author's personal copy - eps.utk.edu

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

http://www.elsevier.com/copyright

Author's personal copy

Pingos on Earth and Mars

Devon M. Burr a,b,�, Kenneth L. Tanaka c, Kenji Yoshikawa d

a Earth and Planetary Sciences Department, University of Tennessee Knoxville, Earth and Planetary Sciences Department Building 412, 1412 Circle Dr., Knoxville, TN 37996, USAb Carl Sagan Center, SETI Institute, 515 N Whisman Rd, Mountain View, CA 94043, USAc USGS Astrogeology Team, 2255 N. Gemini Dr., Flagstaff, AZ 86001, USAd Institute of Northern Engineering, University of Alaska, Fairbanks, AL 99775, USA

a r t i c l e i n f o

Article history:

Received 25 March 2008

Received in revised form

9 November 2008

Accepted 12 November 2008Available online 27 November 2008

Keywords:

Mars

Surface

Ground ice

Periglacial

Terrestrial analogues

a b s t r a c t

Pingos are massive ice-cored mounds that develop through pressurized groundwater flow mechanisms.

Pingos and their collapsed forms are found in periglacial and paleoperiglacial terrains on Earth, and have

been hypothesized for a wide variety of locations on Mars. This literature review of pingos on Earth and

Mars first summarizes the morphology of terrestrial pingos and their geologic contexts. That

information is then used to asses hypothesized pingos on Mars. Pingo-like forms (PLFs) in Utopia

Planitia are the most viable candidates for pingos or collapsed pingos. Other PLFs hypothesized in the

literature to be pingos may be better explained with other mechanisms than those associated with

terrestrial-style pingos.

& 2008 Elsevier Ltd. All rights reserved.

1. Introduction

Pingos are meso-scale (order 100-m-diameter) ice-coredmounds (Figs. 1–5) that form in periglacial terrain as a result ofpressurized groundwater flow and progressive freezing. Thesespecific formation conditions make them helpful climatic andhydrologic markers. The potential of the ice core to hostpsychrophilic (i.e., extremophilic cold-loving) microorganismsmakes them attractive as potential astrobiologic targets. Theseconsiderations make identifying pingos on Mars useful for piecingtogether Mars’ climatic and hydrologic history and for providingpotential biotic or pre-biotic targets. On Mars, pingos have beenhypothesized in the literature as the explanation for a variety ofmound or ring forms at a range of latitudes in the northernhemisphere, with much more limited examples at some similarlatitudes in the southern hemisphere. These hypotheses are basedlargely on their similarity to terrestrial pingos in morphology andsize. Their geologic contexts on Mars are not always similar tothose on Earth. However, the local or regional Martian geologydoes provide indications of water and/or periglacial processes.Thus, pingos are commonly although not universally identified onMars based on the axiom that if their formation mechanisms aresimilar, terrestrial and Martian pingos should have a similar

shape. As discussed in Cabrol et al. (2000) and Burr et al. (2005),pingo size is not a function of gravity.

Given pingos’ utility for understanding Martian climate,hydrology, and potential astrobiology niches, a review of Mars’numerous pingo-like forms (PLFs) provides a perspective on theseissues. In the last decade, a series of visible-wavelength camerason Mars spacecraft have provided high-resolution images ofvarious hypothesized periglacial features, including fields ofinferred pingos. This work surveys the published literaturedescribing these PLFs, in order to assess the plausibility of theirorigin as ice-cored mounds and map out the distribution of likelypingos on Mars. First are reviewed the two terrestrial pingo typesand their associated water pressurization mechanisms, as well asother ice-mound features. This review is followed by an historicalsummary of the terrestrial pingo literature by geography locationand context. Then terrestrial pingo morphology is discussed,focusing on features that would be discernable in Mars spacecraftimages. In the discussion of Martian PLFs the information aboutterrestrial pingos is used in conjunction with available Mars datato assess the likelihood that features hypothesized in theliterature to be Martian pingos are pingos in the terrestrial sense.Whereas a wide variety of mound or ring features are found onMars, we focus here only on those features hypothesized inthe literature to be pingos. The Mars data include imagesand compositional spectral information from the Thermal Emis-sion Imaging Spectrometer (THEMIS), the Thermal EmissionSpectrometer (TES), the Compact Reconnaissance Imaging Spec-trometer for Mars (CRISM), the Mars Orbiter Camera (MOC), theHigh-Resolution Imaging Science Experiment (HiRISE), and the

ARTICLE IN PRESS

Contents lists available at ScienceDirect

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

Planetary and Space Science

0032-0633/$ - see front matter & 2008 Elsevier Ltd. All rights reserved.

doi:10.1016/j.pss.2008.11.003

� Corresponding author at: Earth and Planetary Sciences Department, University

of Tennessee Knoxville, Earth and Planetary Sciences Department, Building 412,

1412 Circle Dr., Knoxville, TN 37996, USA.

E-mail addresses: [email protected] (D.M. Burr), [email protected] (K.L. Tanaka),

[email protected] (K. Yoshikawa).

Planetary and Space Science 57 (2009) 541–555


High-Resolution Stereo Camera (HRSC), as well as topographicinformation from the Mars Orbiter Laser Altimeter (MOLA).

2. Terrestrial pingos

2.1. Pingo types: water pressurization mechanisms

Terrestrial pingos are canonically divided into two types,depending on the mechanism for pressurizing the groundwater.Open-system or hydraulic pingos form through artesian pressure(Muller, 1959; Holmes et al., 1968; French, 1996). Water issupplied under hydraulic pressure develop surrounding areasand flows beneath a low-permeability layer, which is commonlyprovided by permafrost. Permafrost is defined as a terrain whosetemperature is less than 0 1C for two years (Associate Committeeon Geotechnical Research, 1988). In areas of continuous perma-frost, permeability to groundwater is low because of the freezingconditions, although localized taliks—sites of unfrozen ground—-

may occur beneath shallow lakes. In comparison, the discontin-uous permafrost zone is characterized by ‘through taliks’ orunfrozen ground that connects the unfrozen ground below thepermafrost and the active layer that undergoes seasonal thawingabove the permafrost. ‘Through taliks’ increase the permeabilityof discontinuous permafrost regions over continuous permafrostregions. Most open-system pingos are found in the discontinuouspermafrost zone, in or near areas of significant relief, wherethe relief supplies the hydraulic gradient (potential) and the

discontinuous permafrost provides the permeability for ground-water upwelling. Artesian groundwater environments causerepeated injection of water via taliks to the surface, followed by

ARTICLE IN PRESS

Fig. 1. (a) An oblique aerial photograph showing three open-system pingos in a

mountain river valley, Spitsbergen, Norway. The two pingos on the left are

collapsed and have central ponds. ‘Ny’ pingo on the right shows both radial cracks

from dilation and diagonal cracks trending from upper left to lower right. A glacier

is located in the center background (flow direction is out of the photo), with two

moraines in front of it. (b) A generalized schematic of the hydrological and

glaciological situation in the photo in (a). Subglacial meltwater percolates through

an unfrozen zone in the discontinuous permafrost to form a pingo.

Fig. 2. (a) An oblique aerial photo showing an open-system pingo (arrow) at the

toe of a fan (to left), Hulahula River, Brooks, Range, Alaska. (b) A generalized

schematic showing the inferred formation mechanism for the pingo in (a).

Groundwater flow occurs from higher elevation to lower elevation through an

unfrozen zone (‘talik’) within the permafrost. The hydraulic head is provided by

elevation and confinement within the permafrost.

Fig. 3. (a) A ground photograph of a pingo located adjacent to South Kunlun Fault,

Kunlun Mountain, China (4700 m a.s.l.). Artesian pressure of sub-permafrost water

is 22 m higher than the ground surface. Polygonal terrain due to ice wedging is

visible in the foreground. (b) A generalized schematic of the situation inferred for

the photo in Fig. 3a. Pingos develop along minor faults associated with major fault

systems as a result of flow along the minor faults.

D.M. Burr et al. / Planetary and Space Science 57 (2009) 541–555542


freezing. This process leads to the development of the massive iceforming the core of the pingo. Thus, this pingo type is commonlyfound in glacial valleys (Fig. 1), at the toes of alluvial fans near

river bottoms (Fig. 2), in highly fractured or faulted uplands(Fig. 3), and on rebounding marine terraces (Fig. 4; Yoshikawaand Harada, 1994). In each of these settings, surroundingelevated terrain imparts artesian pressure to the groundwater,and discontinuous permafrost permits its upwelling at lowerelevations.

In contrast to open system or hydraulic pingos, closed systemor hydrostatic pingos (Fig. 5) are found in regions of continuouspermafrost (as reviewed in Mackay, 1979, 1998), which creates animpermeable layer at depth. Most closed-system pingos form indrained thaw lakes, in which the saturated lake basin sedimentsare exposed to the atmosphere. This exposure causes thesaturated sediments to freeze, forming new permafrost. However,residual ponds within the lakes retard permafrost growth,resulting in locally thinner permafrost and lesser resistance todeformation. As the permafrost aggrade downward, the porewater is expelled ahead of the freezing front. Blockage of thisdownward pore water flow by the continuous permafrost at depthredirects the flow inward from the basin edges. Injection ofexpelled water beneath the overlying permafrost, followed byfreezing of the injected water, causes progressive doming of thepermafrost overburden. This doming, commonly sited beneath theleast permafrost aggradation, is a result of the massive ice formingthe core of the pingo.

For both pingo types, the 9% expansion due to phase changefrom water to ice is insufficient to account for the size of pingos;groundwater flow is required to supply the volume of a pingo icecore. Both open- and closed-system pingos may undergo verticalpulsing (periods of uplift) as a result of injection of groundwaterto a sub-pingo water ice lens (Mackay, 1998; Yoshikawa, 2008).

2.2. Other ice-cored mounds

Closed-system (hydrostatic) pingos are often associated withfrost mounds (Mackay, 1998). Frost mounds are smaller, seasonal,ice-cored features that may precede pingo formation or developon the sides of pingos after they form. Both pingos and frostmounds form through similar processes, namely, the pressuriza-tion of ground water during freezing of the ground (Pollard andvan Everdingen, 1992). Their morphology is likewise similar inbeing mounded often with cracks. However, terrestrial pingos andfrost mounds differ in size by roughly one to two orders ofmagnitude. That is, pingos may be hundreds of meters in diameter(Figs. 1–5), whereas frost mounds are typically meters indiameter. Pingos and frost mounds also differ in the length oftime they exist. Dating of closed-system pingos by the willowsgrowing on their surfaces indicates ages on the order of athousand years (Mackay, 1973, 1986), whereas frost mounds formover months and do not last more than one season (Pollard andvan Everdingen, 1992).

Other large ice-mound features include palsas and lithalsas.Whereas pingos contain a massive ice core, palsas and lithalsasare comprised largely of segregated ice. Segregated ice is a resultof cryosuction, the process by which ice exerts a suction effect onthe surrounding pore water by capillary action (Davis, 2001). Thissuction continues until the thickness of the water-poor layerbeneath the ice layer is too great for further cryosuction. Becauseof the limit this thickness imposes on ice layer development,maximum palsa and lithalsa heights are of order 10 m on Earth.If the climate is sufficiently cold, subsequent layers of ice mayform below the water-poor layer. The term palsa refers tosegregation ice mounds associated with peat bogs, whichconstitute most segregated ice mounds on Earth; the term lithalsa(or ‘mineral palsas’) refers to segregation ice mounds that form insedimentary material and are considerably more rare (Gurney,2001; Pissart, 2002).

ARTICLE IN PRESS

Fig. 4. (a) An aerial photo showing a swath of pingos (arrows) about 1 km in total

length along a formerly submerged coastline on the river delta of Adventdalen,

Spitsbergen. The overburden of pingo ice is composed mainly of fine silt and mud,

but has a thin peat layer. The largest pingo is approximately 4.5 m high, 500 m

long, and 140 m wide, and is situated approximately 25 km from the former shore

line. (See Yoshikawa and Harada, 1995, for details.) (b) A generalized schematic of

the hypothesized situation in the photo in Fig. 4a. The marine terrace is

rebounding from glacial isostasy, and developing new permafrost. Deglaciation

permits upwelling of artesian groundwater through unfrozen zones (‘taliks’) in the

newly forming permafrost. New permafrost thickness is �30 m.

Fig. 5. (a) A black-and-white aerial photograph showing two closed-system pingos

on the Tuktoyaktuk Peninsula, NWT, Canada. ‘Split’ is located to the lower left and

‘Ibyuk’ is approximately at image center. Dilational cracking is visible on both

pingos, and thermal contraction polygons are visible on the surrounding terrain.

‘Ibyuk’ sits within a drained thaw lake, and ‘Split’ has been inundated by the Arctic

Ocean (just off the image to the left). ‘Ibyuk’ is �300 m in diameter for scale. (b) A

generalized schematic depicting formation of a closed-system pingo (after Mackay,

1998, Fig. 4). Arrows show the direction of permafrost aggradation. Pore water

expulsion ahead of the aggrading permafrost causes subsurface pooling of the pore

water and upward (hydrostatic) pressure. Freezing of this pore water creates the

pingo ice (blue).

D.M. Burr et al. / Planetary and Space Science 57 (2009) 541–555 543


Palsas and lithalsas have a very different genesis in comparisonwith hydrostatic or even hydraulic pingos, but basic similaritiesand differences may be suggested between these two types of ice-cored mounds. Palsas and lithalsas are composed of segregationice formed generally through cryosuction (Davis, 2001), whereaspingos are composed mainly of massive ice formed throughgroundwater pressurization. As with open-system or hydraulicpingos, palsas and lithalsas require an influx of (water) volume,along with a less significant contribution from phase-change-induced volume increase, to induce doming of the overlyingsurface. Like pingos, terrestrial palsas and lithalsas are associatedwith regions of permafrost. However, pingos are surrounded by(continuous or discontinuous) permafrost which is required tofocus groundwater flow during formation (see, e.g., Jorgenson etal., 2008), whereas palsas and lithalsas are themselves loci ofpermanently frozen ground, thus constituting the permafrost.Pingo and lithalsa morphologies are very similar (Gurney, 2001),including ‘ramparted depressions’ as relict forms (e.g., Matthewset al., 1997). However, the lower height attained by lithalsasshould produce less pronounced relict forms, or scars. As expectedfrom the limitations of the cryosuction formation mechanism,terrestrial palsas and lithalsas and their relict forms reported inthe literature are generally smaller and shorter than pingos (e.g.,Worsley et al., 1995; Matthews et al., 1997), although systematicmeasurements are not available to quantify this characterization.

2.3. Morphology

Pingo morphology varies with growth stage (Fig. 6; see alsoMackay, 1998, his Fig. 4 for diagrams illustrating closed-systempingo genesis and growth, and Burr et al., 2005, their Fig. 5 forphotographs of pingos in various stages of growth). Incipientpingos are low mounds (Fig. 6a). The diameter of the closed-system pingos is largely set at this stage, based on the size of the

residual ponds which retard permafrost growth and allow forpingo development. The controls on open-system pingo size areless well understood. Pingo diameters may reach up to 2 km, butare commonly reported to be of order of 100 m. The surfaces ofincipient pingos are generally smooth (Fig. 6a), unless the terrainhas already been cracked by thermal contraction caused bysudden drops of sub-freezing temperature (see, e.g., French, 1996or Davis, 2001 for descriptions of thermal contraction cracking).During their growth, the flow of groundwater continues to causeaccretion of ice onto the pingo ice core (Mackay, 1998). Activelygrowing pingos have a convex, domical or mound shape (Fig. 6b).Fully developed pingos may reach up to several tens of meters inheight, with the largest known examples being about 50 m tall.During this growth stage, the pingo commonly develops dilationalcracks (Fig. 6c), as the ice core expands beyond the overburdentensile strength (see, e.g., Mackay, 1998, his Fig. 9). Thesedilational cracks are generally oriented radially to the pingocenter (Fig. 5a), but may exploit any pre-existing weaknesses,such as thermal contraction cracks. Semi-concentric dilationalcracks may (rarely) develop as normal faults, with the upthrownside towards the pingo center (Mackay, 1998). Ice core growthmay steepen the pingo sides beyond the angle of repose, causingslumping of material from the pingo. Exposure through crackingof the central ice core causes melting and/or sublimation of theice, collapse of the pingo summit, and formation of a summitcrater (Fig. 6d). This summit crater may have marked topographyand the perennial water that results from the ice core melting maycreate a pond within the crater (Fig. 7), which enhances thepermafrost degradation. In addition to melting, vapor diffusionthrough the sedimentary substrate may also contribute to ice corecollapse. In its final stage, collapsed pingos form relict features or‘pingo scars’ (cf., Flemal, 1976) (Fig. 6e). These raise-rimmeddepressions are a result of the transport of sediment by slumpingdown the sides of the pingo, and so may show slight centrifugal(outward) displacement of mass from the original pingo circum-ference. Because of this mass displacement, the floor of the pingomay be slightly lower than the original or surrounding landsurface.

Pingos of different developmental morphologies may be co-located. Of the approximately 1350 closed-system pingos on theTuktoyaktuk Peninsula, only about 50 are actively growing(Mackay, 1973). The remaining pingos are in various morpholo-gical stages of collapse.

ARTICLE IN PRESS

Fig. 6. Cartoon showing profile view of pingo morphology during growth and

collapse: (a) incipient pingo, (b) growing pingo, (c) growing or collapsing pingo

with dilational cracking and slumping (dark areas) on the sides, (d) pingo in

advanced stage of collapse, (e) collapsed or relict pingo or pingo scar. In all cases,

stippled pattern connotes subsurface ice. Aspect ratio is exaggerated for ease of

viewing.

Fig. 7. Ground photo of the summit of Ibyuk Pingo (see Fig. 5 for aerial view),

showing the topography created by dilational cracking and a frozen pond (outlined

in black arrows) created by ice core melting. Persons at left suggest scale.



Features associated with the surrounding periglacial terrain,such as polygonal cracking, are commonly observed on the pingos.Seasonal frost mounds may develop on the sides of pingos, andnarrow circumferential annuli of lower terrain (‘moats’) have beendocumented surrounding a few pingo bases (e.g., Mackay, 1979).Sediment slumps are commonly visible on the sides of pingos, andicings may be observed at the pingo base due to discharge from awater lens beneath the ice core (Mackay, 1998).

2.4. Substrate

Pingos are almost always found in unconsolidated sedimentarysubstrates, which allow both groundwater flow and deformationduring pingo growth. Closed-system pingos in the North Slope ofAlaska are reported only in regions where finer-grained sedimentsoverlie coarser-grained sediments; no pingos are reported inadjacent lake basins that are fine-grained throughout (Carter andGalloway, 1979). On the Tuktoyaktuk Peninsula, 98% of the pingosform in drained thaw lakes, where fine-grained lacustrinematerial commonly overlies glaciofluvial sands. As an exampleof this stratigraphy, the overburden of Ibyuk Pingo (Figs. 5 and 7)is comprised of �4 m of lake silts, �2 m of diamicton with glacialboulders, and �9 m of medium glaciofluvial sands (Mackay, 1998).This stratigraphy of fine-grained over coarse-grained sedimentsallows pore water movement (inflow) through the coarser,subsurface sediments, while limiting water transport (outflow)through the overlying finer-grained sediments. This limitation ofwater outflow through the fine-grained capping layer createsupward pressure. At the same time, the atmospheric conditions inperiglacial regions cause freezing of this water. As discussedabove, the result of continued ground water inflow, and to a lesserextent the expansion due to the phase change of the water to ice,is uplift of the sedimentary overburden over a massive ice core.

Open-system pingos can develop with a wider variety ofsubstrates. Open-system pingos have been reported in clay(Yoshikawa and Harada, 1995), in sand (Yoshikawa, 1991), and insilt (Yoshikawa et al., 2003). For open-system pingo formation,water temperature should be near freezing in order to form amassive ice core. In addition, discharge should be moderate inorder to prevent breakthrough of the permafrost layer and theformation of spring flow instead of formation of a pingo(Yoshikawa, 1993). However, pingos have also been observed inshale (Matsuoka et al., 2004) and in sandstone (Muller, 1959).Shale and sandstone may have extensive fractures, especially inSvalbard (Yoshikawa, 2003) and Greenland (Muller, 1959), whichmay allow artesian groundwater to heave up the overburden,developing a pingo. However, the extreme scarcity of suchobservations suggests this is very unusual condition.

2.5. Historical, geographic, and contextual overview

This section describing pingos on Earth concludes by present-ing an overview of historical research on terrestrial pingos bygeographic region. We focus on features that have been moreeasily accessible to past investigators (and presumably to futureresearchers as well). Regional and local effects, such as proximityto bodies of water and geohydrologic conditions, account for thesegeographic associations. Such factors are poorly understood inMartian cases, which are thus more conservatively grouped andpresented by latitudinal region (see Section 3 below).

2.5.1. North American studies

Porsild (1938) was the first to suggest the word pingo (Inuit for‘small hill’) as a generic term for ice cored hills in permafrostregions, based on his observations in North America. Pingos in the

North American arctic coast regions are commonly closed system.The first published studies date back to the early 20th century.These early workers explored multiple working hypotheses,including pingo formation as (1) the erosional remnants of pre-existing mounds of uncertain origin, (2) as glacial deposits ofcombined sediments and ice, and (3) by pressurization ofgroundwater by mechanisms similar to those that give rise tohydraulic or open-system pingos (e.g., Liffingwell, 1919). Con-temporary observations show that closed-system pingos (e.g., Fig.5) predominate on the North Slope and Seward Peninsula, Alaska,and the Mackenzie Delta in Canada (Mackay, 1979, 1998). Theseareas are dominated by continuous permafrost and heavilypopulated with thaw-lake basins, in which 98% of these coastalpingos formed.

In contrast, inland pingos are commonly open-system types.About 760 pingos are found in Central Alaska and the YukonTerritory, Canada, of which only a few are closed-system pingos(Holmes, 1967; Hughes, 1969; Hamilton and Obi, 1982). Both ofthese inland areas are dominated by discontinuous permafrost.Many of these pingos formed at the toes of alluvial fans (e.g., Fig.1), where sub-permafrost water was likely to attain a maximumhydraulic head and where localization of artesian flow wouldperhaps occur. However, theoretical calculations and limitedgroundwater data from Alaska (Holmes et al., 1968) suggest thathydraulic pressure alone is insufficient to overcome the tensilestrength of the permafrost layer and the static load of theoverburden. Holmes et al. (1968) suggested that freezingpressures probably are also required to lift open-system pingos.

2.5.2. European studies and Greenland

Open-system pingos in East Greenland have been describedsince the late 1800s and researched since the 1950s (Muller, 1959;Cruickshank and Colhoun, 1965; Washburn, 1969; O’Brien, 1971).Early calculations of the total up-doming pressure required tocounteract the cohesion and weight of the overlying permafrostmaterial suggested that 600–2200 kPa of pressure was needed toform pingos. Subsequent work focused on pingo hydrology. Theorigins of open-system pingos in the Svalbard archipelago, in theArctic Ocean north of Norway, were described as resulting fromgroundwater accumulation beneath a glacier’s terminal moraineand flow toward the coast through the unfrozen zone beneath thepermafrost (Liestøl, 1977, 1996). This mechanism is consistentwith the Svalbard glaciers, which exhibit surging behavior withexcess water at their base. Of the approximately 80 pingos inSvalbard (Yoshikawa and Harada, 1995), almost all are of theopen-system type. Pingos are situated in river beds in centralSpitsbergen (the largest of the Svalbard islands) and about 20–30%of these have artesian flow caused by the flow of subglacialgroundwater. Pingos in Svalbard also occur in near-shore upliftareas and at fault zones (Yoshikawa and Harada, 1995), showingthe diversity of locations that may be conducive to groundwaterflow and pingo formation.

No active pingos are known in the Scandinavian Peninsula or inthe European mainland, although many palsas are reported inScandinavian countries. Relict pingos in Europe, i.e. in Ireland(Mitchell, 1973), Great Britain (Pissart, 1963; Watson, 1971),Belgium (Pissart, 1965), and Germany(Weigand, 1965), have beenthe subject of research for the evidence they provide ofpaleopermafrost and paleoclimate conditions.

2.5.3. Eurasian studies

Pingos in Siberia have been studied since the mid-20th century(see Tikhomirov, 1948). They are concentrated in the Yakutiaregion of north-central Siberia, but can be found from the WhiteSea in the west to the Bering Strait in the east. These pingos are

ARTICLE IN PRESS



located at the toes of alluvial fans or in valley bottoms, which arequintessential locations for open-system pingos in Alaska. Pingosin the Boguchansk region of central Siberia are open-systempingos (Serov and Leshchikov, 1978), located on permafrost withtemperature profiles similar to those in central Alaska.

In China, pingos are found in two major regions. The firstregion is the northeast part of China, continuous from Siberia andMongolia. The second region is the Tibetan Plateau, which maypresent useful analog conditions to Mars. The Tibetan Plateau hasa widespread area of permanently frozen terrain above 4000 melevation, producing cold, dry permafrost. In addition, initialstudies show that many of the groundwater systems have brinywater (Yoshikawa et al., 2006a, b). Pingos have been reported andsampled near Kunlun Pass, northern Tibet (An, 1980), but to ourknowledge, very few other pingo studies have been carried out onthe Tibetan Plateau. Related permafrost and periglacial features inTibet were investigated along the Golmud–Lasha Road (e.g., Kuhle,1985; Wang, 1982).

3. Martian PLFs

Knowledge of terrestrial pingos, in conjunction with knowl-edge of Martian conditions, allows assessment of PLFs on Mars.Martian PLFs discussed in the literature may be grouped bylatitude. We discuss Martian PLFs moving from the polar regions,through the mid-latitudes, to the equator. Fig. 8 shows the genericlocations of the various study areas presented in the literature.Detailed locations are provided in the analysis below. A summaryof information on PFLs by location is provided in Table 1.

3.1. Polar PLFs: large impact craters

Pingos have been hypothesized by Sakimoto and colleagues fornearly circular mounds in circum-polar impact craters polewardof 601; all but one are in the northern latitudes (Sakimoto,2005a, b; Bacastow and Sakimoto, 2006). These impact craters are4�8 km in diameter, and the inset mounds occupy a large portionof the crater. These mounds are surrounded by partial or completemoats, and MOLA topography shows that they usually have ashallower equator-ward exposure than poleward exposure. Theyoccasionally show central pits.

The central pits, being difficult to produce through uniformsedimentary accumulation, are interpreted as indicative of pingoformation (Sakimoto, 2005a). The asymmetry is held to beindicative of a volatile-driven process, namely loss throughsolar-driven sublimation (Sakimoto, 2005a). Numerical modelingshows that sufficient flow volume could be generated by hydraulicpressure from an hypothesized regional aquifer to produce theobserved shapes (Sakimoto, 2005b; Bacastow and Sakimoto,2006). The mounds’ hypothesized groundwater source is aregional aquifer, which would supply water through impactfracturing of the crust. The large size of the PLFs relative toterrestrial pingos is proposed to be a result of the large subsurfacewater supply and a sustained cold climate in Mars’ polar regions.

Korolev crater (Fig. 9), a large (83-km diameter) and well-preserved north polar crater (731N, 1631E), provides an example ofa polar crater with ‘pingo-like fill’ (Bacastow and Sakimoto, 2006).Previous work on Korolev crater using data from TES, THEMIS, andCRISM indicates that the surface of the mound in Korolev crater iscomposed of water ice (Armstrong et al., 2005; Brown et al.,2007). These data were all collected during the northern summer.In growing terrestrial pingos, the ice core is buried beneath anactive layer (at least �1 m of overburden) (Mackay, 1987, hisFig. 10). Thus, an entirely water ice surface would be inconsistentwith formation analogous to that of terrestrial pingos and wouldinstead be more analogous to a terrestrial icing mound. Icingmounds are a common site in Arctic regions of groundwaterupwelling, where they occur during winter as a result of artesianpressure (Yoshikawa et al., 2007) and are visible on some open-system pingos. Frost cover over regolith could provide theobserved water ice signature, but modeling of the thermal inertiaindicates the water ice was several meters thick (Armstrong et al.,2005). Alternatively, extensive cracking in the overburden couldexpose the water ice to detection, consistent with formation as aterrestrial pingo. Thus, one test of the pingo hypothesis is whetherthe ice detection derives from cracks in the mound. We examinedavailable THEMIS mosaics, MOC images, and one CTX image(P01_001592_2530_XI_73N195W) of Korolev crater and the otherPLFs given in Bacastow and Sakimoto (2006, their Fig. 1), in orderto look for large cracks that may have exposed interior ice todetection in spectral data. For terrestrial-style pingos (i.e., thatformed through water ice accumulation beneath a thin regolithcover), we would also expect to see frost mounds and polygonalterrain, as are commonly associated with terrestrial pingos.

ARTICLE IN PRESS

Fig. 8. Generic locations of Martian PLFs as proposed in the literature and summarized in Table 1, plotted on MOLA shaded relief topography. The plot stretches from 901N

to 901S and 1801W to 1801E.


Author's personal copyARTIC

LEIN

PRESS

Table 1Summary of information and assessment regarding hypothesized pingos on Mars..

Location Pingo

likelihood

Context Data Primary

references

Morphologies Inferred terrain

type

Tectonic or

impact

structures

Potential

periglacial

features

Potential

lacustrine

features

Potential

volcanic

features

Pingo

mechanism(s)

proposed

Polar regions

Circum-polar

regions

Less likely Large

impact

craters

MOLA Sakimoto

(2005)a,b,

Bacastow and

Sakimoto (2006)

Circular; some

have central pits

and moats; some

may underlie

layered deposits

Large impact

craters, minimum

size 8 km

Occurs within

impact craters

tens of km in

diameter

Hydraulic,

sourced from

global aquifer

Mid-latitudes

NW Utopia

Planitia

More likely Impact

crater

MOC,

THEMIS

visible

Osinski and Soare

(2007)a,b Soare et

al. (2005)

Small mounds,

some with

summit

depressions;

circular to

irregular rimmed

depressions,

some with

interior layered

mounds

Astapus Colles

and Vastitas

Borealis interior

unitsa and

superposed crater

Crater floor Polygonal

terrain

(40–300 m

diameter)

Hydrostatic

Utopia basin

33–431 N

More likely Basin floor MOC de Pablo and

Komatsu (2008)

Domes, cones,

rings; most have

central pits

Various units of

possible fluvial

and volcanic

origin, especially

possible

megalahar

depositsa

Wrinkle ridges Polygonal

troughs

several km in

diameter

Megalahar

deposits,

possible lava

flows proximal

to Elysium

Magmatically

driven melting

of permafrost

30–471N,

80–1201E

More likely Mid-

latitudes in

Utopia

Basin

HiRISE,

MOC

Dundas et al.

(2008)

Circular/oval/

irregular; some

domical, some

flat-topped,

profiles; some

have concentric

and radial

grabens

Various northern

plains and

highland unitsa,b

Scattered

wrinkle ridges

Some in areas

of polygonal/

fractured

terrains

None for

pingos; flat

topped

morphologies

Equatorial regions

North of Thira

crater on

floor of

Gusev

crater

Less likely Impact

crater

Viking Cabrol et al.

(2000)

Circular/elongate/

composite/

collapsed

mounds; some

have central pits;

dense cratering

Unit AHbm1c,

interpreted as

fluvio-lacustrine

sediments

Small wrinkle

ridges

Associated

with possible

fluvio-

lacustrine

deposits

Hydrostatic

(closed-

system)

Athabasca

Valles

Less likely Outflow

channel

MOC Burr et al. (2005),

Page and Murray

(2006), Jaeger et

al. (2007, 2008),

Page (2008)

Mounds, pitted

cones, possible

dilation cracks,

irregular rimmed

forms

Cererus Fossae 3

unita, interpreted

as lava flows with

minor fluvial

deposits

Cerberus

Fossae

Polygonal

terrain

Polygonal

terrain, lava

flows common

Hydrostatic

(closed-

system) or

hydraulic

(open-system)

Cerberus Palus Less likely Basin, along

northern

margin of

MFF

MOC Nussbaumer et al.

(2003, 2004),

Nussbaumer

(2005)

Pitted cones; low

population

density

Mixture of MFF

and lava

Inferred

rootless cones

a Tanaka et al. (2005).b Scott et al. (1986–87).c Kuzmin et al. (2000).

D.M

.B

urr

eta

l./

Pla

neta

rya

nd

Spa

ceScien

ce5

7(2

00

9)

541

–5

55

54

7


Neither cracks on the scale of the spectral data nor otherassociated features were apparent in these images. Thus, wesurmise that the water ice inferred from spectral data reflects theoverall surface composition of the mounds, and is not due toexposure through cracking of an interior ice core. This conclusionis consistent with the interpretation of Armstrong et al. (2005) forthe spectral data from Korolev crater.

Moreover, the mound displays local fine, even layering mostlyoverlain by relatively dark, pitted deposits (e.g., MOC image R20-00307). The finely layered deposits are similar to late-stage, ice-rich materials making up upper parts of the north polar plateauthat are locally overlain by dunes, sedimentary veneers, andresidual ice (Rodriguez et al., 2007; Seelos et al., 2007; Tanaka etal., 2008); no PLFs have been identified thus far in these otherstratigraphically similar materials. On the basis of this layering,these northern PLFs have been hypothesized to be outliers ofnorth polar layered terrain locally capped by residual water ice(Tanaka and Scott, 1987; Tanaka et al., 2005). The absence of suchmounds in all polar craters is held to be counterindicative of thisorigin (Bacastow and Sakimoto, 2006). However, this may be afunction of crater diameter, depth, elevation, and latitude. Korolevis the largest crater north of 701N. The next largest, Dokka, is52 km in diameter at 77.21N, 214.41E and also contains a mound(Tanaka et al., 2005), as does an unnamed, 30-km-diameter craterat 77.21N, 89.21E. Somewhat smaller craters along the margin ofand atop Planum Borum also have mounds of layered deposits. Incontrast, Lomonosov (�110 km diameter but at 651N) does nothave a mound. Some smaller craters also include mounds formedby dune deposits, showing that sediment collection in polarcraters is complex (e.g., three craters 16–21 km in diameter near751N, 3451E).

In the southern hemisphere, a regional suite of large mounds,called the Sisyphi Montes, occur at the centers of degraded impactcraters. Ghatan and Head (2002) interpreted these features to besubglacial volcanoes beneath a 41.4-km-thick ice sheet, based ontheir flat tops, conical shapes, summit craters, and associatedchannels. Rodriguez and Tanaka (2006) further observed that thefeatures (1) occur at the centers of highly rimmed and rimlesscircular depressions interpreted to be degraded impact craters, (2)include adjacent moats o200 m deep within the crater floors inabout half the cases, and (3) appear to be organized along possibletraces of Hellas basin rings. Their alternative explanation for thesouthern mounds is that of a regional phase of magmatisminvolving magma ascent along a combination of Hellas basinstructures and brecciated zones beneath larger impact cratersresulting in eruptions at crater centers; the surrounding moats

may have been caused by initial enrichment and subsequentremoval of volatiles in the brecciated crust within the impactcrater floor adjacent to the volcanic mound (Rodriguez andTanaka, 2006). In this scenario, a volcanic, non-impact (orenhanced impact) central peak is formed much later than theimpact event, perhaps during the Hesperian as suggested byGhatan and Head (2002). Other southern, circum-polar impactcraters contain layered and dune-capped mounds that preferen-tially infill the southern parts of the crater floors and in somecases merge with layered deposits of the south polar plateau(Planum Australe). These features have been mapped as outliers ofvolatile-rich, south polar layered deposits (Tanaka and Scott,1987), whose form and preferred poleward occurrence appear tobe related to climate-controlled energy mass-balance (Russell etal., 2004). None of these proposed processes for southern, high-latitude mounds is similar to the formation of pingos by ground-water flow and freezing.

In summary, the various mechanisms proposed for the forma-tion of mounds in large polar impact craters are not similar to theprocesses that form terrestrial pingos. The stratigraphy, form, andorientations of the polar PLFs suggests that these layered moundsare outliers of ice-rich polar layered terrain (Tanaka and Scott,1987; Russell et al., 2004; Tanaka et al., 2005), whereas the flat-topped Sisyphi Montes are likely some sort of volcanic feature(Ghatan and Head, 2002; Rodriguez and Tanaka, 2006). In thesecases, the features commonly associated with pressurized growthof ice beneath a layer of regolith are lacking in available images.Thus, our analysis of available data does not support the hypothesisthat these PLFs in large polar impact craters are perennial ice-coredmounds analogous to terrestrial open-system pingos.

3.2. Mid-latitude PLFs: Utopia Planitia

Pingos have been hypothesized to explain variously-shaped,meso-scale features in the mid-latitudes. Although Dundas et al.(2008) note one PLF in the southern hemisphere, to date mappingof PLFs in the mid-latitudes has focused on Utopia Planitia.

Centered near 501N 1201E, Utopia Planitia is an ancientmultiring impact basin about 3300 km in diameter (McGill,1989). As part of Mars’ northern lowlands, it may have beeninundated by a northern ocean (Baker et al., 1991; Parker et al.,1993; Head et al., 1999). The lowlands are filled with the VastitasBorealis units, interpreted to be Early Amazonian periglaciallyreworked sediments possibly deposited in an ancient ocean, aswell as Hesperian marginal-plains deposits emplaced largely by

ARTICLE IN PRESS

Fig. 9. Korolev crater (near 731N, 1631E) as seen in image mosaics. (a) The MOC wide angle images show the bright frost deposit. (b) The THEMIS daytime IR images show

the relatively high thermal inertia compared to the surrounding terrain. The 50-km scale bar on (a) applies to both images.



mass wasting (Tanaka et al., 2005; Skinner and Tanaka, 2007).Interstitial water within these sediments may have frozen (Chap-man, 1994, 2003), producing an ice-rich permafrost zone over-lying wet sediments. Widespread polygons and possiblethermokarst provide supporting evidence for this scenario(Costard and Kargel, 1995). Channels on the western ElysiumMons slopes may also have contributed wet sediments to thebasin through magmatically induced debris flows and lahars(Christiansen, 1989; Tanaka et al., 1992; Russell and Head, 2003).Local fractured rises, elliptical mounds, pitted cones, and roundeddepressions indicative of sedimentary diapirism and mud volcan-ism involving water-saturated sediments and ice-rich permafrostare prevalent in Hesperian materials in the marginal plainsbetween the Vastitas Borealis plains materials and the Utopiabasin rim materials. These features may be influenced by buriedimpact-basin structure and topography (Skinner and Tanaka,2007). Thus, the necessary elements for terrestrial-style pingoformation—namely, a sedimentary substrate, subsurface ground-water, and freezing conditions, as well as structural anomalie-s—have all been inferred in Utopia Planitia.

Early identification of PLFs in northwestern Utopia focused onthe floor of a crater near 651N, 671E, showing a set of roughlycircular mounds 25–50 m in diameter (Soare et al., 2005). Thesemounds were hypothesized to be pingos on the basis of their sizeand morphology, which included radial cracks and summitdepressions visible in MOC image E03-00299. As discussed underSection 2 above, radial cracking is observed on terrestrial pingosas dilation cracking and summit depressions on terrestrial pingosresult from melting or sublimation of the ice core and subsequentcollapse. Associated features inferred to be periglacial in origin,namely polygonal cracking and a curvilinear (channel-like)feature, led to interpretation of the crater floor as supporting anassemblage of features most consistent with formation byaqueous periglacial processes. The proffered terrestrial analog isthe assemblage of features found in the Tuktoyaktuk Peninsula,Northwest Territories, Canada, which includes one of the largestconcentrations of closed-system (hydrostatic) pingos on Earth(Mackay, 1998 and references therein). Images from HiRISE(e.g., Fig. 10a) show the same suite of features inferred from thelower resolution MOC image, namely, mounds with radial cracksthat in some cases extends onto the surrounding terrain (Fig. 10b)and/or summit depressions (Fig. 10c). In addition, polygonalcracking of the surrounding crater floor and bright curvilinearfeatures are visible, previously hypothesized to be thermal

contraction polygons and meltwater flow features, respectively(Soare et al., 2005). Thus, the higher resolution data are consistentwith the earlier interpretation of the mounds as pingos. The watersource for the pingos was hypothesized to have been atmospheric,tied to obliquity variations (Soare et al., 2005).

Subsequent classification and interpretation of some sub-circular to circular landforms as pingos has been made throughoutUtopia basin (Osinski and Soare, 2007; de Pablo and Komatsu,2007, 2008; Dundas et al., 2008). These features show a range ofmorphologies. Using MOC and THEMIS images, Osinski and Soare(2007) identify five morphologic types, namely: (1) small sub-circular mounds, (2) shallow depressions surrounded by con-centric and radial cracks, (3) sub-circular to irregularly shapedrimmed depressions, (4) sub-circular rimmed features containinglayered mounds, and (5) circular rimmed-depressions surroundedby ejecta. Whereas types 2 and 5 are suggested to be impactcraters, types 1, 3, and possibly 4 are held to resemble pingomorphologies at different stages of growth or collapse. ‘Small sub-circular mounds’, including some with radial cracks (Fig. 11a), aremorphologically similar to incipient or growing pingos, and areinferred by Osinski and Soare (2007) to be closed-system(hydrostatic) pingos. As shown in Mackay (1998), radial (dilation)cracks on pingos may connect with polygonal (thermal contrac-tion) cracks on the surrounding terrain. ‘Sub-circular to irregularlyshaped rimmed depressions’ are similar in size and morphology tocollapsed pingos or pingo scars (Fig. 11b). As noted in Section 2above, terrestrial pingos in various morphological stages ofdevelopment are commonly co-located, so the multiple adjacentmorphologies is consistent with interpretation of Osinski andSoare (2006) that these features are collapsed pingos. In somecases the depth of the surface inside these rimmed depressionsappears noticeably greater than the terrain surface outside therim. As noted in Section 2 above, transport of material away fromthe pingo center by slumping may cause a collapsed pingo centerto be slightly lower than the surrounding terrain, although theamount of this effect is not known. Examples of a thirdmorphology, ‘sub-circular rimmed features containing layeredmounds’ (Fig. 11c), are similar to features located on the floors ofinferred thermokarst depressions and previously interpreted byMorgenstern et al. (2006) to be impact craters partially filled withlayered material. Osinski and Soare (2007) note that not all suchfeatures are located within thermokarst basins and suggest thatthey could be due to pingo regrowth within a slumped basin.Horizontal layering is not characteristic of terrestrial pingos,

ARTICLE IN PRESS

Fig. 10. (a) Subset of HiIRSE image PSP_008492_2450 (centered near 65.41N, 67.31E), showing the floor of an impact crater in Utopia Planitia with features hypothesized by

Soare et al. (2005) to be pingos, thermal contraction polygons, and drainage features. White boxes show the locations of Fig. 10b and c. (b) and (c): Subsets of HiRISE image

PSP_008492_2450, showing hypothesized pingos with (b) with radial cracking (white arrows), and (c) a summit depression. The 50-m scale bar on (b) applies to both

images (b) and (c).



however, and the observed layering in the mounds is notexplained for an ice-growth mechanism consistent with forma-tion of a closed-system pingo. On the basis of other inferredground-ice features in the area, Osinski and Soare (2007)speculate that the water emplacement was recent, although aspecific process is not proposed.

Using both MOC and HRSC images, de Pablo and Komatsu (2007,2008) provide a simpler, tripartite classification of features—domes,cones, and rings—although each of these classes includes a range ofindividual examples. Domes and cones tend to occur in groups of tenor more individual features, whereas rings tend to occur in smallergroups. A high-resolution image of Utopia Planitia shows examples ofthese features (Fig. 12a). Domes are defined by de Pablo and Komatsu(2007, 2008) as circular, domical features with diameters rangingfrom 20 to 200 m (average 90 m) and heights of less than 45 m(Fig. 12b). Most domes are flat without cracks or fissures; a few showshallow summit depressions. Cones are defined as conical featureswith summit depressions and having diameters ranging from 11 to890 m (average 85 m) (Fig. 12c). Rings are defined as annular featurescharacterized by smooth, thin rims, which may form complexstructures. They range in diameter from 30 to 850 m (average150 m) (Fig. 12d). Most of these PLFs are located between �331N and431N (de Pablo and Komatsu, 2008, their Fig. 1), although a fewoutliers are located outside of this concentration. Because the threetypes of features are closely related in space, a common origin wassought for all three types. They are hypothesized to be pingos basedon their morphology, which is similar to pingo morphology atdifferent stages of development, and their location within geologicunits that were inferred to be water-rich (i.e., megalahars). Rootlesscones are presented as another possible hypothesis, but de Pablo andKomatsu (2007, 2008) did not find clear evidence of volcanicmaterials co-located with the PLFs. The water for the PLFs may havederived from melting of ground ice, hydrothermal water flows, orboth mechanisms together. The heat source for these mechanisms isspeculated to be a recent, deep magma chamber, allowing pingoformation under current climatic conditions (de Pablo and Komatsu,2007, 2008).

Using HiRISE images, Dundas et al. (2008) focus on a subset ofthe features proposed as pingos by Osinski and Soare (2007) andde Pablo and Komatsu (2007), namely, near-circular isolatedmounds with steep sides and flat summits (Fig. 13). They noteseveral similarities to terrestrial pingos, including scale, size, andradial and concentric fractures. The similarity of the surroundingterrain with the surface of the PLFs is held to indicate formation

by uplift of the surrounding terrain, and thus to be mostconsistent with formation as pingos. They found a concentrationof these features between �351N and 451N, similar to that of dePablo and Komatsu (2007). Alternative mechanisms besides upliftare discounted as being inconsistent with the observed character-istics of the surface. Conversely, the latitudinal control of themounds is held to be best explained by ground ice processes(Dundas et al., 2008). Pingos are presented as the most similarground ice landform on Earth, although the ‘trapezoidal profile’ ofthe Utopia Planitia mounds is noted as being atypical of terrestrialpingos. Flat-topped pingos have been noted on Earth; a well-documented example from the Tuktoyaktuk Peninsula is Pingo 14(Geyser Pingo) of Mackay (1998). This large closed-system pingo(42400 m in its longest dimension) is located in a drained thawlake, along with two other pingos which are more mound shaped.Its location in a thaw lake supports its identification as a pingo, asdoes its large sub-pingo water lens, determined by yields ofartesian water from repeated drilling. Pingo 14 gained and lostelevation over 18 years through water influx and egress (in partdue to drilling), so that its final elevation was significantly belowits highest elevation. Such elevation gain and loss is not unusual,however, and also occurs in mounded pingos. Pingo 14 is uniquein the morphology of its dilational cracks, which originate fromthe pingo summit, but then bifurcate before reaching the pingoflanks. This crack morphology is not understood, nor is the causefor the unusually flat top of the Utopian features. As suggested byDundas et al. (2008), the flat-topped mounds could be morpho-logically similar to lithalsas. As discussed in Section 2 above,lithalsas tend to have flat upper surfaces and so may be a viableanalogue for the Utopia mounds.

In summary, research in the Utopia Planitia shows that theregional context is favorable to terrestrial-style pingo formationand that the observed features span a range of forms that aresimilar to the range of terrestrial pingos undergoing growth andcollapse. However, some questions remain about these UtopiaPlanitia features, including the water source for the hypothesizedpingos, the timing of their formation, and the flat-toppedmorphology observed in some cases.

3.3. Equatorial PLFs: a range of contexts

In Mars’ equatorial regions, PLFs are found in a plethora oflocations.

ARTICLE IN PRESS

Fig. 11. High-resolution images showing examples of PLFs in Utopia Planitia as classified by Osinski and Soare (2007). In all images, north is up and illumination is from the

left. (a) Subset of HiRISE image PSP_007951_2220 (centered near 41.61N, 82.91E), showing small sub-circular mounds with radial cracks (black arrows denote the most

prominent crack). (b) Subset of HiRISE image PSP_007951_2220, showing sub-circular to irregularly shaped rimmed depressions. (c) Subset of HiRISE image

PSP_008162_2220 (centered near 41.61N, 82.91E), showing sub-circular rimmed features containing layered mounds. Both scale bars are 50 m; scale bar on (b) also applies

to (c).



3.3.1. Gusev crater

Rounded mounds commonly with raised margins in Gusevcrater (Fig. 14) were hypothesized to be pingos by Cabrol et al.(2000). As derived from Viking images, mound lengths weremeasured as ranging up to 2200 m, widths were measured asranging up to 900 m, diameters were measured as ranging from

300 to 1000 m, and heights were measured from shadowmeasurements as being between 35 and 240 m. These moundshave circular, elongated or composite planforms, and some showinterior depressions which may be interpreted as evidence ofcollapse. They were hypothesized to be pingos on the basis of theirplan view morphology and morphometry (ratio of length towidth), which are shown to be roughly similar to published data(Pissart and French, 1976) on terrestrial pingos. An inferredperiglacial origin for other features in Gusev crater appeared toprovide a supporting geologic context to the pingo interpretation(Cabrol et al., 2000).

More recent data, however, do not support these inferencesregarding the material in Gusev crater. Data from the Spirit MarsExploration Rover show that much of Gusev crater is filled withlava (McSween et al., 2006 and references therein), but thematerials in the vicinity of the proposed pingos are less wellcharacterized. THEMIS (Fig. 14b) and MOC images (e.g., S15-00196) show that these PLFs are highly degraded by impactgardening and embayed by a thin sheet deposit of uncertainorigin. Although the size and irregular shape are similar to somehydraulic pingos on Earth, their density appears high for hydraulicpingo formation and their spatial distribution is inconsistent withthat of terrestrial pingos (Bruno et al., 2006). Based on this lack ofa sedimentary substrate and anomalous spatial distribution, wediscount the hypothesis that these PLFs in Gusev crater areterrestrial-style pingos. Instead, they may relate to deflation anddeformation of a debris-flow deposit or other wet sedimentary orhydrovolcanic material. Similar processes but with more distinctvolcanic and tectonic associations have been proposed in the

ARTICLE IN PRESS

Fig. 12. High-resolution images showing examples of PLFs in Utopia Planitia as classified by de Pablo and Komatsu (2008). In all images, north is up and illumination is from

the left. (a) Subset of HiRISE image PSP_002095_2130 (centered near 32.71N, 127.61E), providing context and location of sub-images, which show examples of (b) domes, (c)

cones, and (d) rings. The 50-m scale bar on (b) applies to all three sub-images.

Fig. 13. Subset of HiRISE image PSP_004180_2165 (centered near 36.21N 84.01E)

showing an example of a flat-topped PLFs in Utopia Planitia as identified by

Dundas et al. (2008). North is up and illumination is from the lower left.



Galaxias Fossae region of eastern Utopia, where volcanic flowsand lahars led to deflated and deformed-appearing flows alongwith linear grooves and cratered edifices (Skinner et al., 2007).

3.3.2. Cerberus region/Athabasca Valles

Rings, cones, and mounds have been hypothesized to be pingosat a variety of sites in the Cerberus region. In the Athabasca Vallesoutflow channel (Fig. 15a), a limited area of PLFs (Fig. 15b and c) ina proximal location in the channel was hypothesized to be a fieldof pingos based on their size and morphology (Burr et al., 2005).Small non-concentric inner cones, wide exterior moats, and acontinuum of PLF morphologies with size were all interpreted asmost consistent with pingo formation in comparison with otherpotential analogs. The size of the surrounding polygonal terrainwas also interpreted as being consistent with periglacial pro-cesses, namely, thermal contraction cracking. The geologic contextcould have supported formation of either hydraulic or hydrostaticpingos. Similar reasoning was extended by other workers to PLFsfound in both proximal and distal locations (Page and Murray,2006). In an analysis of MOC images, inferred secondary impactcraters (cf., McEwen et al., 2005) were observed to be crosscut bythe PLFs. Such stratigraphic relationships were interpreted asshowing that the formation of the PLFs postdated formation of theimpact craters, implying a time lag between emplacement of thesubstrate in the channel and formation on that substrate ofthe PLFs. Such timing would have been consistent with pingoformation (which would lag substrate emplacement) and incon-sistent with formation of the PLFs as rootless cones (which wouldoccur near-simultaneously with the substrate emplacement).

ARTICLE IN PRESS

Fig. 14. PLFs in Gusev Crater. (a) MOLA shaded relief image of Gusev Crater. White

asterisk shows location of (b). (b) A portion of THEMIS image V15085003 (centered

near 14.31S 175.81E), showing PLFs of Cabrol et al. (2000), approximately numbers

10–15 (the correspondence between features in the Viking image used in the

original work and the THEMIS image is not exact). North is up, and illumination is

from the left.

Fig. 15. Examples of PLFs in Athabasca Valles. (a) MOLA topography of Athabasca Valles (oriented northeast to southwest across the image), showing locations of (b)

(designated by the letter ‘b’) and approximate locations (designated by x’s) of fields of PLFs as discussed in the multiple cited references. (One field in Page and Murray

(2006) is off the image to the southwest.). (b) Portion of MOC image R0100745 (centered near 9.71N 2041W/1561E), showing a portion of a field of PLFs and subaqueous

dunes on the north side of a streamlined form. North is up and illumination is from the left. The black box outlines the area shown in (c). (c) Portion of MOC image R0100745

showing a close up of PLFs. The black box shows outlines the area shown in (d). (d) Portion of HiRISE image PSP_002661_1895 (centered near 9.61N 204.71W/156.31E)

showing PLFs.



This and various other fields of PLFs in Athabasca Valles havebeen given a variety of interpretations. These interpretationsinclude kettle holes (Gaidos and Marion, 2003), which form fromdeposition and subsequent melting or sublimation of large iceblocks, and basaltic ring structures (Jaeger et al., 2003, 2005),which form through hydrovolcanic interactions and subsequentextensive erosion. The first interpretation of conical landforms inthis channel was as rootless cones, which also result fromhydrovolcanic interactions but do not require subsequent erosion(Lanagan et al., 2001; Greeley and Fagents, 2001; Lanagan, 2004;Jaeger et al., 2007). From analysis of recent HiRISE images, thischannel is interpreted as being draped in lava that once filled thechannel and subsequently drained downslope into Cerberus Palus(Jaeger et al., 2007). In some locations, the moats around PLFs arerimmed with thin, high-relief morphology, interpreted as up-tilted slabs of lava that embays the channel (Fig. 15d). Other PLFsoverlap to form chains along the direction of flow, indicating thatthe volatile source for the PLFs came from below the flow surface.These observations are interpreted as supporting the rootless conehypothesis for the PLFs over the various other hypotheses,including the pingo hypothesis (Jaeger et al., 2007).

These observations and their interpretation remain controver-sial. Important points of discussion include whether the PLFs andthe surrounding polygonal terrain overlie or are overlain bysecondary impact craters (cf., McEwen et al., 2005). Some earlierobservations of the PLFs were not addressed in the analysis ofHiRISE data (Jaeger et al., 2007), such as the observed interiorcones, which are common in pingos (Burr et al., 2005). And a thin,high-relief rim morphology may not be unique to lava, as a similarmorphology can be seen in MOC images of the Utopia PLFs as well(e.g., de Pablo and Komatsu, 2008, their Fig. 5A and F), for whichpingos are the preferred formation hypothesis, as discussed above.However, lack of comparable (e.g., stereo) HiRISE coverage inUtopia Planitia to that over Athabasca Valles makes thiscomparison uncertain at the present time. The reader is referredto the original references, including a recent comment (Page,2008) and response (Jaeger et al., 2008), for further discussion onthis topic.

3.3.3. Cerberus Palus/western Medusae Fossae Formation

A series of abstracts has proposed the idea that conesapparently eroding out from beneath the western Medusae FossaeFormation (Fig. 16) are pingos (Nussbaumer, 2005; Nussbaumer etal., 2003, 2004). This hypothesis is based on the cones’ similarityin morphology and size to terrestrial pingos. As other features inthe region are interpreted to be glacial, the inferred context is alsoconsistent with the hypothesis that the cones are pingos.

Although the features’ conical morphology is pingo-like, thearguments in these abstracts are not well-developed. Potentialwater pressurization mechanisms for the inferred pingos are notexplained, and the interpretation of other features (e.g., positive-relief linear hills) as glacial (e.g., as glacial fluting) is notconvincingly argued in view of previous and more detailedinterpretations (e.g., as yardangs). Similar PLFs nearby werehypothesized to be rootless cones on the basis of the PLFmorphology and the inference that the underlying substrate waslava (Lanagan and McEwen, 2003; Lanagan, 2004). In our view,the interpretation remains viable, but speculative, with bettersupported interpretations available.

4. Summary

A tabular summary of the data and our evaluations of PLFs onMars is provided in Table 1. Of the Martian landforms that havebeen hypothesized in the literature to be pingos, those in Utopia

Planitia are the most likely candidates to be terrestrial-stylepingos or pingo scars. PLFs reported to date elsewhere on Mars areless likely pingo candidates or are better explained by some othermechanism.

That the landforms most likely to be pingos on Mars areclustered in Utopia basin may be due to general conditions in thisnorthern mid-latitude basin. As discussed above (Section 3.2),Utopia basin has been a spatially extensive site for sedimentaccumulation throughout the Amazonian Period. Except for asingle reported example of a pingo forming in bedrock (Muller,1959), all terrestrial pingos form in unconsolidated sediments. Sosediment accumulation appears generally required for pingoformation. In addition, Utopia basin also experienced repeatedaqueous inflow during the Amazonian. Such surface aqueous flowwould provide near-surface groundwater, which is also generallynecessary for pingo formation.

To date, candidate pingos have not been hypothesized in large,southern mid-latitude basins that experienced sedimentary in-filling and aqueous inflow. Landforms in Hellas and Argyre basinshave been hypothesized to be glacial (e.g., Kargel and Strom,1992). Thus, it would be reasonable for the basins to also hostperiglacial (‘near-glacial’) landforms, such as pingos. To date,however, PLFs in these basins have not yet been identified.A limited number of PLFs have been identified in small (�10-kmdiameter) craters in the southern mid-latitudes (Dundas et al.,2008). Additional data analysis is required to determine the extentof PLFs in the southern hemisphere. Such future research will helpto determine the veracity of this seeming north–south asymmetryin PLF occurrence and, if true, its climatologic, hydrologic, andsedimentologic implications.

Acknowledgements

This work was supported by the Mars Fundamental ResearchProgram. We thank Jim Head and Goro Komatsu for constructivecritiques of the original manuscript that led to significant improve-ments, and Windy Jaeger and Jim Skinner for helpful comments.

References

An, Z., 1980. Formation and evolution of permanent ice mound on Qing-ZangPlateau. J. Glaciology Cryopedology 2 (2), 25–30.

Armstrong, J.C., Titus, T.N., Kieffer, H.H., 2005. Evidence for subsurface water ice inKorolev crater, Mars. Icarus 174, 360–372.

ARTICLE IN PRESS

Fig. 16. Portion of MOC image E1701551 (centered near 31N 142.31E/142.31W),

showing PLFs at the edge of the western Medusae Fossae Formation (seen at upper

left). North is up, and illumination is from the left.



Associate Committee on Geotechnical Research, 1988. Glossary of Permafrost andRelated Ground Ice Terms, Vol. 14. National Research Council of Canada,Ottawa, pp. 156 (Technical Memorandam).

Bacastow, A.L., Sakimoto, S.E.H., 2006. Martian north polar crater morphology:implications for an aquifer. Lunar Planet. Sci. Conf. XXXVII (abstract 2239).

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

Brown, A.J., Byrne, S., Roush, T., Herkenhoff, K.E., Bishop, J.L., Hansen, C., Green, R.O.,Russell, P., McEwen, A., Murchie, S.L., CRISM Team, HiRISE Team, 2007. Highresolution observations of Korolev Crater and Mrs. Chippy’s ring duringsummer by CRISM and HiRISE. Lunar Planet. Sci. Conf. XXXVIII (CD-ROM.abstract 2308).

Bruno, B.C., Fagents, S.A., Hamilton, C.W., Burr, D.M., Baloga, S.M., 2006.Identification of volcanic rootless cones and ice mounds on Earth and Mars:using spatial distribution as a remote sensing tool. J. Geophys. Res. 111 (E6),E06017.

Burr, D.M., Soare, R.J., Wan Bun Tseung, J.M., Emery, J.P., 2005. Young (lateAmazonian), near-surface, ground ice features near the equator, AthabascaValles, Mars. Icarus 178 (1), 56–73.

Cabrol, N.A., Grin, E.A., Pollard, W.H., 2000. Possible frost mounds in an ancientmartian lake bed. Icarus 145, 91–107.

Carter, L.D., Galloway, J.P., 1979. Arctic coastal plain pingos in National PetroleumReserve in Alaska. In: Johnson, K.M., Williams, J.R. (Eds.), The United StatesGeological Survey in Alaska. Accomplishments during 1978, US Geol. Surv.Circular, Vol. 804-B, pp. B33–35.

Chapman, M.G., 1994. Evidence, age, and thickness of a frozen paleolake in UtopiaPlanitia, Mars. Icarus 109, 393–406.

Chapman, M.G., 2003. Sub-ice volcanoes ad ancient oceans/lakes: a Martianchallenge. Global Planet. Change 35, 185–198.

Christiansen, E.H., 1989. Lahars in the Elysium region of Mars. Geology 17,203–206.

Costard, F.M., Kargel, J.S., 1995. Outwash plains and thermokarst on Mars. Icarus114, 93–112.

Cruickshank, G.J., Colhoun, A.E., 1965. Observations on pingos and other landformsin Schuchert Dal, Northeast Greenland. Geogr. Annlr. 47A (4), 224–236.

Davis, N., 2001. Permafrost: A Guide to Frozen Ground in Transition. University ofAlaska Press, Fairbanks, AK, p. 351.

de Pablo, M.A., Komatsu, G., 2007. Pingo fields in the Utopia Basin, Mars: geologicaland climatic implications. Lunar Planet. Sci. XXXVIII (abstract 1278).

de Pablo, M.A., Komatsu, G., 2008. Possible pingo fields in the Utopia Basin, Mars:geological and climatic implications. Icarus 2.

Dundas, C.M., Mellon, M.T., McEwen, A.S., Lefort, A., Keszthelyi, L.P., Thomas, N., theHiRISE Team, 2008. HiRISE observations of fractured mounds: possible Martianpingos. Geophys. Res. Lett. 35 (4), L04201.

Flemal, R.C., 1976. Pingos and pingo scars: their characteristics, distribution, andutility in reconstructing former permafrost environments. Quat. Res. 6, 37–53.

French, H.M., 1996. The Periglacial Environment, Second Ed. Addison WeelsyLongman Limited, Essex, England, p. 341.

Gaidos, E., Marion, G., 2003. Geological and geochemical legacy of a cold, earlyMars. J. Geophys. Res. 108 (E6), 5005.

Ghatan, G.J., Head, J.W., 2002. Candidate subglacial volcanoes in the south polarregion of Mars: morphology, morphometry, and eruption conditions. J.Geophys. Res. 107(E7), 2–1, 5048, doi:10.1029/2001JE001519.

Greeley, R., Fagents, S.A., 2001. Icelandic pseudocraters as analogs to some volcaniccones on Mars. J. Geophys. Res., 20527–20546.

Gurney, S.D., 2001. Aspects of the genesis, geomorphology and terminology ofpalsas: 198 perennial cryogenic mounds. Prog. Phys. Geogr. 25 (2), 249–260.

Hamilton, T.D., Obi, C.M., 1982. Pingos in Brooks Range, Northern Alaska, USA. Arct.Alp. Res. 14, 13–20.

Head III, J.W., Hiesinger, H., Ivanov, M.A., Kreslavsky, A., Pratt, S., Thomson, B.J.,1999. Possible ancient oceans on Mars: evidence from Mars Orbiter LaserAltimeter data. Science 286, 2134–2137.

Holmes, G.W., 1967. Location of pingos and pinoglike mounds observed from theground, from aerial reconnaissance, and on aerial photographs in interiorAlaska. US Geological Survey Open-file Report, 13pp.

Holmes, G.W., Hopkins, D.M., Foster, H.L., 1968. Pingos in central Alaska. US Dept.Interior Geol. Surv. Bull. 1241-H, 40.

Hughes, L.O., 1969. Distribution of open system pingos in central Yukon Territorywith respect to glacial limits. Geol. Surv. Can. Pap. 69–34, 8.

Jaeger, W.L., Keszthelyi, L.P., Burr, D.M., McEwen, A.S., Baker, V.R., Miyamoto, H.,Beyer, R.A., 2003. Ring dike structures in the channeled scabland as analogs forcircular features in Athabasca Valles, Mars. Lunar Planet. Sci. Conf. XXXIV(abstract 2045).

Jaeger, W.L., Keszthelyi, L.P., Burr, D.M., Emery, J.P., Baker, V.R., McEwen, A.S.,Miyamoto, H., 2005. Basaltic ring structures as an analog for ring features inAthabasca Valles, Mars. Lunar Planet. Sci. XXXVI (abstract 1886).

Jaeger, W.L., Keszthelyi, L.P., McEwen, A.S., Dundas, C.M., Russell, P.S., 2007.Athabasca Valles, Mars: a lava-draped channel system. Science 317, 1709–1711.

Jaeger, W.L., Keszthelyi, L.P., McEwen, A.S., Dundas, C.M., Russell, P.S., 2008.Response to comment on ‘‘Athabasca Valles, Mars: a lava-draped channelsystem’’. Science 320, 1588c.

Jorgenson, T., Yoshikawa, K., Kanevskiy, M., Shur, Y., Romanovsky, V., Marchenko, S.,Grosse, G., Brown, J., Jones, B., 2008. Permafrost characteristics of Alaska,Institute of Northern Engineering, University of Alaska Fairbanks, 1:7,200,000scale.

Kargel, J.S., Strom, R.G., 1992. Ancient glaciation on Mars. Geology 20 (1), 3–7.Kuhle, G.M., 1985. Permafrost and periglacial indicators on the Tibetan Plateau

from the Himalaya Mountains in the south to the Quilian Shan in the north(28–401N). Z. Geomorph. N.F. 29 (2), 183–192.

Kuzmin, R.O., Greeley, R, Landheim, R., Cabrol, N.A., Farmer, J.D., 2000. Geologicmap of the MTM-15182 and MTM-15187 quadrangles, Gusev crater—Ma’adimVallis region, Mars. USGS map I-2666.

Lanagan, P.D., 2004. Geologic history of the Cerberus Plains, Mars, Ph.D. Thesis,Tuscon, University of Arizona, p. 142.

Lanagan, P.D., McEwen, A.S., 2003. Cerberus plains volcanism: constraints ontemporal emplacement of the youngest flood lavas on Mars. Sixth InternationalConference on Mars, Pasadena, CA, abstract no. 3215.

Lanagan, P.D., McEwen, A.S., Keszthelyi, L.P., Thordarson, Th., 2001. Rootless coneson Mars indicating the presence of shallow equatorial ground ice in recenttimes. Geophys. Res. Lett. 28 (12), 2365–2367.

Liestøl, O., 1977. Pingos, springs, and permafrost in Spitsbergen. Norsk Polarin-stitutt Arbok, 7–29.

Liestøl, O., 1996. Open system pingos in Spitsbergen. Norsk Geogr. Tidsskr., 50.Leffingwell, E. de K., 1919. The Canning River region, northern Alaska. US Geol. Surv.

Prof. Pap., 109, 251.Mackay, J.R., 1973. The growth of pingos: Western Arctic coast, Canada. Can. J. Earth

Sci. 10, 979–1004.Mackay, J.R., 1979. Pingos of the Tuktoyaktuk Peninsula area, Northwest Territories.

Geogr. Phys. Quat. 23 (1), 3–61.Mackay, J.R., 1986. Growth of Ibyuk Pingo, Western Arctic Coast, Canada, and some

implications for environmental reconstructions. Quat. Res. 26, 68–80.Mackay, J.R., 1987. Some mechanical aspects of pingo growth and failure, western

Arctic coast, Canada. Can. J. Earth Sci. 24, 1108–1119.Mackay, J.R., 1998. Pingo growth and collapse, Tuktoyaktuk Peninsula area, western

arctic coast, Canada: a long-term field study. Geogr. Phys. Quat. 52 (3),271–323.

Matsuoka, N., Sawaguchi, S., Yoshikawa, K., 2004. Present-day periglacialenvironments in central Spitsbergen, Svalbard. Geographical Review of Japan77, 276–300.

Matthews, J.A., Dahl, S.-O., Berrisford, M.S., Nesje, A., 1997. Cyclic development andthermokarstic degradation of palsas in the Mid-Alpine Zone of Leirpullan,Dovrefjell, Southern Norway. Permafrost Periglac. Processes 8, 107–122.

McEwen, A.S., Preblich, B.S., Turtle, E.P., Artemieva, N.A., Golombek, M.P., Hurst, M.,Kirk, R.L., Burr, D.M., Christensen, P.R., 2005. The rayed crater zunil andinterpretations of small impact craters on Mars. Icarus 176, 351–381.

McGill, G.E., 1989. Buried topography of Utopia, Mars—persistence of a giantimpact depression. J. Geophys. Res. 94, 2753–2759.

McSween, H.Y., Ruff, S.W., Morris, R.V., Bell, J.F., Herkenhoff, K., Gellert, R., Stockstill,K.R., Tornabene, L.L., Squyres, S.W., Crisp, J.A., Christensen, P.R., McCoy, T.J.,Mittlefehldt, D.W., Schmidt, M., 2006. Alkaline volcanic rocks from theColumbia Hills, Gusev crater, Mars. J. Geophys. Res. 111 (E9) CiteID E09S91.

Mitchell, G.F., 1973. Fossil pingos in Camaross Townland, Co. Wexford, In:Proceedings of the Royal Irish Academy, vol. 73B(16), pp. 269–281.

Muller, F., 1959. Observations on pingos (in German). Meddelelser om Grønland153, 127 (Translated by D.A. Sinclair, National Research Council of Canada,Ottawa, TT-1073, 1963, p. 117).

Nussbaumer, J., 2005. Extent and further characteristics of former glaciated terrainin Elysium Planitia, Mars. Lunar Planet. Sci. XXXVI CD-ROM (abstract 1949).

Nussbaumer, J., Jaumann, R., Hauber, E., 2003. Evidence for a surfing ice-sheet inElysium Planitia, Mars. Sixth International Conference on Mars CD-ROM(abstract 3018).

Nussbaumer, J., Jaumann, R., Hauber, E., 2004. Evidence of recent glaciation inElysium Planitia, Mars. Lunar Planet. Sci. XXXV CD-ROM (abstract 1244).

O’Brien, R., 1971. Observations on pingos and permafrost hydrology in SchuchertDal, N.E.Greenland. Meddelelser om Grønland 195 (1), 1–20.

Osinski, G.R., Soare, R.J., 2007. Circular structures in Utopia Planitia, Mars: ImpactV. Periglacial origin and implications for assessing ground-ice content. LunarPlanet. Sci. Conf. XXXVIII CD-ROM (abstract 1609).

Page, D.P., 2008. Comment on ‘‘Athabasca Valles, Mars: a lava-draped channelsystem’’. Science 320, 1588b.

Page, D.P., Murray, J.B., 2006. Stratigraphical and morphological evidence for pingogenesis in the Cerberus plains. Icarus 183, 46–54.

Parker, T.J., Gorsline, D.S., Saunders, R.S., Pieri, D.C., Schneeberger, D.M., 1993.Coastal geomorphology of the Martian northern plains. J. Geophys. Res. 98 (E6),11061–11078.

Pissart, A., 1963. Les traces de ‘pingos’ du Pays de Galles et du Plateau des HautesFragnes. Z. Geomorphol. 7, 147–165.

Pissart, A., 1965. Les pingos des Hautes Fragnes. Le probleme de leur genese. Ann.Soc. Geol. Belg. 88, 277–289.

Pissart, A., 2002. Palsas, lithalsas and remnants of these periglacial mounds. Aprogress report. Prog. Phys. Geogr. 26 (4), 605–621.

Pissart, A., French, H.M., 1976. Pingo investigations, north-central Banks Island,Canadian Arctic. Can. J. Earth Sci. 13, 937–946.

Pollard, W.H., van Everdingen, R.O., 1992. Formation of seasonal ice bodies. In:Dixon, J.C., Abrahams, A.D. (Eds.), Periglacial Geomorphology. Wiley, New York,pp. 281–304.

Porsild, A.E., 1938. Earth mounds in unglaciated Arctic Northwestern America.Geogr. Rev. 28, 45–58.

Rodriguez, J.A.P., Tanaka, K.L., 2006. Sisyphi Montes and southwest Hellas Paterae:possible impact, cryotectonic, volcanic, and mantle tectonoic processes alongHellas Basin rings. Fourth Mars Polar Science Conference (Abstract 8066).

ARTICLE IN PRESS



Rodriguez, J.A.P., Tanaka, K.L., Langevin, Y., Bourke, M., Kargel, J., Christensen, P.,Sasaki, S., 2007. Recent aeolian erosion and deposition in the north polarplateau of Mars. Mars 3, 29–41.

Russell, P.S., Head, J.W., 2003. Elysium–Utopia flows as mega-lahars: a model ofdike intrusion, cryosphere cracking, and water-sediment release. J. Geophys.Res. 108.

Russell, P.S., Head, J.W., Hecht, M.H., 2004. Evolution of ice deposits in the localenvironments of Martian circum-polar craters and implications for polar caphistory. Lunar Planet. Sci. Conf. XXXV (abstract 2007).

Sakimoto, S.E.H., 2005a. Central mounds in Martian impact craters: assessment aspossible perennial permafrost mounds (pingos). Lunar Planet. Sci. Conf. XXXVI(abstract 2099).

Sakimoto, S.E.H., 2005b. Martian polar craters: possible polar materials apparenteffects and post-impact modification by apparent perennial permafrost mound(pingo). Role of Volatiles and Atmospheres on Martian Impact Craters (abstract3036).

Scott, D.H., Tanaka, K.L., Greeley, R., Guest, J.E., 1986–87. Geologic maps of thewestern and eastern equatorial and polar regions of Mars. USGS map I-1802-A,B, C.

Seelos, F., Murchie, S., Pelkey, S., Seelos, K., CRISM Team, 2007. CRISM multispectralsurvey campaign—status and initial mosaics. Lunar Planet. Sci. 38 Abstract2336 (CD-ROM).

Serov, A.G., Leshchikov, F.N., 1978. Permafrost in the Territory of the Boguchanskreservoir. Regional hydrogeology and engineering geology of East Siberia.Novosibirsk, Nauka, pp. 1–9.

Skinner Jr., J.A., Tanaka, K.L., 2007. Evidence for and implications of sedimentarydiapirism and mud volcanism in the southern Utopia highland–lowlandboundary plain, Mars. Icarus 186, 41–59.

Skinner Jr., J.A., Skinner, L.A., Kargel, J.S., 2007. Re-assessment of hydrovolcanism-based resurfacing within the Galaxias Fossae region of Mars. Lunar Planet. Sci.Conf. 38 Abstract 1998 (CD-ROM).

Soare, R.J., Burr, D.M., Wan Bun Tseung, J-M., 2005b. Pingos and a possibleperiglacial landscape in Utopia Planitia. Icarus 174 (2), 373–382.

Tanaka, K.L., Scott, D.H., 1987. Geologic map of the polar regions of Mars(1:15,000,000). US Geological Survey Miscellaneous Investigations Series,Map I-1802-C.

Tanaka, K.L., Chapman, M.G., Scott, D.H., 1992. Geologic map of the Elysium regionof Mars (1:5,000,000). US Geological Survey Miscellaneous InvestigationsSeries, Map I–2147.

Tanaka, K.L., Skinner, J.A., Hare, T.H., 2005. Geologic map of the northern plains ofMars (1:15,000,000). US Geological Survey Scientific Investigations Series, MapSIM-2888.

Tanaka, K.L., Rodriguez, J.A.P., Skinner Jr., J.A., Bourke, M.C., Fortezzo, C.M.,Herkenhoff, K.E., Kolb, E.J., Okubo, C.H., 2008. North polar region of Mars:advances in stratigraphy, structure, and erosional modification. Icarus 196,318–358.

Tikhomirov, B.A., 1948. Ogeograficheskom rasprostranenii bugrov-baidzharakhovna severe Evrazii (On the geographical distribution of pingos in northernEurasia). Priroda 37, 51–53.

Wang Shaoling, 1982. Frozen ground and periglacial features in the southeastQinghai Province. 1. Chinese Tibet Expedition 1981. Braunschweig-symposium,14–16/4.

Washburn, A.L., 1969. Weathering, frost action, and patterned ground in theMesters Vig district, Northeast Greenland. Meddeer. Grønland 176 (4), 1–303.

Watson, E., 1971. Remains of pingos in Wales and the Isle of Man. Geol. J. 7,381–392.

Weigand, G., 1965. Fossile pingos in Mitteleuropa. Wurzburger Geogr. Arb. 16, 152.Worsley, P., Gurney, S.D., Collins, P.E.F., 1995. Late Holocene ‘mineral palsas’ and

associated vegetation patterns: a case study from Lac Hendry, NorthernQuebec, Canada and significance for European Pleistocene thermokarst. Quat.Sci. Rev. 14, 179–192.

Yoshikawa, K., 1991. Age of growth two pingos, Saqaq Dalen, West CentralGreenland. Permafrost Periglac. Processes 2, 245–252.

Yoshikawa, K., 1993. Notes on open system pingo ice, Adventdalen, Spitsbergen.Permafrost Periglac. Processes 4, 327–334.

Yoshikawa, K., 2008. Stable isotope composition of ice in seasonally andperennially frozen mounds. In: Kane, D.L., Hinkel, K.M., (Eds.), Proceedings ofNinth International Conference on Permafrost, vol. 2, pp. 1997–2002.

Yoshikawa, K., Harada, K., 1995. Observations on Nearshore pingo growth,Adventdalen, Spitsbergen. Permafrost Periglac. Processes 6, 361–372.

Yoshikawa, K., White, D., Hinzman, L., Goering, D., Petrone, K., Bolton, W., Ishikawa,N., 2003. Water in permafrost: case study of aufeis and pingo hydrology indiscontinuous permafrost. In: Phillips, M., Springman, S.M., Arenson, L.U.,(Eds.), Proceedings, Eighth International Conference on Permafrost, vol. 2,1259–1264.

Yoshikawa, K, Zhao, L., Baisheng, Y.E., 2006a. Historical variability of the icing(aufeis) in the Brooks Range, Alaska and Kunlun Mountain, China, AsianConference on Permafrost Lanzhou, China August 7–9, 2006.

Yoshikawa, K., Leuschen, C., Ikeda, A., Harada, K., Gogineni, P., Hoekstra, P.,Hinzman, L., Sawada, Y., Matsuoka, N., 2006b. Comparison of geophysicalinvestigations for detection of massive ground ice (pingo ice). J. Geophys. Res.111, E06S19.

Yoshikawa, K., Hinzman, L.D., Kane, D.L., 2007. Spring and aufeis (icing) hydrologyin Brooks Range, Alaska. J. Geophys. Res. 112, G04S43.

ARTICLE IN PRESS


author's personal copy - eps.utk.edu

Documents