Candidate subglacial volcanoes in the south polar region ... · Morphology, morphometry, and eruption conditions ... and large summit craters relative to the summit ... trending slightly

Candidate subglacial volcanoes in the south polar region of Mars:

Morphology, morphometry, and eruption conditions

Gil J. Ghatan and James W. Head IIIDepartment of Geological Sciences, Brown University, Providence, Rhode Island, USA

Received 17 May 2001; revised 19 February 2002; accepted 27 March 2002; published 12 July 2002.

[1] A number of isolated mountains mapped in the south polar region of Mars aredocumented using Mars Global Surveyor (MGS) Mars Orbiter Laser Altimeter (MOLA)and Mars Orbiter Camera (MOC) data. These mountains have average separationdistances of �175 km, they are typically 30–40 km in diameter and 1000–1500 m high,and their bases fall near an elevation of �1200 m. A significant number of the populationare located on or very close to a 660 km long line extending toward the south pole. Thesummits of a number of these mountains have unusual shapes: flat-topped, flat-toppedwith a cone, and large summit craters relative to the summit diameter. Sinuous channelsare found in association with the margins of several of the mountains. Several modes oforigin are considered for these mountains, including impact, tectonic, and volcanic. Thesemountains occur in intimate association with the Hesperian-aged Dorsa ArgenteaFormation, a unit containing features and structures interpreted to be related to a thick,areally extensive ice sheet. On the basis of these observations and analyses it is concludedthat the most plausible origin for these mountains is volcanic and that they representextrusion of lava from vents, many of which lay underneath an ice sheet. The unusualshape of many of the mountains is consistent with construction of a volcanic edificeunderneath an ice sheet and melting of adjacent ice-rich material, sometimes formingdrainage channels. The topography of these mountains suggests that the ice sheet averagedat least 1.4 km thick at the time of eruption. These features appear to be associated withEarly Hesperian-aged ridged plains (Hr). INDEX TERMS: 6225 Planetology: Solar System

Objects: Mars; 5480 Planetology: Solid Surface Planets: Volcanism (8450); 5416 Planetology: Solid Surface

Planets: Glaciation; 5462 Planetology: Solid Surface Planets: Polar regions; KEYWORDS: Mars, south polar,

Dorsa Argentea Formation, subglacial volcanism, Hesperian

1. Introduction

[2] In their mapping of the south polar region of Mars,Tanaka and Scott [1987] documented 17 features locatedbetween 20� and 340�W (Figure 1a) as (m), described asmountains of a high and rugged nature, primarily foundwithin the Dorsa Argentea Formation (Hd). These moun-tains were first observed in Mariner 9 images [Murray et al.,1972] and were mapped by several authors [Peterson, 1977;Condit and Soderblom, 1978; Scott and Carr, 1978]. Due tothe resolution of Mariner images and lack of high-resolutiontopography, a definite age and origin for the mountains werenot able to be determined. Several possible origins wereproposed, such as remnants of crater rims, central peaks,and volcanoes [Murray et al., 1972; Peterson, 1977; Conditand Soderblom, 1978; Scott and Carr, 1978]. Even with thehigh-resolution images from the Viking mission, a moredetailed assessment of the age and origin of the mountainswas not possible [Tanaka and Scott, 1987]. Also mapped byTanaka and Scott [1987] in the south polar region wereseveral features labeled as (v), confidently interpreted as

volcanoes, and (v?), features that appeared to be volcanoesbut whose origin was less certain. The ages of thesevolcanoes were also unknown. A single occurrence of(v?) was mapped in the region containing the mountains(Figure 1a).[3] Using new Mars Orbiter Laser Altimeter (MOLA)

and Mars Orbiter Camera (MOC) data from the MarsGlobal Surveyor Mission, we have been able to reexaminethese features at much higher resolution than previouslypossible. We have performed morphometric and morpho-logic analyses of the mountains and the feature interpretedto be a volcano in the region that reveal in great detail theirshapes, dimensions, and slopes [Ghatan and Head, 2001;Head and Ghatan, 2001]. We have further investigated therelationship of the structures to the surrounding plains,along with their association with nearby features, andexamined their detailed characteristics where MOC imageswere available.[4] We examine a range of hypotheses for the origin of

the mountains (e.g., impact cratering, tectonism, volcan-ism), addressing those origins which were previously pro-posed by other authors. Detailed mapping of the regionstrongly suggests that the middle Hesperian Dorsa ArgenteaFormation (Hd) mapped by Tanaka and Scott [1987] is a

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 107, NO. E7, 10.1029/2001JE001519, 2002

Copyright 2002 by the American Geophysical Union.0148-0227/02/2001JE001519$09.00

2 - 1

Figure

1.

(a)Portionofthegeologic

map

ofthesouth

polarregiongeneralized

from

Tanaka

andScott[1987]and

showingthedistributionofthefeaturesmapped

asmountains(m

),shownin

green.Other

unitsincludeHd(theDorsa

ArgenteaForm

ation),

HNu

(undivided

material),Npl(rough,heavily

cratered

terrain),

V(volcanoes),

Hpl(smooth,

moderatelycratered

terrain),andAm

(mantlematerialfoundwithin

Cavi).Themap

coversfrom

�20�W

to340�W

and

55�S

to75�S.Themountainsarelocatedpredominantlywithin

Hdorin

close

associationwithit.(b)MOLA

color

topographyoverlain

onMOLAshaded

relief

map

oftheregionshownin

Figure1a.Only

those

areasabove800m

elevation

havecolortopographyoverlain,withredindicatinghighelevationsandblueindicatinglow.See

colorversionofthisfigure

atbackofthisissue.

2 - 2 GHATAN AND HEAD: MARTIAN SOUTH POLAR REGION SUBGLACIAL VOLCANOES

volatile-rich unit and represents the formerly large extent ofa south polar ice sheet [Head and Pratt, 2001]. Since theage and origin of the mountains are unknown, there may bea relationship between them and this deposit. We find that anumber of the mountains are excellent candidates forsubglacial volcanoes, having most likely erupted into theHesperian-aged ice sheet. Evidence exists for (1) a sequenceof morphologies related to interaction with the overlyingice-rich layer and (2) drainage of some meltwater intoadjacent low-lying areas.

2. Description and Characterization

2.1. The Mountains

[5] The 17 mountains mapped by Tanaka and Scott[1987] occur within an area 930 km � 250 km, centeredat 65�S and 355�W (Figure 1a). Using MOLA data, it ispossible to construct a view of the topography of the regionin which the mountains are located (Figure 1b). From thisview, four additional features that meet the criteria ofTanaka and Scott [1987] are found that were not discernablefrom original Viking images, bringing the total number ofmountains to 21 (Figure 1a). Also, the feature mapped as(v?) in this region does not appear in the high-resolutiontopographic data. Distances between the mountains rangefrom 64 to 430 km, with an average separation from theirnearest neighbor of 175 km. The mountains are predom-inantly located within a broad, regional low, roughly corre-sponding to the local deposits of the Dorsa ArgenteaFormation (Figures 1a and 1b).[6] A clear linearity exists in the location of the moun-

tains (Figures 1b and 2a). Specifically, mountains 2, 4, 5, 6,7, 11, and 13 form a line �660 km in length, with most ofthe remaining mountains positioned within �100 km of theline. This line, trending slightly NW-SE (58�S, 358�W to69�S, 355�W), extends poleward to a large mountainousarea �100 km south of South crater. The large mountainousarea, about 60 km � 200 km in dimension, is located mostlyin the formation Tanaka and Scott [1987] mapped asHesperian-Noachian undivided (HNu) (Figure 1a).[7] From the MOLA data the physical characteristics of

the mountains were measured (Table 1). The heights of themountains are generally centered between 1 and 1.5 km(mean of 1.17 km), with some minor variation above andbelow this central range (Figure 3a). The lowest of the groupare mountains 1, 2, 9, 16, and 17. Mountains 1 and 2 are thefeatures found farthest from the cap but are part of thealignment. Mountains 9, 16, and 17 are the three featuresfound farthest from the alignment, situated 392, 680, and530 km away, respectively. Mountain 13 is the tallest, with aheight of 2.2 km, and is located near the middle of the DorsaArgentea Formation.[8] The basal diameters of the mountains range from �20

to 70 km (mean of 40.3 km) and tend to be clusteredbetween 30 and 40 km (Figure 3b). The main exceptions aremountains 3, 16, 18, 20, and 21. Mountains 3 and 18 are thetwo widest features and lie outside of the Dorsa ArgenteaFormation boundaries. Mountain 18 was unmapped byTanaka and Scott [1987], has a width of 70 km, and issituated 415 km away from the alignment. Mountain 21,also unmapped, has a width of 49 km. It falls at the edge ofthe Dorsa Argentea Formation adjacent to an impact crater

that is 25 km in diameter and is located 160 km away fromthe alignment. Mountain 16 has a width of 52 km. Mountain20 has the smallest diameter, 20 km, and sits 190 km awayfrom the alignment.[9] The mountains generally have a narrow range of basal

elevations centered around 1200 m (mean of 1176 m)(Figure 3c). The two exceptions are mountains 12 and 16,which have basal elevations of 1400 and 200 m, respec-tively. Mountain 12, whose basal elevation is near 1200 m,is located 160 km from the alignment. The level of thesurrounding regional plains, located at 1200 m elevation,consists of the Dorsa Argentea Formation (Hd) and a unit ofunconsolidated material mapped as HNu.[10] The range of summit elevations is between 750 and

�3100 m (mean of 2345 m), but over 60% have summitelevations between 2 and 3 km (Figure 3d). Here again themain exception is mountain 16, which has a summitelevation of only 750 m.[11] Volumes of the mountains were determined from the

MOLA data. Volumes range from 75 to 1964 km3 andaverage around 700 km3, with some variation above andbelow. The main exceptions are mountains 1, 2, and 20,with volumes around 150 km3, and mountains 15, 18, and21, with volumes of 1935, 1964, and 1134 km3, respec-tively. Mean flank slopes were also measured for themountains, with values ranging from 2.4� to 13.7� andclustering between 6� and 11�. The three exceptions aremountains 2, 9, and 16, with slopes of 2.4�, 3.96� and 4�,respectively.[12] To summarize the above physical descriptions

(Table 1), the heights of the mountains are generallybetween 1 and 1.5 km (Figure 3a), while the widths areclustered between 30 and 40 km (Figure 3b). The basalelevations are clustered at �1.2 km (Figure 3c), and thesummit elevations are clustered between 2 and 3 km(Figure 3d). Typical volumes are �700 km3, and flankslopes are between 6� and 11�. Those mountains thatexhibit some variation from one or more of the abovetrends tend to be separate from the main alignment of themountains (for example, mountain 16) or are farther awayfrom the cap (for example, mountains 1 and 2).[13] While remarkably similar in morphometric charac-

teristics, the mountains do display some variation in theirmorphology. The various morphologies present can be seenin MOLA shaded relief views (Figure 4) and the corre-sponding profiles (Figure 5) as well as in MOC wide-angleimages (Figure 6). The overall shapes of the mountains fallinto five categories: (1) low-domed, (2) flat-topped, (3) flat-topped with a cone, (4) cone-shaped with a summit crater,and (5) cone-shaped. The mountains predominantly fall intoone of the cone-shaped varieties (4 and 5). The low-domemountains are found only outside the boundary of the DorsaArgentea Formation. This variety contains two of the threemountains that are farthest away from the main alignment(mountains 9 and 16) as well as those farthest from the cap(mountains 1 and 2). The mountains that are cone-shapedwith a summit crater are found either just inside or justoutside the boundary of the Hd. The summit craters tend tobe large relative to the summit diameter of the mountain andreach depths up to �200 m. The cone-shaped mountains arelocated within the center of the Dorsa Argentea Formationof the region. The exception to this is mountain 17, which is

GHATAN AND HEAD: MARTIAN SOUTH POLAR REGION SUBGLACIAL VOLCANOES 2 - 3

located just outside Hd in the unit mapped as Npl [Tanakaand Scott, 1987]. The flat-topped varieties are scattered,falling both within and outside of Hd. However, all arelocated north of 66�S (Figures 1a, 1b, 2a, and 7).

2.2. Surrounding Features and Associations

[14] Surrounding many of the mountains are what appearto be shallow depressions, circularly shaped and extendingfor about one diameter of the mountain in every direction.This is most clearly seen among mountains 2, 4, 5, 6, 7, 10,

11, and 17 (Figure 2b). At their deepest the depressions are50–100 m below the level of the regional plains. Thedepressions then gently grade up into the level of the plains.The exception is mountain 17, with a depression reachingdepths of 250–300 m. Also, the depression around moun-tain 11 extends out from the mountain for only �10 km.There is no lip to the ‘‘rim’’ of the depressions.[15] Seen extending away from many of the mountains

are channels that are carved into the adjacent regional plains(for example, the Dorsa Argentea Formation). The best

Figure 2. (a) Perspective view of aligned mountains looking toward the current south polar cap. MOLAcolor topography is overlain on MOLA shaded relief map. Red is high, and purple is low. Note thecircular depressions around many of the mountains, which stand significantly lower then the mountains.Also note the various morphologies of the mountains. Vertical exaggeration is 70�. (b) Enlarged view ofportion of area containing the mountains. MOLA color topography is overlain on MOLA shaded reliefmap; red is high, and purple is low. Note the channel (indicated by arrows) emanating away from the areanear mountains 5, 6, and 7. Also note the circular depressions and marginal troughs surrounding thedifferent mountains. See color version of this figure at back of this issue.


example of these channels is seen with mountains 5, 6, and 7(Figure 2b). The channel (indicated by the arrows) appearsto breach a narrow, rocky rim surrounding the three moun-tains and to extend away from it for �250 km. In addition,areas of low-lying terrain are seen in Figure 2b surroundingmountains 8, 9, 10, 18, and 21. These marginal troughs areabout as deep as the circular depressions and also gentlygrade back into the level of the surrounding plains. How-ever, the troughs are more expansive and are not restricted tothe symmetrical shape of the circular depressions.[16] Surrounding some of the mountains, fine-scale pol-

ygonal patterns can be found in MOC images. The polygonsare <100 m across. MOC image coverage of the mountainsis sparse, but from the available images, polygons appear tobe present in the low-lying areas, especially in the shallowdepressions surrounding the mountains. This is best seenwith mountain 11 (Figure 8), for which a MOC imagetraverses part of the surrounding depression, and then ontosome of the surrounding plains. Polygons are seen only inthe depression.

3. Discussion and Interpretation

[17] In order to interpret the origin and evolution of themountains, we consider features that are found on Marswith similar physical characteristics and spacing. WhatMartian phenomena could lead to a series of topographichighs, aligned as many of the mountains are? Possiblecandidates that have been previously proposed are remnantsof impact crater rims, central peaks, and volcanoes [Murrayet al., 1972; Peterson, 1977; Condit and Soderblom, 1978;

Scott and Carr, 1978]. A fourth possible origin, not pre-viously considered, is rootless volcanic cones (oftenreferred to as pseudocraters). Each shall be assessed interms of its ability to explain the three key factors pertainingto the mountains: their morphologies, their morphometriccharacteristics, and their spacing and alignment.

3.1. Remnants of Impact Crater Rims

[18] Impact craters on Mars show a wide range of sizesand complexities [e.g., Mouginis-Mark, 1979; Barlow,1988, 1990; Strom et al., 1992], dependent on the size,speed, and angle of the impactor, along with the materialcomposing the bolide and the substrate [e.g., Melosh, 1989;Strom et al., 1992]. The generally circular rim can bedegraded through fluvial, eolian, tectonic, or volcanicmeans or through bombardment by later impacts. Suchdegradation could remove all but the most resistant partsof a crater rim, leaving behind an isolated mesa. Cratersmay also be buried beneath sediment and then laterexhumed, leaving much of the rim intact [Grant andSchultz, 1990].[19] Remnants of impact crater rims do not compare well

to the characteristics of the mountains. On the basis of thephysical dimensions of the mountains, if these featuresformed as parts of crater rims, then the craters would havehad to have been rather large in size; a crater severalhundred kilometers in diameter would be necessary tocreate a rim at least as wide as the mountains, and littleevidence is seen of the former presence of such features.[20] For some of a crater rim to remain while the rest

erodes away is indicative that the remaining part of the rim

Table 1. Physical Characteristics of the Mountains

MountainApproximateVolume, km3

AverageSlope, deg

ApproximateHeight, km

ApproximateWidth, km

SummitElevation, m

BaseElevation, m Shape

1 142.9 5.4 0.47 28.8 1680 1210 low-dome2 215.5 2.41 0.479 37.2 1668 1189 low-dDome3 2671.7 5.42 1.022 72-124 2376 1354 flat-topped4 583 7.76 1.277 36 2608 1331 cone-shaped

with summitcrater

5 557.8 4.83 0.699 37.8 1971 1272 flat-topped6 455.4 9.14 1.455 34.2 2784 1329 flat-topped

with cone7 344.9 8.34 0.971 33.6 2250 1279 cone-shaped

with summitcrater

8 415.6 11.29 1.533 29.4 2687 1154 cone-shaped9 231.9 3.96 0.676 37.8 1839 1163 low-dome10 277.9 10.4 1.157 27 2345 1188 flat-topped11 397 7.46 1.79 40.8 2961 1171 flat-topped

with cone12 366.3 8.36 1.226 30 2627 1401 cone-shaped13 592.2 10.76 2.248 31.2 3412 1164 cone-shaped14 462.4 7.39 1.032 46.2 2172 1140 other15 1934.5 5.45 1.735 45.6 2940 1205 cone-shaped16 529.3 4 0.554 51.6 754 200 low-dome17 204.8 9.03 0.846 24 2178 1332 cone-shaped18 1964 6.45 1.362 69.6 2497 1135 flat-topped19 410.92 7.01 1.091 37.2 2013 922 cone-shaped

with summitcrater

20 75 11.51 0.995 20.4 2358 1343 cone-shapedwith summit

crater21 1133.5 13.71 1.919 49.8 3135 1216 other


is composed of a more resistant material. The chance thatthe number of large impacts required to form the mountainswould have occurred in the same region of Mars atapproximately the same time, and that resistant parts ofthe various crater rims would have been aligned to lead tothe present alignment of the mountains, seems highlyunlikely. Further, almost all the mountains display a clearcircular outline to their structure and decrease in diameterapproximately symmetrically up the structure toward itspeak. This degree of regularity seems unlikely for remnantsof a crater rim.

3.2. Central Peaks

[21] As crater size increases, the physical characteristicsof a crater become more complex [Pike, 1980] with the

onset of central peaks, flat floors, and wall terraces (hencetheir designation as a complex crater). The onset diameterfor craters displaying central peaks on Mars ranges from2 km [Wood et al., 1978] to 5 km [Cintala, 1977] but isgenerally is about 3–3.5 km [Pike, 1980].[22] Upon initial inspection, central peaks might well

appear to be a suitable explanation for the mountains. Theyare clearly topographic highs and are broadly similar inshape to the mountains (Figure 9a). However, detailedexamination shows that the dimensions of the mountainsare different from those of Martian central peaks [Stromet al., 1992; Hale and Head, 1981; Pike, 1985]. Estimatesof the diameter of the craters that should be associated withcentral peaks similar in size to the mountains [Hale andHead, 1981] are several tens of kilometers greater than the

Figure 3. (a) Histogram of the heights of the mountains. The majority of the mountains have heightsbetween 1 and 1.5 km. (b) Histogram of the widths of the mountains. The majority of the mountains havewidths between 30 and 40 km. (c) Histogram of the basal elevations of the mountains. The majority of themountains have basal elevations around 1200 m. (d) Histogram of the summit elevations of themountains. The majority of the mountains have summit elevations between 2 and 3 km.


diameter of the circular depression found around many ofthe mountains (Table 2). For some mountains the expectedcrater diameter is greater than three times the diameter ofthe depression. If the circular depressions are the remains ofcrater rims, they are severely degraded. It would be highlyunlikely that the source of the erosion of the crater rimswould leave the central peaks as well preserved as the

mountains are found to be. Degraded craters are found inthe surrounding plains that are of the size necessary toproduce central peaks having the width of the mountains;however, all of these craters lack any peaks. Those cratersthat do contain peaks are crisp, <100 km in diameter, andhave peaks that are much smaller in size than the mountains(Figure 9a).

Figure 4. (a) Gallery of cut outs from MOLA shaded relief map, displaying the various morphologiesseen among the mountains. Each box is 120 km � 120 km. (b) Annotated sketch maps for each boxshown in Figure 4a. Seen around mountains 2, 5, and 7 are circular depressions. Also, small channels canbe seen emanating from the area surrounding the some of the mountains.


[23] The depth of the expected craters would be estimatedto be between 1 and 3 km [Strom et al., 1992], significantlydeeper than the depth of the circular depressions, which areonly 100 m at their deepest. Also, the alignment of themountains is not explained by a central peak origin (Figures1a and 2a); it would be improbable to have as many cratersof the size required to produce such peaks lined up as theywould need to be. Further, the mountains stand hundreds ofmeters higher (Figures 1b and 2a) than any possible ‘‘rim,’’not a usual characteristic of central peaks (Figure 9a).

[24] For all the above reasons, a central peak origin doesnot appear consistent with the physical parameters of thevast majority of the mountains. The only exception might bemountain 17 (Figure 9b). This mountain is one of thefarthest of the mountains from the alignment (530 km)and is not located within the Dorsa Argentea Formationbut is situated in the Noachian cratered Unit (Npl) (Figure1a), surrounded by craters of similar size containing centralpeaks. However, the peaks found within the craters nearmountain 17 are smaller in size than mountain 17. Further-

Figure 5. Series of MOLA topographic profiles, corresponding to the mountains shown in Figure 4,displaying the various morphologies found among the mountains. Also shown for context is the level ofthe regional plains, which consist predominately of Hd. Vertical exaggeration is 45�.

Figure 6. Gallery of MOC wide-angle images, displaying the various morphologies seen among themountains. From upper left to lower right, images are M11-02780, M03-00362, M04-00561, M07-03961, M03-07301, and M04-00913. Each image has a resolution of about 232 m/pixel.


more, the depth of the circular depression around mountain17 is only a few hundred meters, whereas the depths ofnearby craters of diameters similar in size to that of thecircular depression around mountain 17 are over 1 km. Ifthis mountain is a central peak remnant, significant degra-dation of the crater has occurred and the central peak musthave been enlarged through subsequent deposition or con-struction.

3.3. Rootless Volcanic Cones

[25] The flow of lava over a water-rich substrate canlead to the formation of crater-shaped volcanic cones [e.g.,Thorarinsson, 1953; Greeley and Fagents, 2001]. Phreaticeruptions, resulting from the interaction between the lavaand water, occur, followed by the accumulation of pyro-clastic material into small, rootless constructs. The line-arity of a lava flow might then lead to a linearity amongthese rootless volcanic cones. However, the dimensionsand shapes of the mountains are not at all consistent withrootless cones observed on Earth [Thorarinsson, 1953] orsimilar features observed on Mars [Frey et al., 1979; Freyand Jarosewich, 1982; Lanagan et al., 2001; Keszthelyiet al., 2000; Greeley and Fagents, 2001; Fagents et al.,2002]. Features interpreted to be rootless volcanic coneson Mars, such as those located near the Olympus Mons

Aureole [Hodges and Moore, 1994], have widths rangingup to 1 km. The smallest width found among the moun-tains is �20 km.

3.4. Volcanoes

[26] Volcanoes on Mars come in many different shapesand sizes, some of which are similar to the mountains.Specifically, Uranius [Hodges and Moore, 1994] andZephyria Tholi [Stewart and Head, 2001] (Figure 10)have similar morphometric characteristics to the moun-tains, their heights, widths, and volumes all falling withinthe ranges established for the mountains (Figure 11). Thesummit craters found among at least four of the mountains(Table 1) could be interpreted as collapse calderas [Crum-pler et al., 1996]. The alignment of some of the mountainscould also be explained by dike swarms propagating to thesurface [Wilson and Head, 1994, 1998, 2000, 2001a].[27] Many volcanic edifices show evidence for lateral

dike propagation, commonly in the form of flanking riftzones and flank eruptions. If the mountains are volcanoes,the lack of a significant edifice in the vicinity suggests thatthey and their alignment represent primarily vertical dikepropagation from depth, and are not the result of lateral dikepropagation from a major edifice reservoir [e.g., Wilson andHead, 1994]. The spacing and total distance covered by the

Figure 7. Map of the same region displayed in Figure 1a, showing the distribution of the variousmorphologies of the mountains. Mountains 14 and 21 are not shown, since their morphologies do not fallinto one of the main categories. The Dorsa Argentea Formation lies within and poleward of the solid line.


mountains are empirically similar to those of some of thesmaller Tharsis Montes [Scott and Tanaka, 1986] andvolcanic edifices observed in Iceland [Bjornsson andEinarsson, 1990].[28] An interpretation related to volcanic construction

appears to explain both the morphometry of the mountainsand their relatively isolated state from each other. However,as mentioned earlier, while similar in most morphometric

descriptions, the mountains do display some variation intheir morphology from each other (Figures 4, 5, and 6). Inaddition, they display some variation in their morphologyfrom the above mentioned Martian volcanoes (Figure 10).Specifically, the slopes of a number of the mountains tend tobe steeper than those of similar sized Martian volcanoes[Kortz and Head, 2001; Head et al., 1998; Smith et al.,1998], leading to broader summits for the non-cone-shaped

Figure 8. Portion of MOC narrow-angle image M07-05379, which traverses part of the circulardepression around mountain 11 (upper portion of image) and then moves up onto the surrounding plains(lower portion of image). Clearly seen is the restriction of the polygonal patterns to the low-lying areas.


mountains and high width-to-height ratios; this also leads tomany mountains that are quite conical. The exceptions tothis are the low-domed mountains, which together accountfor the lowest four slopes among the mountains. In addition,for those mountains containing summit craters, the craterstend to be large relative to the mountain [Crumpler et al.,1996], in some cases dominating the entire summit. Also,the depths of the summit craters reach depths of only�200 m. On the basis of these characteristics and compar-isons, we believe that the interpretation of the mountains asvolcanoes is most consistent with the data. The severalmorphological variations from other Martian volcanoes ofsimilar size, as well as the surrounding features found inassociation with the mountains, may hint at an environment

of eruption different from that typical of the subaerialedifices elsewhere on Mars.[29] The location of these mountains and their surround-

ing environment may provide clues as to their origin. Mostof the mountains are situated within the Hesperian-agedDorsa Argentea Formation (Hd), whose origin has previ-ously been attributed to eolian mantle or lava flows [Tanakaand Scott, 1987]. The relation of the mountains to Hd isintimate in that they are predominantly surrounded by it(Figure 1) and the common basal elevation shared by themountains, �1.2 km, is the same as the level of the top ofsurrounding plains, largely the Dorsa Argentea Formation.Recent detailed mapping of the Dorsa Argentea Formation[Head and Pratt, 2001] supports the interpretation that it isa volatile-rich unit, and it has been interpreted as represent-ing the remnants of a formerly larger south polar ice sheet.This interpretation is supported by the presence of esker-likeridge systems [Head and Hallet, 2001a, 2001b], large-scalecollapse features, flat marginal areas interpreted to beponded debris, and sinuous channels leading from themargins of the deposit and interpreted to be drainage frommeltback products [e.g., Head et al., 2001; Milkovich et al.,2002]. An alternative interpretation is presented by Kolband Tanaka [2001].[30] Here we use the terms ‘‘ice cap’’ and ‘‘ice sheet’’

according to Head and Pratt [2001], following the terres-trial convention [e.g., Benn and Evans, 1998]. Ice cap willrefer to an ice-rich body <50,000 km2 in area, such as thecurrent Amazonian-aged south polar ice cap. Ice sheet willrefer to an ice-rich body >50,000 km2 in area, such as the

Figure 9. Two portions of a MOLA shaded relief map of the region, comparing the mountains to nearbycraters with central peaks. (a) Seen is a crater with diameter similar to that of the circular depressionsaround the mountains, whose central peak is smaller, and has a different appearance than that of themountains. (b) Seen is a comparison view mountain 17 and its circular depression with a nearby cratercontaining a central peak.

Table 2. Comparison of the Circular Depressions to Estimated

Crater Diameter

Mountain Expected CraterDiameter,a km

Diameter of CircularDepression, km

1 107 732 138 1194 134 855 141 1206 127 817 125 8710 101 6211 152 5117 89 49

aHale and Head [1981].


Hesperian-aged ice sheet. The composition of the residualice (Api) and layered terrain (Apl) of the Amazonian-agedMartian ice caps is not definitely known [Thomas et al.,1992]. It is known that there is a volatile component and adust component. The volatile component has been argued tobe mostly water ice [Clifford, 1987, 1993; Mellon, 1996],mostly CO2 ice, or any ratio in between, including CO2

clathrates [Kargel, 1998; Hoffman et al., 2002]. However, itis generally accepted that water is the dominant volatilecomponent. The dust component is inferred from albedo[Herkenhoff and Murray, 1990; Thomas et al., 1992] andfrom images such as those of the layered terrain [Murray etal., 1972; Cutts, 1973], which clearly show alternating darkand light layers. This is mentioned because a Martian ice

cap or ice sheet is not necessarily a pure ice body as may bethought to be the case with their terrestrial analogs. In arelative sense, it is accepted that Api has a high ice-to-dustratio [Kieffer, 1990; Mellon, 1996] and Apl has a lower, butstill high, ice-to-dust ratio [Cutts et al., 1979].[31] What might be said about the volatile component of

the Hesperian-aged south polar ice sheet? What is left ofthis ice sheet is what has been mapped as the DorsaArgentea Formation and related unit HNu and will bereferred to as the deposits of the ice sheet. These units havebeen shown by Head and Pratt [2001] to be volatile rich,and as mentioned above, evidence exists for meltback anddrainage. The meltback that occurred indicates that volatileswere lost and that the larger ice sheet once had a higher ice-

Figure 10. (a) Portion of Viking MDIM showing the morphology of Uranius Tholus. Center of image isat 26.5�N and 98�W. (b) Viking high-resolution image showing the morphology of Zephyria Tholus.Center of image is located at 20�S and 187�W. Image number F430S023 (70 m/pixel). (c) MOLAtopographic profile across Uranius Tholus. Orbit number 10832.4. Vertical exaggeration is 15�. (d)MOLA topographic profile across Zephyria Tholus. Orbit number 12263.2. Vertical exaggeration is 15�.


to-dust ratio than the deposits now have. What exactly thisratio was is unknown. However, estimates made by Headand Pratt [2001] of the thickness of the ice sheet suggestthat in places it was as much as 2500 m thick. Such avoluminous loss of material from the area suggests that themajority of the loss was volatile in nature and that theHesperian-aged ice sheet likely had a high ice-to-dust ratio.[32] The presence of an expansive ice sheet at the south

pole during the Hesperian may have influenced the con-struction and evolution of the mountains. Since the ages ofthe mountains are unknown, the possibility exists that thesefeatures interpreted to be volcanic edifices formed byeruption into or beneath the formerly larger Hesperian icesheet. We now examine this possibility to see if it might beable to explain the variations of the morphologies of themountains from some Martian volcanoes of similar size.

4. Subglacial Eruptions

4.1. Terrestrial Subglacial Eruptions

[33] Subglacial eruptions on the Earth are perhaps bestrecognized by the nature of table mountain structures suchas those located in Antarctica [Smellie and Skilling, 1994;Skilling, 1994; Smellie, 2000, 2001], British Columbia[Mathews, 1947, 1951], Alaska [Hoare and Coonrad,1978], and Iceland [Nielson, 1937; Van Bemmelen andRutten, 1955; Einarsson, 1966] and by the rapid drainageof large volumes of meltwater through jokulhlaups [Majorand Newhall, 1989], most prominent in Iceland [Bjornsson,1974, 1975, 1977, 1992; Gudmundsson et al., 1995; Jons-son, 1982; Nye, 1976; Thorarinsson, 1957; Tomasson,1974, 1996]. However, the shape of the construct developedfrom eruption into an ice sheet is not restricted to the low,flat-topped mounds of table mountains [Nielson, 1937;Einarsson, 1966; Hickson, 2000]. Conical features can alsoresult from eruption into an ice sheet [Nielson, 1937;Mathews, 1947; Hickson, 2000]. These structures are

known as subglacial mounds. As with most terrestrialanalogs to Martian structures, table mountains (also referredto as tuyas) and subglacial mounds on Earth are signifi-cantly smaller than the Martian features we interpret toshare a similar genetic origin. Terrestrial subglacial moundsare documented as reaching heights of up to 1 km andvolumes of only a few cubic kilometers [Hickson, 2000].[34] The morphological form of the structure built by

subglacial eruptions is controlled by a number of factors[e.g., Wilson and Head, 2001b], such as the thickness of theice sheet [Smellie and Skilling, 1994], type of ice sheet(polar versus temperate), the volume of meltwater producedand confined by the ice [Smellie, 2001], type of eruption(localized vent versus fissure eruption), size of cauldronformed on the ice surface, composition of the magma, rateof eruption, and volume of magma erupted [Hickson, 2000;Smellie, 2000]. Such differences in morphology are docu-mented even for the same eruption, the 1934 eruption ofGrimsvotn, Iceland, where two eruptive centers within afew kilometers of each other produced edifices of differentmorphologies [Nielson, 1937].[35] Much of a subglacial eruption actually takes place

subaqueously. A general sequence of events for a subglacialeruption is as follows (e.g., Figure 12). During the initialstages of eruption, large volumes of meltwater are producedby the increase in magmatic heat. This leads to formation ofa depression in the surface of the overlying ice [Bjornsson,1974; Gudmundsson et al., 1997]. Glacier hydrology ispredominantly controlled by the topography of the icesurface [Nye, 1976; Benn and Evans, 1998]. The depressionin the ice surface will cause the meltwater to drain towardthe site of eruption and to collect around the forming edificeas an englacial lake [Bjornsson, 1975; Smellie, 2001]. Theenglacial lake is enclosed on all sides by the ice wallbarriers.[36] Further eruptions occur into the englacial lake,

which quenches the lava, makes it less mobile, and causesit to either crumble or form pillows [Gudmundsson et al.,1997; Smellie and Skilling, 1994; Skilling, 1994; Smellie2000, 2001]. The quenched lava acts as a good insulator,keeping further erupted lava out of contact with the ice. Theeasiest way for lava to flow in this englacial lake is side-ways, where it is confined by the ice walls, thus causing themound to have high width-to-height ratios and relativelysteep slopes [Einarsson, 1966]. If the mound reaches aheight sufficient to breach the level of the englacial lake,effusive subaerial capping flows will occur, leading to arelatively flat top, making the edifice less susceptible toerosion [Smellie and Skilling, 1994; Skilling, 1994; Smellie2000, 2001]. If the eruption turns subaerial, pyroclasticactivity can also occur, and a cone of material can beformed on top of the otherwise low, and often flat-topped,mound [Allen, 1979]. This pyroclastic activity could be dueto either release of volatiles from the magma or phreato-magmatic interactions with the surrounding meltwater.Alternatively, it may date to a time after the edifice hasbeen constructed when the ice sheet has retreated [Allen,1979]. If the surface of the englacial lake is of sufficientheight not to be breached, an entirely conical feature can beformed [Hickson, 2000]. Important in the constructionalhistory of the volcano are the stability of the englacial lakeand how the meltwater is drained.

Figure 11. Plot of height versus width, comparing themountains Uranius and Zephyria Tholi and features in thenorthern plains interpreted to be table mountains [Allen,1979].


[37] The general structure of an ice sheet is that of animpermeable ice layer overlain by a permeable layer com-posed of snow and firn. The permeable layer is referred toas the passage zone [Smellie and Skilling, 1994]. On theEarth the passage zone is located within �150 m of the icesheet surface [Smellie, 2001]. There exist two main mech-anisms by which the englacial lake can drain, discussed indetail by Smellie and Skilling [1994] and Smellie [2001] andbriefly summarized here. The first mechanism involvesfloating of the surrounding ice sheet. If the hydrostaticpressure at the base of the englacial lake reaches themagnitude at the base of the surrounding ice, the ice sheetwill be lifted off its bed, and the meltwater will drainsubglacially, underneath the ice along the basement top-ography, and will carve out channels. This is presumablythe mechanism by which the majority of meltwater drainsfrom underneath the Vatnajokull ice cap in Iceland [Bjorns-son, 1974, 1975, 1992]. In order for this mechanism tooccur, the height of the englacial lake required to reach thecritical hydrostatic pressure must be less than the heightrequired to reach the passage zone. If the passage zone is

reached before the critical hydrostatic pressure is reached,the second mechanism of drainage results. Meltwater accu-mulates within the ice chamber, increasing the height of theenglacial lake. When the level of the passage zone isreached, the overflow of meltwater drains supraglacially,through the permeable firn and snow. This leads to a morestable lake level than if drainage were to occur subglacially.Some subglacial drainage may still occur occasionally iflava melts the ice at the base of the ice sheet [Bjornsson,1975; Bjornsson and Kristmannsdottir, 1984].[38] Supraglacial drainage through the passage zone

requires a relatively thick permeable layer and is thusthought to be the predominant mechanism of drainage inthick glaciers and ice sheets. Estimates of the thickness of theHesperian-aged ice sheet performed by Head and Pratt[2001] suggest that in places it was as much as 2500 mthick. Also, the lower gravity of Mars relative to that of Earthwould suggest that there would be less compaction of snowinto ice in a Martian ice sheet, and thus a Martian ice sheetwould be expected to have a thicker passage zone relative toa terrestrial one. Together, the estimates of the ice sheet

Figure 12. General sequence of events for a subglacial eruption, showing the development of the initialmelt cavity, englacial meltwater lake, and final morphology of the volcanic structure. Adapted fromSmellie and Skilling [1994].


thickness, as well as the lower gravity of Mars, suggest thatthe predominant mechanism of drainage of meltwater pro-duced from eruption into the Hesperian ice sheet would havebeen supraglacial drainage through the passage zone.[39] After the subglacial eruptions have ceased, melt-

water can still accumulate from the increase in geothermalheat near the erupted structure [Bjornsson and Gudmunds-son, 1993; Gudmundsson et al., 1995]. Thus jokulhlaupscan continue to form, leading to continued drainage of wateralong the basement topography and the carving out ofchannels. Once the geothermal heat level returns to pre-eruption conditions, the ice is able to reform around andupon the volcano. Such glacial activity could lead to erosionof the flanks of the edifice [Van Bemmelen and Rutten,1955]. As pointed out by Chapman and Tanaka [2001], thepostconstructional environment provides ample opportunityfor degradation of the ideal tuya form. The identification ofwell-preserved examples may be indicative of an ice sheetthat was melted and did not reform.[40] The heights of the mountains can be used as an

estimate of the thickness of the ice sheet at the time ofconstruction of the volcanoes. The conically shaped moun-tains would have erupted entirely within the englacial lake,while the flat-topped varieties would have just breached thesurface of the lake. Since the lake surface corresponds tothe level of the passage zone, which is below the surface of

the ice sheet, estimates made from the heights of thevolcanoes would be minimum estimates. These observa-tions provide a basis for comparison to the characteristics ofthe Martian mountains interpreted to be volcanoes eruptedin the proximity of the Hesperian-aged ice sheet [Head andPratt, 2001].

4.2. The Mountains

[41] Many of the morphological differences observedbetween the mountains and Martian subaerial volcanoesof similar size could be explained by the processes involvedin subglacial eruptions, as described above (Figure 12). Thehigh width-to-height ratios of the mountains (Figure 11),along with their relatively steep slopes, and in some casesflat tops, are very similar to those predicted by the sequenceof events outlined for a subglacial eruption (Figure 12).[42] Some of the mountains display a conical feature on

top of an otherwise flat-topped summit (mountains 6 and 11)(Figure 5). Such a conical feature could be accounted for bythe transition in the eruption from a subaqueous phase to asubaerial phase, or more simply from subaerial eruptions at atime when the ice sheet had retreated from the area [Allen,1979]. Most of the mountains do display a conical shape,although some are irregular in outline. How is this morphol-ogy accounted for among the others attributed to subglacialorigin? As documented byMathews [1947], there are severalconical volcanoes interspersed among table mountains innorthern British Columbia. These conical features likelyerupted at a time when the overlying ice was sufficientlythick that breach of the englacial lake was not possible by theerupting structure. In such a situation the ice walls wouldhave confined accumulation of the pillow material andhyaloclastites. A conical feature would result and is a likelyscenario for the origin of the more conical and irregularlyconical mountains under discussion here. The cone-shapedmountains all fall within the center of the regional depositsof the Dorsa Argentea Formation (Figure 7). Such locationsare consistent with the origin just outlined, as the formerlylarge Hesperian-aged ice sheet [Head and Pratt, 2001]would be expected to be have been thicker toward the poleand thinner farther away. The only exception is mountain 17,which, as discussed earlier, is a likely candidate for animpact crater central peak.[43] An interesting terrestrial analog for the spacing and

alignment of the mountains is a trio of volcanic edificeserupted subglacially in Iceland (Figure 13). This trio consistsof Kverkfjoll in the northeast, Grimsvotn in the middle, andThordarhyrna and the Laki craters in the southwest [Grond-vold and Johannesson, 1983; Bjornsson and Einarsson,1990]. The alignment spans �120–130 km, with spacingbetween the different centers of 30–70 km. A similar triocan be found within the alignment of the Martian mountains,specifically, mountains 5, 6, and 7, which span �123 km(Figure 2b). As indicated by the Icelandic features, eruptionsat one center do not require simultaneous eruption at one ofthe other centers. Additionally, the morphologies among theIcelandic structures in this trio are not constant, nor are thoseexhibited by mountains 5, 6, and 7. It should be noted thatwhile this Icelandic analog serves to show that differentmorphologies can be constructed underneath the same icesheet, the conditions at Iceland are certainly different fromthe Martian volcanoes. Specifically, there is no indication of

Figure 13. Annotated sketch map of the western portionof the Vatnajokull ice cap, Iceland, showing the spacing andalignment of the volcanic centers erupted beneath the ice.Inset shows outline of Iceland, with the location of its fourice caps, and a box around the enlarged view of westernVatnajokull.


a rift system, as is the case with Iceland, other than thealignment of the mountains.[44] The relatively large size of the summit calderas found

on some of the mountains may also have a relationship to asubglacial origin. During times between eruptions, melt-water may have collected within the summit craters in afashion similar to that occurring with the Grimsvotn calderain Iceland today [Bjornsson, 1983]. In Grimsvotn the melt-waer lake has served to both erode away at the banks of thesummit crater as well as fill in the crater with eroded material[Gudmundsson, 1989]. The summit craters found on at leastfour of the south polar mountains may share a similarevolutionary history to that of the Grimsvotn cladera.[45] The low-domed mountains could likely have erupted

entirely subaerially. Together they account for the fourshallowest slopes and lowest heights among the mountains(Table 1). Also, all four of these mountains are locatedoutside the boundary of the Dorsa Argentea Formation(Figure 7).[46] Finally, the interpretation of the mountains as sub-

glacial volcanoes is also supported by the surroundingfeatures discussed earlier. The channels found extendingaway from some of the mountains may be indicative ofsome basal meltwater drainage (Figure 2b). Such meltwatercould have accumulated during the constructional history ofthe mountains and been released as jokulhlaups by meltingof the surrounding ice walls along the basement. Asmentioned in section 4.1, the main mechanism of meltwaterdrainage is inferred to be supraglacial drainage through thepassage zone [Smellie, 2001].[47] The circular depressions located around several of the

mountains are remarkably similar to depressions observedaround some volcanoes in the north polar region of Mars[Grosfils and Sakimoto, 2001]. These volcanoes are slightlysmaller than the south polar mountains, but their correspond-ing circular depressions extend outward for about onediameter of the volcano and are at most �100 m deep,similar to the observations made of the circular depressionsaround the south polar mountains. Grosfils and Sakimoto[2001] interpreted the north polar depressions as resultingfrom volatile loss from the surrounding plains and groundsubsidence due to emplacement of the volcanoes, andinferred underlying sills. While there is no evidence forunderlying sills beneath the south polar mountains, theapproximate central location of the mountains within thecircular depressions suggests that the two are related and thatmelting and loss of volatiles could have resulted from heatconduction from the magma reservoir as well as sillemplacement [Head and Wilson, 2002]. Avolatile loss originfor the circular depressions is consistent with the volatile-rich nature of the Dorsa Argentea Formation and the formerpresence of an ice sheet [Head and Pratt, 2001].[48] Polygonal patterns are found primarily in the circular

depressions and marginal troughs surrounding many of themountains. If these date from the Hesperian period, thepolygons may have formed in the process of formation ofthe circular depressions, when volatiles were removed orwhen the ground subsided. Other possible origins for thecracking are from either mud desiccation or ice-wedgecracking [Seibert and Kargel, 2001] following pooling ofmeltwater in the depressions around the mountains or insome other low-lying areas acting as basins for meltwater

drainage. Alternatively, the polygonal terrain could be dueto more recent cycling of a volatile-rich upper layer [Laneand Christensen, 2000].[49] On the basis of the above detailed analysis of the

mountains and surrounding features found in associationwith them, we interpret several of the mountains as havingbeen constructed into the Hesperian-aged ice sheet. Thewell-preserved state of the mountains indicates that littlepostconstructional erosion has occurred. This suggests thatonce the ice was melted and removed from the vicinity ofthe mountains, the ice sheet did not reform [Chapman andTanaka, 2001]. Therefore the mountains could have likelyplayed a significant role in local meltback of the Hesperian-aged ice sheet. This is consistent with the work ofMilkovichet al. [2002], which suggests that meltback of the Hesperianice sheet likely resulted from bottom-up, rather than the top-down, melting.[50] The Dorsa Argentea Formation, the deposits of the

Hesperian ice sheet, dates to the middle part of the HesperianPeriod [Tanaka and Scott, 1987], suggesting that the largerice sheet had melted back and retreated by that time. There-fore the sources of the melting must have operated prior tothat time. Among the most prominent candidates for bottom-up melting in this region is the emplacement of the Hes-perian-aged ridged plains (Hr), which have been mapped byTanaka and Scott [1987] throughout much of the south polarregion. The ridged plains material forms numerous patchesof ‘‘broad, planar, moderately cratered surfaces marked bylong, linear to sinuous ridges.., embays all adjacent units..,’’and is interpreted as ‘‘extensive flows of low-viscosity lavaerupted at high rates from local sources..’’ [Tanaka andScott, 1987]. According to Tanaka and Scott [1987], Hr lieslargely stratigraphically below the Dorsa Argentea Forma-tion and was emplaced during the Early Hesperian. Thus, ifthe Dorsa Argentea Formation represents the remnants of apreviously larger volatile-rich ice sheet unit, then theemplacement of volcanic deposits of the ridged plains (Hr)during the early part of the Hesperian could have served as aheat source to melt portions of the ice sheet.[51] Indeed, such melting could have occurred in several

different modes. Recent mapping of the northern lowlandshas shown evidence for the widespread occurrence of Hr,lying below the Vastitas Borealis Formation (Late Hesper-ian) and resurfacing the northern lowlands [Head et al.,2002]. This new northern lowlands occurrence brings thetotal area of Hr to �30% of the surface of Mars, strengthen-ing arguments that this phase of volcanism and tectonismmay have represented a significant global event on theplanet [e.g., Watters, 1993]. Such important global eventscould have involved a period of enhanced global heat flow,potentially leading to enhanced basal melting of thickvolatile-rich deposits. Second, local intrusion of Hr intovolatile-rich deposits or ice sheets could obviously lead tothe equivalent of terrestrial supraglacial, intraglacial, andsubglacial melting, in which features such as tuyas mighthave formed. Thus the widespread nature and style ofoccurrence of the Early Hesperian ridged plains, togetherwith their geographic and temporal proximity, mean that theemplacement of Hr is a clear candidate for a source ofbottom-up melting of the unit that became the DorsaArgentea Formation. Furthermore, the individual occurren-ces of the mountain structures could well mark the mani-


festations of an Hr source vent erupted under an ice sheet,producing these unusual morphologic features, rather thanlaterally extensive volcanic plains.

4.3. Estimates of the Thickness of the Ice Sheet

[52] If these features are correctly interpreted to bevolcanic eruptions beneath and in the vicinity of an ancientice sheet, then the thickness of the ice sheet can beestimated from the heights and morphologies of the moun-tains. Specifically, the heights of the flat-topped mountainsand cone-shaped mountains can be used to develop con-straints on the minimum thickness of the ice sheet intowhich the volcanoes erupted. The flat-topped mountains areinterpreted to have erupted up to about the level of theenglacial lake. Different authors have proposed variousconnections between the level of the lake and the thicknessof the cap [e.g., Smellie and Skilling, 1994]. However, itcan be agreed that the maximum level of the lake is lessthan the thickness of the overlying ice. Therefore anestimate of the meltwater lake level can provide a minimumvalue for the thickness of the ice. This is especially true forMars, where the passage zone, corresponding to the lakelevel, would be expected to be farther from the surface ofthe ice sheet. The heights of the flat-topped south polarmountains provide an estimate of between 700 and 1300 m.[53] As discussed in section 4.2, the mountains that are

more cone-shaped could have erupted entirely within theenglacial lake. Discounting mountain 17, which may wellbe a central peak, the remaining cone-shaped mountainsprovide ice sheet thickness estimates ranging from 1200 to2200 m. Interestingly, the tallest of the cone-shaped moun-tains (13) is found in the middle of the regional deposit ofthe Dorsa Argentea Formation. The remaining cone-shapedmountains decrease in height closer toward the edge of thedeposit (Figure 7, Table 1).[54] The above estimates combine to give a minimum

estimate of the thickness of the overlying ice sheet, rangingfrom 700 m to 2.2 km and averaging to �1.4 km. Thickerestimates generally come from mountains closer toward themiddle of the regional deposit of Hd, and smaller estimatescome from mountains found closer toward the boundary ofthe deposit. This trend is consistent with the expectation thatthe ice sheet would have thinned toward its edges. Theestimates are also consistent with the thickness of theformerly large Hesperian-aged south polar ice sheet asestimated by Head and Pratt [2001].[55] An alternative explanation for the different ice sheet

thicknesses obtained from the heights of the mountains is avariation in ice sheet thickness with time rather than simplyin location. If the mountains were constructed at somewhatdifferent times, then the heights of the mountains may beindicating the thickness of the ice sheet at these times,implying that it was not constant. However, there does notappear to be any good evidence to suggest a significant agevariation among the mountains. The simplest explanationfor the varying heights seems to be a variation in ice sheetthickness with location.

4.4. Other Candidate Martian Subglacial Volcanoes

[56] To date, other features on Mars interpreted as sub-glacial eruptions have been low, elongated, steep-sided

mounds, considered to be analogous to terrestrial tablemountains [Hodges and Moore, 1978; Allen, 1979; Chap-man, 1994; Chapman and Tanaka, 2001]. To some degree,this has been due to the limited resolution of Viking images.However, planetwide surveys in search of volcano-iceinteraction have also been restricted to features resemblingIcelandic table mountains and moberg ridges. Such featuresare not the sole type of subglacial volcanoes found on Earthand therefore should not be considered the sole type thatmight be found on Mars.[57] The majority of the table mountains found on Mars

are located in the northern lowlands. The structures aregenerally smaller than the mountains found near the southpole and longer than they are wide, with widths rangingfrom 1 to 16 km in diameter and heights reaching up to 1km (Figure 11). An exception to this is a feature in HebesChasma which stands 4 km above the chasma floor [Chap-man and Tanaka, 2001]. They have steep slopes, and manyare flat-topped [Hodges and Moore, 1978; Allen, 1979;Chapman, 1994; Chapman and Tanaka, 2001]. What reasonis likely to explain the difference in morphology andmorphometry between the south polar features and the tablemountains of the northern lowlands?[58] A possible explanation might be found with the thick-

ness and type of ice layer into which the volcanoes erupted[Smellie, 2000]. Allen [1979] proposes that the table moun-tains he describes erupted into a subsurface ground ice layer.Chapman [1994] proposes that the table mountains found inUtopia erupted into a frozen paleolake, of thickness estimatedto be�180m. The south polar mountains would have eruptedinto a formerly large south polar ice sheet. The heights of theflat-topped and conical features have been used to arrive at aminimum value of 1.4 km for the average thickness of the icesheet. Clearly, a different set of constraining variables wouldhave been involved in the formation of the south polarmountains and the table mountains found elsewhere. Addi-tionally, the state of the base of the ice sheet (wet or dry)would have played a role in the eruption [Smellie and Skilling,1994]. Unfortunately, there is no method by which to deter-mine such conditions on Mars at this time.

5. Conclusions

[59] Several features previously mapped as mountains inthe south polar region of Mars [Tanaka and Scott, 1987]have been reexamined using MOLA data. From the spac-ing, alignment, and morphometries of the mountains, andfrom comparison to other Mars features, we interpret themountains as volcanoes. However, their morphologies showsome variation from other similarly sized Martian volca-noes. The mountains are located within the Dorsa ArgenteaFormation, a volatile-rich unit interpreted as the deposits ofa larger Hesperian-aged ice sheet [Head and Pratt, 2000].Their location within the Dorsa Argentea Formation, thecorrelation of their basal elevations with the elevation ofHd, and the unknown age of the mountains hint at apossible interaction between the volcanoes and the formerlylarge south polar ice sheet. Comparisons were madebetween the south polar mountains of Mars and terrestrialsubglacial volcanic constructs. Subglacial eruptions onEarth are known to produce structures of varying sizesand shapes, depending on the specific eruption conditions.


The morphologies of the mountains fit well with a plausiblesubglacial origin. The association of the mountainswith surrounding features supports this scenario. Channelsextending away from the volcanoes may represent melt-water drainage, and cracking found in shallow depressionssurrounding the structures may date from the time of themelting and could represent pooling of water followed bydesiccation, ice wedging, or volatile overturn.[60] If these interpretations are correct, thickness esti-

mates of the larger Hesperian-aged south polar ice sheetinterpreted to be represented by the Dorsa Argentea For-mation [Head and Pratt, 2001] can be made from the heightsof the flat-topped and cone-shaped mountains. Such esti-mates suggest a minimum average thickness of �1.4 km forthe ice sheet. Eruption into ice of this thickness is animportant difference from the eruption conditions inferredfor other Martian subglacial constructs [Allen, 1979; Chap-man, 1994]. The table mountains interpreted to occur onMars appear to have erupted into ice bodies significantlythinner than that of the formerly large south polar ice sheet.This difference appears to account for the distinct morphol-ogies observed between the south polar mountains and theMartian table mountains. The well-preserved state of themountains suggests that little postcontructional erosion hastaken place and could mean that the mountains are respon-sible for local meltback of the Hesperian ice. In addition, aspointed out by Head and Pratt [2001], if life originated inthe early history of Mars, then the Dorsa Argentea Formationand related units may provide an excellent location toexplore for fossil evidence [e.g., McKay and Stoker, 1989].Specifically, volcanic eruptions in this water-rich environ-ment, such as those described here, should be consideredcarefully as possible sites for further detailed analysis.

[61] Acknowledgments. We gratefully acknowledge fruitful discus-sions with John Smellie, Ian Skilling, Ken Tanaka, Mary Chapman, andSusan Sakimoto. Thanks are extended to the Rhode Island NASA SpaceGrant (P. H. Schultz, PI) for providing the internships during which thiswork was accomplished. We gratefully acknowledge NASA grants NAG5-9780 and NAG5-10690 to J.W.H., which helped fund the work andattendance at the Lunar and Planetary Science Conference to present theseresults. Comments from an anonymous reviewer significantly improved thiswork. We thank Steve Pratt for advice on data processing and volumecalculations. Peter Neivert and Anne Cote assisted with preparation of thefigures and manuscript. We also wish to thank the Brown MOLA ScienceGroup, with special thanks extended to Harry Hiesinger and Brad Thom-son. The authors acknowledge the use of Mars Orbiter Camera imagesprocessed by Malin Space Science Systems that are available at http://www.msss.com/moc_gallery/.

ReferencesAllen, C. C., Volcano-ice interactions on Mars, J. Geophys. Res., 84, 8048–8059, 1979.

Barlow, N. G., Crater size-frequency distribution and a revised Martianchronology, Icarus, 75, 285–305, 1988.

Barlow, N. G., Martian impact craters: Correlations of ejecta and interiormorphologies with diameter, latitude, and terrain, Icarus, 87, 156–179,1990.

Benn, D., and D. Evans, Glaciers and Glaciation, 734 pp., John Wiley,New York, 1998.

Bjornsson, H., Explanation of Jokulhlaups from Grimsvotn, Vatnajokull,Iceland, Jokull, 24, 1–16, 1974.

Bjornsson, H., Subglacial water reservoirs, jokulhlaups and volcanic erup-tions, Jokull, 25, 1–14, 1975.

Bjornsson, H., The cause of Jokulhlaups in the Skafta River, Vatnajokull,Jokull, 27, 71–78, 1977.

Bjornsson, H., A natural calorimeter at Grimsvotn: An indicator of geother-mal and volcanic activity, Jokull, 33, 13–18, 1983.

Bjornsson, H., Jokulhlaups in Iceland: Prediction, characteristics, and si-mulation, Ann. Glaciol., 16, 95–106, 1992.

Bjornsson, H., and P. Einarsson, Volcanoes beneath Vatnajokull, Iceland:Evidence from radio echo-sounding, earthquakes and jokulhlaups, Jokull,40, 147–168, 1990.

Bjornsson, H., and M. T. Gudmundsson, Variations in the thermal output ofthe subglacial Grimsvotn Caldera, Iceland, Geophys. Res. Lett., 20,2127–2130, 1993.

Bjornsson, H., and H. Kristmannsdottir, The Grimsvotn Geothermal Area,Vatnajokull, Iceland, Jokull, 34, 25–50, 1984.

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

Chapman, M. G., and K. L. Tanaka, Interior trough deposits on Mars:Subice volcanoes?, J. Geophys. Res., 106, 10,087–10,100, 2001.

Cintala, M. J., Martian fresh crater morphology and morphometry—A pre-Viking review, in Impact and Explosion Cratering, edited by J. Roddyand R. Pepin, pp. 575–591, Pergamon, New York, 1977.

Clifford, S. M., Polar basal melting on Mars, J. Geophys. Res., 92, 9135–9152, 1987.

Clifford, S. M., A model for the hydrologic and climatic behavior of wateron Mars, J. Geophys. Res., 98, 10,973–11,016, 1993.

Condit, C. D., L. A. Soderblom, Geologic map of the Mare Australe area ofMars, U.S. Geol. Surv. Misc. Invest. Ser. Map, I-1706-C, 1978.

Crumpler, L. S., J. W. Head, J. C. Aubele, Calderas on Mars: Character-istics, structure, and associated flank deformation, in Volcano Instabilityon the Earth and Other Planets, edited by W. J. McGuire et al., Geol.Soc. Spec. Publ., 110, 307–348, 1996.

Cutts, J. A., Nature and origin of the layered deposits of the Martian polarregions, J. Geophys. Res., 78, 4231–4249, 1973.

Cutts, J. A., K. R. Blasius, and W. J. Robert, Evolution of Martian polarlandscapes: Interplay of long term variations in perennial ice cover anddust storm intensity, J. Geophys. Res., 84, 2975–2994, 1979.

Einarsson, T., Physical aspects of sub-glacial eruptions, Jokull, 16, 167–174, 1966.

Fagents, S. A., P. Lanagan, R. Greeley, Pseudocraters on Mars: A conse-quence of lava-ground ice interaction, in Volcano-Ice Interaction onEarth and Mars, Geol. Soc. Spec. Publ., in press, 2002.

Frey, H., and M. Jarosewich, Subkilometer Martian volcanoes: Propertiesand possible terrestrial analogs, J. Geophys. Res., 87, 9867–9879, 1982.

Frey, H., B. L. Lowry, and S. A. Chase, Pseudocraters on Mars, J. Geophys.Res., 84, 8068–8075, 1979.

Ghatan, G. J., and J. W. Head, Candidate subglacial volcanoes in the southpolar region of Mars, Lunar Planet. Sci. [CD-ROM], XXXII, abstract1039, 2001.

Grant, J. A., and P. H. Schultz, Gradational epochs on Mars: Evidence fromwest-northwest Isidis Basin and Electris, Icarus, 84, 166–195, 1990.

Greeley, R., and S. A. Fagents, Icelandic pseudocraters as analogs to somevolcanic cones on Mars, J. Geophys. Res., 106, 20,527–20,546, 2001.

Gronvold, K., and H. Johannesson, Eruption in Grimsvotn 1983: Course ofevents and chemical studies of the tephra, Jokull, 34, 1–11, 1983.

Grosfils, E. B., and S. E. Sakimoto, Topographic constraints on magmareservoir volume and depth for small near-polar volcanoes in the northernplains of Mars, Lunar Planet. Sci. [CD-ROM], XXXII, abstract 1111, 2001.

Gudmundsson, M. T., The Grimsvotn caldera, Vatnajokull: Subglacial to-pography and structure of caldera infill, Jokull, 39, 1–19, 1989.

Gudmundsson, M. T., H. Bjornsson, and F. Palsson, Changes in jokulhlaupsizes in Grimsvotn, Vatnajokull, Iceland, 1934–91, deduced from in-situmeasurements of subglacial lake volume, J. Glaciol., 41, 263–272, 1995.

Gudmundsson, M. T., F. Sigmundsson, and H. Bjornsson, Ice-volcano in-teraction of the 1996 Gjalp subglacial eruption, Vatnajokull, Iceland,Nature, 389, 954–957, 1997.

Hale, W., and J. W. Head, Central peaks in Martian craters: Occurrence bysubstrate, ejecta type and rim diameter, Lunar Planet. Sci., XIII, 295–296, 1981.

Head, J. W., and G. J. Ghatan, Hesperian-aged ice-rich deposits near thesouth pole: Evidence for large-scale melting and drainage in the southernNoachis Terra region of Mars, Lunar Planet. Sci. [CD-ROM], XXXII,abstract 1062, 2001.

Head, J. W., and B. Hallet, Origin of sinuous ridges in the Dorsa ArgenteaFormation: Additional criteria for tests of the esker hypothesis, LunarPlanet. Sci. [CD-ROM], XXXII, abstract 1366, 2001a.

Head, J. W., and B. Hallet, Origin of sinuous ridges in the Dorsa ArgenteaFormation: New observations and tests of the esker hypothesis, LunarPlanet. Sci. [CD-ROM], XXXII, abstract 1373, 2001b.

Head, J. W., and S. Pratt, Extensive Hesperian-aged south polar ice sheet onMars: Evidence for massive melting and retreat, and lateral flow andponding of meltwater, J. Geophys. Res., 106, 12,275–12,299, 2001.

Head, J. W., and L. Wilson, General environments and geological settingsof magma/H2O interactions, in Volcano-Ice Interaction on Earth andMars, Geol. Soc. Spec. Publ., in press, 2002.


Head, J. W., et al., Characterization of major volcanic edifices on Marsusing Mars Orbiter Laser Altimeter data, Lunar Planet. Sci. [CD-ROM],XXIX, abstract 1322, 1998.

Head, J. W., S. Milkovich, and S. Pratt, Meltback of Hesperian-aged ice-rich deposits at the south pole: Evidence for proglacial drainage channelsand lakes, Lunar Planet. Sci. [CD-ROM], XXXII, abstract 1156, 2001.

Head, J. W., M. A. Kreslavsky, and S. Pratt, Northern lowlands of Mars:Evidence for widespread volcanic flooding and tectonic deformation inthe Hesperian Period, J. Geophys. Res., 107(E1), 10.1029/2000JE001445,2002.

Herkenhoff, K. E., and B. C. Murray, Color and albedo of the south polarlayered deposits on Mars, J. Geophys. Res., 95, 1343–1358, 1990.

Hickson, C. J., Physical controls and resulting morphological forms ofQuaternary ice-contact volcanoes in Western Canada, Geomorphology,32, 239–261, 2000.

Hoare, J. M., and W. L. Coonrad, A tuya in Togiak Valley, SouthwestAlaska, J. Res. U.S. Geol. Surv., 6, 193–201, 1978.

Hodges, C. A., and H. J. Moore, Tablemountains of Mars, Lunar Planet.Sci., IX, 523–525, 1978.

Hodges, C. A., and H. J. Moore, Atlas of Volcanic Landforms on Mars,U.S. Geol. Surv. Prof. Pap., 1534, 1994.

Hoffman, N., J. Kargel, and K. Tanaka, Basal melting of a CO2-rich icecapon Mars, Lunar Planet. Sci. [CD-ROM], XXXIII, abstract 1509, 2002.

Jonsson, J., Notes of the Katla volcanoglacial debris flows, Jokull, 32, 61–68, 1982.

Kargel, J. S., Possible composition of Martian polar caps and controls onice-cap behavior, paper presented at First International Conference onMars Polar Science, Lunar and Planet. Inst., Houston, Tex., 1998.

Keszthelyi, L. P., A. S. McEwen, and T. Thordarson, Terrestrial analogs andthermal models for Martian flood lavas, J. Geophys. Res., 105, 15,027–15,049, 2000.

Kieffer, H. H., H2O grain size and the amount of dust in Mars’s residualnorth polar cap, J. Geophys. Res., 95, 1481–1493, 1990.

Kolb, E. J., and K. L. Tanaka, South polar region of Mars: Preliminary1:3,000,000-scale mapping results, Lunar Planet. Sci. [CD-ROM],XXXII, abstract 2082, 2001.

Kortz, B. E., and J.W. Head, Comparisons of volcanic fields on Venus, Earth,and Mars, Lunar Planet. Sci. [CD-ROM], XXXII, abstract 1422, 2001.

Lanagan, P. D., A. S. McEwen, L. P. Keszthelyi, and T. Thordarson, Root-less cones on Mars indicating the presence of shallow equatorial groundice in recent times, Geophys. Res. Lett., 28, 2365–2367, 2001.

Lane, M. D., and P. R. Christensen, Convection in a catastrophic flooddeposit as the mechanism for the giant polygons on Mars, J. Geophys.Res., 105, 17,617–17,627, 2000.

Major, J. J., and C. G. Newhall, Snow and ice perturbation during historicalvolcanic eruptions and the formation of lahars and floods, Bull. Volcanol.,52, 1–27, 1989.

Malin, M. C., K. S. Edgett, S. D. Davis, M. A. Caplinger, E. Jensen, K. D.Supulver, J. Sandoval, L. Posiolova, and R. Zimdar, M03-00362, M03-07301, M04-00561, M04-00913, M07-03961, M07-05379, M11-02780,Malin Space Science Systems Mars Orbiter Camera Image Gallery(http://www.msss.com/moc_gallery/), 1999, 2000.

Mathews, W. H., ‘‘Tuyas,’’ flat-topped volcanoes in northern British Co-lumbia, Am. J. Sci., 245, 560–570, 1947.

Mathews, W. H., The Table, a flat-topped volcano in southern BritishColumbia, Am. J. Sci., 249, 830–841, 1951.

McKay, C. P., and C. R. Stoker, The early environment and its evolution onMars: Implications for life, Rev. Geophys., 27, 189–214, 1989.

Mellon, M. T., Limits of the CO2 content of the Martian polar deposits,Icarus, 124, 268–279, 1996.

Melosh, H. J., Impact Cratering: A Geological Process, Oxford Univ.Press, New York, 1989.

Milkovich, S. M., J. W. Head, and S. Pratt, Meltback of Hesperian-aged ice-rich deposits near the south pole of Mars: Evidence for drainage channelsand lakes, J. Geophys. Res., 107, 10.1029/2001JE001802, in press, 2002.

Mouginis-Mark, P. J., Martian fluidized crater morphology: Variations withcrater size, latitude, altitude, and target material, J. Geophys. Res., 84,8011–8022, 1979.

Murray, B. C., L. A. Soderblom, J. A. Cutts, R. P. Sharp, D. J. Milton, andR. B. Leighton, Geological framework of the south polar region of Mars,Icarus, 17, 328–345, 1972.

Nielson, N., A volcano under an ice-cap, Vatnajokull, Iceland, 1934–1936,Geogr. J., 90, 6–23, 1937.

Nye, J. F., Water flow in glaciers: Jokulhlaups, tunnels, and veins, J. Gla-ciol., 17, 181–207, 1976.

Peterson, J. E., Geologic map of the Noachis quadrangle of Mars, U.S.Geol. Surv. Misc. Invest. Ser. Map, I-910, 1977.

Pike, R. J., Control of crater morphology by gravity and target type: Mars,Earth, Moon, Proc. Lunar Planet. Sci. 11th, 2159–2189, 1980.

Pike, R. J., Some morphologic systematics of complex impact craters,Meteoritics, 20, 49–68, 1985.

Scott, D. H., and M. H. Carr, Geologic map of Mars, U.S. Geol. Surv. Misc.Invest. Ser. Map, I-1083, 1978.

Scott, D. H., and K. L. Tanaka, Geologic map of the western equatorialregion of Mars, U.S. Geol. Surv. Misc. Invest. Ser. Map, I-1082A, 1986.

Seibert, N. M., and J. S. Kargel, Small-scale Martian polygonal terrain:Implications for liquid surface water, Geophys. Res. Lett., 28, 899–902, 2001.

Skilling, I. P., Evolution of an englacial volcano: Brown Bluff, Antarctica,Bull. Volcanol., 56, 573–591, 1994.

Smellie, J. L., Subglacial eruptions, in Encyclopedia of Volcanoes, editedby H. Sigurdsson, pp. 403–418, Academic, San Diego, Calif., 2000.

Smellie, J. L., Lithofacies architecture and construction of volcanoeserupted in englacial lakes: Icefall Nunatak, Mount Murphy, eastern MarieByrd Land, Antarctica, Spec. Publ. Int. Assoc. Sedimentol., 30, 9–34,2001.

Smellie, J. L., and I. P. Skilling, Products of subglacial volcanic eruptionsunder different ice thickness: Two examples from Antarctica, Sediment.Geol., 91, 115–129, 1994.

Smith, D. E., et al., Topography of the northern hemisphere of Mars fromthe Mars Orbiter Laser Altimeter, Science, 279, 1686–1692, 1998.

Stewart, E. M., and J. W. Head, Ancient Martian volcanoes in the AeolisRegion: New evidence from MOLA data, J. Geophys. Res., 106,17,505–17,513, 2001.

Strom, R. G., S. K. Croft, and N. G. Barlow, The Martian impact crateringrecord, in Mars, edited by H. H. Kieffer et al., pp. 383–423, Univ. ofAriz. Press, Tucson, 1992.

Tanaka, K. L., and D. H. Scott, Geologic map of the polar regions of Mars,U.S. Geol. Surv. Misc. Invest. Ser. Map, I-1802C, 1987.

Thomas, P., S. Squyres, K. Herkenhoff, A. Howard, and B. Murray, inMars, edited by H. H. Kieffer et al., pp. 298–320, Univ. of Ariz. Press,Tucson, 1992.

Thorarinsson, S., The crater groups in Iceland, Bull. Volcanol., 14, 3–44,1953.

Thorarinsson, S., The jokulhlaup from the Katla area in 1955 comparedwith other jokulhlaups in Iceland, Jokull, 7, 21–25, 1957.

Tomasson, H., Grimsvotnalaup 1972: Mechanism and sediment discharge,Jokull, 24, 27–39, 1974.

Tomasson, H., The jokulhlaup from Katla in 1918, Ann. Glaciol., 22, 249–254, 1996.

Van Bemmelen, R. W., M. G. Rutten, Tablemountains of Northern Iceland,217 pp., E. J. Brill, Cologne, Germany, 1955.

Watters, T. R., Compressional tectonism on Mars, J. Geophys. Res., 98,17,049–17,060, 1993.

Wilson, L., and J. W. Head, Mars: Review and analysis of volcanic eruptiontheory and relationships to observed landforms, Rev. Geophys., 32, 221–263, 1994.

Wilson, L., and J. W. Head, Evolution of magma reservoirs within shieldvolcanoes on Mars, Lunar Planet. Sci. [CD-ROM], XXIX, abstract 1128,1998.

Wilson, L., and J. W. Head, Tharsis-radial graben systems as the surfacemanifestation of plume-related dike intrusion complexes: Models andimplications, Lunar Planet. Sci. [CD-ROM], XXXI, abstract 1371,2000.

Wilson, L., and J. W. Head, Giant dike swarms and related graben systemsin the Tharsis province of Mars, Lunar Planet. Sci. [CD-ROM], XXXII,abstract 1153, 2001.

Wilson, L., and J. W. Head, Heat transfer and melting in subglacial volcaniceruptions: Implications for volcanic deposits and meltwater volumes,Lunar Planet. Sci. [CD-ROM], XXXII, abstract 1213, 2001.

Wood, C. A., J. W. Head, and M. J. Cintala, Interior morphology of freshMartian craters: The effects of target characteristics, Proc. Lunar Planet.Sci. Conf. 9th, 3691–3709, 1978.

��G. J. Ghatan and J. W. Head III, Department of Geological Sciences,

Brown University, Box 1846, Providence, RI 02192, USA. ([email protected])


Figure 1. (a) Portion of the geologic map of the south polar region generalized from Tanaka and Scott[1987] and showing the distribution of the features mapped as mountains (m), shown in green. Otherunits include Hd (the Dorsa Argentea Formation), HNu (undivided material), Npl (rough, heavily crateredterrain), V (volcanoes), Hpl (smooth, moderately cratered terrain), and Am (mantle material found withinCavi). The map covers from �20�W to 340�W and 55�S to 75�S. The mountains are locatedpredominantly within Hd or in close association with it. (b) MOLA color topography overlain on MOLAshaded relief map of the region shown in Figure 1a. Only those areas above 800 m elevation have colortopography overlain, with red indicating high elevations and blue indicating low.

GHATAN AND HEAD: MARTIAN SOUTH POLAR REGION SUBGLACIAL VOLCANOES

2 - 2

Figure 2. (a) Perspective view of aligned mountains looking toward the current south polar cap. MOLAcolor topography is overlain on MOLA shaded relief map. Red is high, and purple is low. Note thecircular depressions around many of the mountains, which stand significantly lower then the mountains.Also note the various morphologies of the mountains. Vertical exaggeration is 70�. (b) Enlarged view ofportion of area containing the mountains. MOLA color topography is overlain on MOLA shaded reliefmap; red is high, and purple is low. Note the channel (indicated by arrows) emanating away from the areanear mountains 5, 6, and 7. Also note the circular depressions and marginal troughs surrounding thedifferent mountains.

GHATAN AND HEAD: MARTIAN SOUTH POLAR REGION SUBGLACIAL VOLCANOES

2 - 4

Documents

Candidate subglacial volcanoes in the south polar region ... · Morphology, morphometry, and eruption conditions ... and large summit craters relative to the summit ... trending slightly