grand Forks, ND 58202, Uni b Space Science and Engineering Department, Southwest Research Institute, San Antonio cDepartment of Geosciences, Oregon State University, Corvallis, OR 97331, United State dCEPS/NASM MRC 315, Smithsonian Institution, Washington, DC 20013, United States eGeosciences and Engineering Division, Southwest Research Institute, San Antonio, TX 7 a r t i c l e i n f o a b s t r a c t Ica journal homepage: wwwll rights reserved.sae Fossae Formation, Mars, ons an d, Danielle Wyrick e tates 78238, United States , United Statesle at ScienceDirect s sevier .com/locate / icarus examples are found with aspect ratios as high as 50:1 in the Alti- (bl ies us 2dichotomy boundary (Sakimoto et al., 1999, Bradley et al., 2002): a great circle inclined 28 to the equator that divides the northern lowlands from the southern, cratered highlands (Greeley and Guest, 1987). The MFF is considered to be one of the youngest deposits on Mars based on stratigraphic relationships (Scott and Tanaka, 1986; Greeley and Guest, 1987; Head and Kreslavsky, 2004; Hynek et al., 2003). A wide variety of hypotheses have been proposed for the geo- logic origin of the MFF: ignimbrites (Scott and Tanaka, 1982; Mandt et al., 2008), aeolian deposits (Greeley and Guest, 1987), paleopolar deposits (Schultz and Lutz, 1988), exhumed fault rocks (Forsythe and Zimbelman, 1988), carbonate platform (Parker, 1991), rafted pumice deposits (Mouginis-Mark, 1993), lacustrine deposits (Malin and Edgett, 2000), ash fall tuff (Tanaka, 2000; Brad- ley et al., 2002; Hynek et al., 2003), and deposits associated with a subsurface aquifer (Nussbaumer, 2005). Most of these origins have been challenged by later work and only the ash fall tuff, aeolian (loess), and the ignimbrite hypotheses remain most viable after examination of recent datasets. Based on the material properties evident in the formation, we have argued that an ignimbrite origin is the most likely of the three remaining viable hypotheses (Mandt et al., 2008). In addition to providing information about the nature Fig. 1. MOLA shaded relief with map showing locations of images with yardangs boundaries based on Scott and Tanaka (1986) and Greeley and Guest (1987). Boundar region is on the far right side of the map.472 K. Mandt et al. / Icarof the MFF lithology, the observations documented in this paper outline evidence for the development of two prominent erosional remnant forms of the MFF and the dissimilar environmental condi- tions required to form both of them. 3. Methods The MOC Narrow Angle images we used had a typical resolution of one to three meters per pixel (e.g., Malin and Edgett, 2001). Though newer missions are providing higher resolution images, MOC was chosen for the present study due to the extent of its cov- erage of the MFF, which allows a synoptic analysis of the forma- tion. This approach is important because this formation spans more than one-fourth of the equatorial circumference of the planet and thus could have considerable variability at the scale of high resolution images. The extent of coverage by MOC images allowed us to produce maps of the yardangs and kilometer-scale mesas in order to evaluate the formation as a whole. Each of the images was studied individually and within the con- text of its location on a Mars Odyssey Thermal Emission Imaging System (THEMIS) mosaic (e.g., Christensen et al., 2004). Maps were created identifying the locations of images showing mesas and yardangs within the MFF to highlight their extent throughout theplano-Puna of the Central Andes. The degree of induration and the vertical induration pro?le plays a major role in aspect ratio, and properties such as jointing and layers with different degrees of induration can also alter the morphology of a yardang (Mandt et al., 2008).formation (Fig. 1), and the inferred wind direction based on yard- ang orientation (Fig. 2a). We then examined the 541 images asso- ciated with yardangs and 96 images associated with collapse features and mesas to develop models for their formation. 4. Yardangs Yardangs are the most commonly observed feature in the MFF; they form ?elds covering very large areas. Yardangs are linear aerodynamic forms formed by aeolian erosion, and they are best developed in arid regions (Breed et al., 1979). They can be as large as many kilometers in length and 100?200 m high (Ward, 1979). Terrestrial yardangs formed by a strong unidirectional wind erod- ing lithologically consistent material are ??ideal? in form when they are well-streamlined, and typically have an aspect ratio (length to width) of about 3:1 (McCauley et al., 1997), but terrestrial ack) and collapse features and mesas (white). Upper, middle and lower member of map stretch from 120 to 230W Longitude and 15S to 15N Latitude. The Tharsis 04 (2009) 471?477We have found that there is an erosional progression in the development of yardangs from initiation as v-shaped depressions to proto-yardangs and then fully ?edged yardangs. The v-shaped depressions, interpreted to be de?ation hollows caused by the re- moval of less resistant material (Bradley et al., 2002), are observed in areas dominated by yardangs. The depth of these depressions ranges from 10?50 m and a progression from the depressions to the yardangs can be seen as illustrated in Figs. 3 and 4. The yard- angs in the MFF appear to be initiated through de?ation/excava- tion that exposes a resistant block of material that is subsequently sculpted into an aerodynamic landform in a well understood process (e.g. Breed et al., 1979). Although Bradley et al. (2002) conducted a statistical analysis of yardang direction with the goal of evaluating the lithology of MFF material and concluded that wind direction was not the sole deter- minant of yardang orientation, this is in contrast to other studies on Mars that show that yardangs re?ect the orientation of strongly unidirectional winds (e.g., Ward, 1979; Mandt et al., 2008). More- over, studies of terrestrial yardangs have shown that yardang ori- entation strongly re?ects wind direction (Bailey et al., 2007; de Silva et al., 2009). We therefore believe that the orientation of the yardangs can be used as a proxy for the dominant wind direc- tion that formed them. Our analysis thus presents the ?rst high Fig. 2. MOLA topography overlain on Viking mosaic of the MFF. (a) Overview of entire formation. Black lines indicate yardang orientation that was used as a proxy for wind direction. Yellow arrows suggest dominant wind direction and white boxes outline the enlarged portions below. (b) Enlarged view of the western portion of the MFF. The yardang orientation shows an exposed layer that provides evidence of changing climatic conditions in this region over time. (c) Central portion of the MFF showing two distinct environmental regimes. The lower portion has abundant yardangs formed by strong unidirectional winds while the outlined ??sheltered? portion shows no evidence of the unidirectional winds. (d and e) Enlarged view of the east-central portion of the MFF showing areas where topography has clearly controlled the direction of the wind. (f) Enlarged view of the region north of the Gordii Dorsum showing the transition region, area 2, located between a region of yardangs subjected to strong unidirectional winds, area 1, and a region with collapse features and the resulting mesas and buttes, area 3. Fig. 3. Erosional progression from v-shaped depressions to yardangs. All images are illuminated from the bottom left and north is up. (a) MOC E10-00390, centered at 156W and 1N, showing a single v-shaped depression. (b) MOC M07-00371 centered at 163W and 4.5N, showing a cluster of v-shaped depressions. (c) MOC M21-01904 centered at 148W and 0.5N, showing yardangs with remainders of v-shaped depressions at the tip. (d and e) MOC E03-01400 centered at 173W and 2.5S, showing fully formed curvilinear yardangs. Image (d) shows the remainder of an elongated v-shaped depression. K. Mandt et al. / Icarus 204 (2009) 471?477 473 474 K. Mandt et al. / Icarus 204 (2009) 471?477resolution graphical presentation of wind direction in the MFF re- gion. In the areas with v-shaped depressions, the prominent wind direction is inward from the point of the v. Fully-developed yard- angs form a linear shape in the prominent wind direction, are wider on the side facing the wind and narrower in the tail. The pri- mary wind direction in the area of the MFF differs from one section of the formation to another (Fig. 2a). The far western portion (westward of 200W Longitude) has been eroded by a predomi- nantly east?west wind. In Fig. 2b, the majority of the yardangs are oriented in the ENE?WSW direction. Some yardangs are ori- ented perpendicular to this direction and are at a lower elevation. This suggests erosional exposure of an underlying layer with yard- angs that formed prior to deposition of the layer overlying these Fig. 4. THEMIS images (Christensen et al., 2004) showing yardang erosional progression a infrared image I10241007, centered at 204W and 3.6N, showing progression from v centered at 201W and 11.2N, showing erosional progression from feathered scarps (b Fig. 5. Erosional progression from troughs to mesas. All images are illuminated from t showing chains of pits formed by probable subsurface collapse. (b) MOC E05-02396, cent deeper trenches. (c) MOC E10-01357, centered at 147W and 13N, showing the developm 02348, centered at 213W and 3.3N, showing a group of mesas resulting from this prog buttes at the ?nal stages of this erosional progression.yardangs. It is clear from this that the wind direction when the old- er yardangs formed was perpendicular to the most recent wind re- gime. The central portions (between 150 and 180W Longitude) of the MFF have been eroded by a wind that appears to be coming off of the dichotomy boundary while the far eastern portion (eastward of 150W Longitude) has been exposed to winds coming off of the Tharsis rise. Evidence of topographic in?uence on the wind is clearly seen (Fig. 2b, d and e). 5. Collapse features and mesas Chains of pits and trenches, referred to as collapse features, with size scales ranging from one to tens of kilometers are ob- t larger scale. Illumination is from the lower left in both images. (a) THEMIS daytime -shaped depressions to curvilinear yardangs and (b) THEMIS visible V01054003, ottom of image) to curvilinear yardangs. he bottom left and north is up. (a) MOC E03-01084, centered at 181W and 15N, ered at 182W and 2.3N, showing similar collapse features that have progressed to ent of individual mesas as the deepening of the trenches progresses. (d) MOC E04- ression. (e) MOC E05-01356, centered at 213W and 3.3N, showing heavily eroded served at the edges of the MFF, where the apparent thickness is less than 250 m, assuming the base material below the entire forma- tion is relatively ?at (Watters et al., 2007; Carter et al., 2009). At the very edges of the deposit, less-than-kilometer scale erosional mesas coincide with areas where the collapse features are seen. MOC images (Figs. 5 and 6) show a clear progression from collapse features to mesas. The evolution appears to begin with pit chains and trenches forming by collapse. The collapsed material is, as a re- The formation of yardangs is well understood, and our observa- K. Mandt et al. / Icarus 2Fig. 6. MOC E11-02357, centered at 211W and 2.4N, showing full progression from collapse features to mesas. Blocks at the base of the buttes suggest an indurated layer overlying a more friable base layer.tions of the yardangs of the MFF are consistent with yardang for- mation in differentially indurated lithologies on Earth (de Silva et al., 2009). Where a well-indurated capping layer is present, yardang formation proceeds at local anomalies that expose the less indurated layer. Knobs and reentrants can expose the lower softer layers that subsequently get preferentially excavated by the wind, armed with saltating particles. As the depression develops in the softer material, excavation focuses within the rim of the growing depression, creating local eddies and turbulence (Breed et al., 1979). As the indurated upper layer forms a resistant caprock, the wind becomes bifurcated, producing v-shaped depressions that result in the development of the yardang form. Where an indu- rated capping layer is present, mega-yardangs (Goudie, 2007) of extreme aspect ratio can develop, aided by sculpting, through block collapse of the upper indurated parts while poorly indurated lithologies produce smaller, less well-de?ned yardangs (de Silva et al., 2009). Strong unidirectional winds and a lithology with a vertical induration pro?le are the primary factors that result in the mega-yardang formation in the MFF. The mesas and buttes represent different phenomena, though the lithology of the material in which the collapse features, mesas and buttes form resembles that in which the yardangs form: ?ne- grained with evidence of upper layers of caprock in many areas. Collapse features similar to the ones found in the MFF have been observed at much larger scale elsewhere on Mars, and have been interpreted to be the result of dilational faulting and fracturing (Wyrick et al., 2004). In the case of the MFF collapse features, dila- tional faulting and fracturing could possibly account for the origi- nal collapse features. The fact that they are only found at thesult, of a less cohesive nature and more susceptible to aeolian ero- sion and mass wasting. Over time these features become deeper and wider, eventually leaving only the mesas. Images of this pro- gression suggest that the top layer of material is more indurated than the material below it, as evidenced by the way the polygonal forms remain during the removal of material and appear to tilt (as shown in Fig. 6). Blocks at the bases of many mesas in the highest resolution images also support a more lithi?ed top layer as the more friable base layer is removed. Recent modi?cation of ?ne material is evident in the dunes between, and on top of, the mesas suggesting that aeolian processes may play a recent role in local erosion of materials. The regions dominated by collapse features and mesas are spa- tially separated from yardangs by 50?350 km (Figs. 1 and 2c), although some transition zones between the two regions can be seen in Viking mosaics (Fig. 2f). Regions showing collapse features and mesas are clearly not exposed to a strong unidirectional wind because such a mechanism would shape the mesas into yardangs and erase much of the evidence of collapse. A study of the lower member of the MFF, with a speci?c focus on a region containing a large number of collapse features and lacking yardangs (McColley et al., 2005), interpret the collapse fea- tures to be troughs created by subsurface collapse due to the re- lease of volatiles. This region is mapped as the ?sheltered region? outlined in Figs. 2b and 1 shows that these features cover a very extensive area. Collapse features are seen all over the surface of Mars at varying scales and shapes. The larger-scale collapse fea- tures have been shown to form by collapse due to an underlying fault or fracture system (Wyrick et al., 2004). 6. Discussion 04 (2009) 471?477 475margins of the deposit may mean that the fracture systems are only exposed near the surface at locations where the thickness is least, and therefore become susceptible to further mass wasting us 2and erosion. We suggest that the collapse is driven by the removal of support from within the lithology, along the margins of the deposit. The association of the collapse features and mesas with cliff recession is striking. The mesas and buttes are clearly remnants of larger forms that are left behind as the cliff recedes (Fig. 6), and this removal of material is facilitated by the formation of the pit chains. We propose that this has taken place due to loss of material, and hence support, from within the deposit near their margins, rather like sapping of groundwater in terrestrial deposits, only in this case it is a material property inherent to the MFF. These collapse features have been suggested to be the result of release of volatiles (McColley et al., 2005), and if the MFF is predominantly ignimbrite (Scott and Tanaka 1982, 1986; Mandt et al., 2008), vol- atiles could have been trapped during the formation process and later released. However, volatiles in ignimbrites typically escape upwards. These volatiles have two sources; magmatic volatiles contained in the juvenile (magmatic) material usually pumice, and external gas (trapped and ingested ambient air, gas from com- bustion of vegetation, water) that may be incorporated into the ?ows as it moves. Observations of historical eruptions in 1902 Mt. Pel?e, Martinique, in 1980 at Mt. St. Helens, Washington, in 1991 at Mt. Pinatubo in the Phillipines, and in 1992 at Volcan Las- car, Chile, emphasize the presence of massive ash clouds above the moving pyroclastic ?ows. These observations con?rm sedimento- logical data from ancient deposits that indicate that gas loss is con- tinuous during the ?ow process (Sparks, 1978; Wilson, 1984). Juvenile gas is released as pumice is broken and crushed during ?ow and external gas is streamed upwards through the ?ow. How- ever, post depositional gas loss may actually account for most of the gas loss from ignimbrites; this is enhanced if an external gas source like surface water is present (e.g. Keating, 2005). This is evi- denced by vertical gas escape structures and vapour phase alter- ation pro?les in ancient deposits, and active degassing of the upper surfaces of historic ignimbrites such as those described above. Active degassing months to years after deposition has been observed, while qualitative and quantitative thermal and chemical considerations suggest time scales of at least decades to maybe centuries for complete cooling and degassing of thick (many tens of meters) (e.g. Ross and Smith, 1961; Sheridan and Ragan, 1976; Riehle et al., 1995). The pathways for gas loss are usually strongly indurated and recrystallized, due to vapor phase devitri?cation and crystallization, and sintering in upper zones ignimbrites (Cas and Wright, 1992; de Silva, 1989; de Silva et al., 2009). These pathways would not be weak areas along which pits could develop; therefore a different mass loss process to post depositional degassing of ignimbrite is required. Recent Mars Advanced Radar for Subsurface and Ionosphere Sounding (MARSIS) data consistent with either an ice-rich or very porous deposit (Watters et al., 2007), raising the possibility that ice is subliming at the exposed front of the MFF, although analysis of the Shallow Radar (SHARAD) does not interpret the data to show strong evidence of ice within MFF (Carter et al., 2009). If ice were to sublime, a volume de?cit could be created and result in forma- tion of collapse features and subsequent cliff recession. Such a sce- nario is an attractive possibility, potentially supporting the paleopolar deposit origin of the MFF (Schultz and Lutz, 1988). However, laboratory results of the dielectric properties presently are not able to distinguish between a ?ne-grained dry deposit and a porous material where the void spaces are ?lled with water ice (Watters et al., 2007; Carter et al., 2009), so that at present the radar sounding results are unable to provide direct evidence of either the presence or the absence of ice from within MFF. In any 476 K. Mandt et al. / Icarcase, both volatile release and ice sublimation would likely have produced a more ubiquitous pattern of pit distribution rather than the polygonal fractured patterns as indicated by the pit chains andtroughs. It is possible that volatile release or ice sublimation could have been preferentially emplaced and transported through high- er-permeability faults and fractures, but it is not required for the polygonal pit chain and trough formations. The loss of support from within the MFF to produce the collapse features at the edge of the formation remains problematic. Because ice sublimation and volatile loss do not provide good explanations, we are unsure of what material is being lost out of the margins of the MFF to produce the collapse structures, mesas and buttes. Nonetheless, because the mesas and buttes form in material with similar lithology to the yardang material but lack the aerodynamic form of the yardangs shows that there is not a strong unidirec- tional wind in the areas where the mesas and buttes form. This suggests local climate variability within the region where the MFF is found. Current models of circulation on Mars show generally west- ward winds for this region, northwest in the northern spring and summer and southwest in the northern summer (Zurek, 1992; Fen- ton, 2003). Our observations of the yardangs along with previously documented variations in yardang orientation (Bradley et al., 2002) show spatial variations that suggest a more complex surface pat- tern (Fig. 2a), and long-term variation in primary wind direction between one episode of deposition and the next (Fig. 2b). Topogra- phy clearly plays an important role in near-surface winds (Fig. 2d and e). We note that some workers (Anderson et al., 1999) have at- tempted to look at erosion vs. deposition on the surface of Mars with a 25  40 km resolution wind model, but found no correlation to speci?c geologic features. Our observations of the difference be- tween mesas and yardangs in the MFF suggest that local environ- mental conditions vary on the spatial scale of 50?350 km in the equatorial region of Mars (Fig. 2f). The differing erosional processes also have important implica- tions for analysis and mapping of the MFF. This formation was originally mapped into three members of one formation (Scott and Tanaka, 1986; Greeley and Guest, 1987), but is typically ad- dressed as a single entity (e.g. Parker, 1991). Across its expanse of over 5000 km there exist enormous differences in the overall thickness of the deposits, and in their degree of degradation show- ing that the MFF is not a homogenous unit (Mandt et al., 2008). The variations in erosional style documented here further emphasize how diverse MFF is. This emphasizes the importance in recognizing the MFF for what its name implies, a ??formation? ? a lithostrati- graphic division consisting of two or more members, each of which consists of two or more beds (Barnes and Lisle, 2004, p. 94). Forma- tions consist of a variety of materials that have contributed to their ?nal composite state. 7. Conclusions Mesas and yardangs in the MFF on Mars are the ?nal result of progressive erosional processes that are unique to each feature. The yardangs are streamlined by a dominant locally unidirectional wind, but no streamlining is seen in the mesas, buttes and associ- ated collapse features suggesting the absence of a strong unidirec- tional wind in these areas. The development of mesas and buttes from collapse features due to progressive recession of the margin of the MFF through removal of support from the interior of the lithology is important to understanding the lithology of the MFF material when not degraded by a strong unidirectional wind. This variability of erosional forms within the deposit emphasizes the overall complexity of this extensive formation that must be taken into account when evaluating its origin and the recent (Amazo- nian) geological history of this region of Mars. 04 (2009) 471?477As seen by the erosional progressions illustrated above, aeolian processes dominate for yardangs while collapse and mass wasting in response to loss of support from within the deposits produces the mesas. The spatial separation of mesas from yardangs suggests quite different local environmental conditions in their respective locations within the equatorial region of Mars. These erosional fea- tures are revealing environmental variability at a scale not yet incorporated into current wind modeling. Acknowledgments K.M. acknowledges research support from a 2006 North Dakota View Consortium Grant. S.deS. thanks North Dakota Space Grant and Oregon Space Grant for support to conduct this work. JZ was supported by NASA Grant NNX07AP42G for portions of this work. Hynek, B.M., Phillips, R.J., Arvidson, R.E., 2003. Explosive volcanism in the Tharsis region: Global evidence in the martian geologic record. J. Geophys. Res. 108 (E9), 15.1?15.16. Keating, G.N., 2005. The role of water in cooling ignimbrites. J. Volcanol. Geotherm. Res. 142 (1?2), 145?171. Mandt, K.E., de Silva, S., Zimbelman, J.R., Crown, D.A., 2008. The origin of the Medusae Fossae Formation, Mars: Insights from a synoptic approach. J. Geophys. Res. 113, E12011. doi:10.1027/2008JE003076. Malin, M.C., Edgett, K.S., 2000. Sedimentary rocks of early Mars. Science 290 (5498), 1927?1937. Malin, M.C., Edgett, K.S., 2001. Mars Global Surveyor Mars Orbiter Camera: Interplanetary cruise through primary mission. J. Geophys. Res. 106 (E10), 23429?23570. McCauley, J.F., Grolier, M.J., Breed, C.S., 1977. Yardangs of Peru and other desert regions. USGS, Interagency Report, Astrogeology, p. 81. K. Mandt et al. / Icarus 204 (2009) 471?477 477We appreciate comments of Peter Schultz on an earlier version of this manuscript and the reviews of two anonymous journal reviewers that allowed us to hone our message more clearly. References Anderson, F.S., Greeley, R., Xu, P., Lo, E., Blumberg, D.G., Haberle, R.M., Murphy, J.R., 1999. Assessing the martian surface distribution of aeolian sand using a Mars general circulation model. J. Geophys. Res. 104 (E8), 18991?19002. Bailey, J.E., Self, S., Wooller, L.K., Mouginis-Mark, P.J., 2007. Discrimination of ?uvial and eolian features on large ignimbrite sheets around La Pacana Caldera, Chile, using Landsat and SRTM-derived DEM. Remote Sens. Environ. 108 (1), 24?41. Barnes, J.W., Lisle, R.J., 2004. Basic Geological Mapping. Chichester, West Sussex, England; Hoboken, Wiley, NJ. Bradley, B.A., Sakimoto, S.E.H., 2001. Relationships between the Medusae Fossae Formation (MFF), ?uvial channels, and the dichotomy boundary southeast of Nicholson Crater, Mars. In: 32nd Lunar Planetary Science Conference, 1335 (abstract). Bradley, B.A., Sakimoto, S.E.H., Frey, H., Zimbelman, J.R., 2002. Medusae Fossae Formation: New perspectives from Mars Global Surveyor. J. Geophys. Res. 107 (E8), 2.1?2.17. Breed, C.S., Grolier, M., McCauley, J.F., 1979. Eolian features of the Western Desert of Egypt and some applications to Mars. J. Geophys. Res. 84 (B14), 8205?8221. Carter, L.M., Campbell, B.A., Watters, T.R., Phillips, R.J., Putzig, N.E., Safaeinili, A., Plaut, J.J., Okubo, C.H., Egan, A.F., Seu, R., Biccari, D., Orosei, R., 2009. Shallow Radar (SHARAD) sounding observations of the Medusae Fossae Formation. Icarus 199, 295?302. Cas, R.A.F., Wright, J.V., 1992. Volcanic Successions, Modern and Ancient: A Geological Approach to Processes, Products and Successions. Chapman Hall, London. Christensen, P.R., and 10 colleagues, 2004. The Thermal Emission Imaging System (THEMIS) for the Mars 2001 Odyssey Mission. Space. Sci. Rev. 110, 85?130. de Silva, S.L., Bailey, J., Mandt, K., 2009. Yardangs in terrestrial ignimbrites: Synergistic remote and ?eld observations on Earth with applications to Mars. Planetary and Space Science, in review, submitted for publication. de Silva, S.L., 1989. Geochronology and stratigraphy of the ignimbrites from the 21 30?S to 23 30? S portion of the Central Andes of N. Chile. J. Volcanol. Geotherm. Res. 37 (2), 93?131. Fenton, L.K., 2003. Aeolian processes on Mars: Atmospheric modeling and GIS analysis. Ph.D. thesis, CalTech, Pasadena, CA. Forsythe, R.D., Zimbelman, J.R., 1988. Is the Gordii Dorsum escarpment on Mars an exhumed transcurrent fault? Nature 336 (6195), 143?146. Goudie, A.S., 2007. Mega-Yardangs: A Global Analysis. Geogr. Compass 1 (1), 65?81. Greeley, R., Guest, J.E., 1987. Geologic map of the eastern equatorial region of Mars. U.S. Geological Survey Miscellaneous Investigations Series Map, I-1802-B. U.S. Geological Survey, Washington, DC. Head, J.W., Kreslavsky, M.A., 2004. Medusae Fossae Formation: Ice-rich airborne dust deposited during periods of high obliquity? In: 35th Lunar Planetary Science Conference, 1635 (abstract).McColley, S.M., Head, J.W., Neukum, G., 2005. The Medusae Fossae Formation: Geological characteristics and topographic and stratigraphic relationships of the lower member along southeastern Elysium Planitia. In: 36th Lunar Planetary Science Conference, 1184 (abstract). Mouginis-Mark, P., 1993. The in?uence of oceans on martian volcanism. In: 24th Lunar Planetary Science Conference, pp. 1021?1022 (abstract). Nussbaumer, J., 2005. Extent and further characteristics of former glaciated terrain in Elysium Planitia, Mars. In: 36th Lunar Planetary Science Conference, 1949 (abstract). Parker, T.J., 1991. A comparison of the martian Medusae Fossae Formation with terrestrial carbonate platforms. In: 22nd Lunar Planetary Science Conference, pp. 1029?1030 (abstract). Riehle, J.R., Miller, T.F., Bailey, R.A., 1995. Cooling, degassing, and compaction of rhyolitic ash-?ow tuffs: A computational model. Bull. Volcanol. 57, 319?336. Ross, C.S., Smith, R.L., 1961. Ash-?ow tuffs: Their origin, geologic relations, and identi?cation, Prof. Pap. U.S. Geol. Surv. vol. 366, p. 81. Sakimoto, S.E.H., Frey, H.V., Garvin, J.B., Roark, J.H., 1999. Topography, roughness, layering, and slope properties of the Medusae Fossae Formation from Mars Orbiter Laser Altimeter (MOLA) and Mars Orbiter Camera (MOC) data, J. Geophys. Res. 104 (E10), 24,141?24,154. Schultz, P.H., Lutz, A.B., 1988. Polar wandering on Mars. Icarus 73, 91?141. Scott, D.H., Tanaka, K.L., 1982. Ignimbrites of Amazonis Planitia region of Mars. J. Geophys. Res. 87, 1179?1190. Scott, D.H., Tanaka, K.L., 1986. Geologic map of the western equatorial region of Mars, U.S. Geological Survey Miscellaneous Investigations Series Map, I-1802-A. U.S. Geological Survey, Washington, DC. Sheridan, M.F., Ragan, D.M., 1976. Compaction of Ash-?ow Tuffs, Developments in Sedimentology 2. Elsevier, New York. 2, 677?717. Sparks, R.S.J., 1978. Gas release rates from pyroclastic ?ows: An assessment of the role of ?uidisation in their emplacement. Bull. Volcanol. 41, 1?9. Tanaka, K.L., 2000. Dust and ice deposition in the martian geologic record. Icarus 144, 254?266. Ward, A.W., 1979. Yardangs on Mars: Evidence of recent wind erosion. J. Geophys. Res. 84, 8147?8166. Watters, T.R., Campbell, B.A., Carter, L.M., Leuschen, C.J., Plaut, J.J., Picardi, G., Orosei, R., Safaeinili, A., Clifford, S.M., Farrell, W.M., Ivanov, A.B., Phillips, R.J., Stofan, E.R., 2007. Radar sounding of the Medusae Fossae Formation, Mars: Equatorial ice, or dry, low-density deposits. Science 318, 1125?1128. Wilson, C.J.N., 1984. The role of ?uidisation in the emplacement of pyroclastic ?ows, 2: Experimental results and their interpretation. J. Volcanol. Geotherm. Res. 20, 55?84. Wyrick, D., Ferrill, D.A., Morris, A.P., Colton, S.L., Sims, D.W., 2004. Distribution, morphology, and origins of martian pit crater chains. J. Geophys. Res. 109, E6005. Zimbelman, J.R., Patel, S.D., 1998. Photogeologic constraints on the emplacement of the Medusae Fossae Formation, Mars. In: 29th Lunar Planetary Science Conference, 1085 (abstract). Zurek, R.W., Barnes, J.R., Haberle, R.M., Pollack, J.B., Tillman, J.E., Leovy, C.B., 1992. Dynamics of the atmosphere of Mars. In: Kieffer, H.H., Jakosky, B.M., Snyder, C.W., Matthews, M.S. (Eds.), Mars. University of Arizona Press, Tucson, pp. 835? 933.