sign in sign up
<<
Home , Noachian, Tholus

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 106, NO. E8, PAGES 17,505-17,513,AUGUST 25, 2001

Ancient volcanoes in the Aeolis region' New evidence from MOLA data

Emily M. Stewart and JamesW. Head Departmentof Geological Sciences,Brown University, Providence,Rhode Island

Abstract. Orbiter Laser Altimeter (MOLA) altimetricdata confirm the volcanicorigin of a -agedhilly feature(Zephyria ) in the Aeolis areaof Mars. Similar characteristics sharedby at least one nearbytopographic high suggestthat volcanicactivity representedby such featuresmay have beenmore widespreadthan previously thought. Morphologicalstructures and morphometricdata for Zephyriaare consistentwith a stratovolcanoorigin, in which the edifice formedby mixed explosiveand effusiveeruptions, and was alteredby subsequentflank erosion,to producea distinctivetruncated cone with concaveupward flanks. Criteriaare outlined to help detectand distinguishthe origin of suchfeatures in otherancient regions of Mars.

1. Introduction As on Earth, the topographic and slope characteristics of Feature A could provide information about its eruptive history. An unusually symmetricalcone located in the Aeolis region In this paper we use Mars Orbiter Laser Altimeter (MOLA) to- of Mars (20øS, 187øW) with a flat-floored summit crater and dis- pographicdata [•t al., 1998, 1999; Zuber et al., 2000] to tinctive radial dissection of its flanks was first recognized by study the shapesand slope characteristicsof this feature and to and Spudis [1978] and interpreted to have a volcanic assessits origin. Furthermore, we examine the characteristics origin (Feature A, Plate 1). In their global survey of Martian of other, more degradedtopographic highs nearby to assessthe volcanoes,Hodges and Moore [1994] describedthe featureas a likelihood that they represent the same class of feature, and single, symmetric, crateredcone in the hilly Noachian high- thus have the same origin, as Feature A. lands with a basal diameter of-30 km and a circular summit cra- ter 8 km across. On the basis of its shape and its striking mor- 2. Stratovolcanoes on Earth phological similarity to Earth stratovolcanoes (e.g. Tongariro in New Zealand;Figure 4lB of Hodgesand Moore [1994]), they On Earth, compositevolcanoes or stratovolcanoesare rela- interpreted this feature to be the best candidatein their global tively large constructionalvolcanic featuresbuilt of layers of survey for a composite cone or stratovolcano. Tongariro vol- small-volume flows and pyroclastic deposits [Bates and cano is a truncated composite cone composedof andesitic ash Jackson, 1984], unlike shield volcanoes, which are built pre- and lava flows which has subsequentlyundergone flank erosion dominantly of lava flows with a widerrange of lengths. The by glaciers and streams[Gregg, 1960]. greaterexplosiveness of stratovolcanoeruptions relative to FeatureA (Zephyria Tholus) is formedwithin the hilly unit shield volcano effusions is usually attributed to higher viscos- of the lower Noachian highlandplateau sequence (Nplh), one of ity and/orvolatile contentof stratovolcanomagmas [Cas and the oldest recognized stratigraphic units on Mars [Scott and , 1987, pp. 386-390]. The pyroclastic deposits and Tanaka, 1986]. Nplh consists of rough, hilly, and fractured lava flows associated with stratovolcanoes tend to accumulate material with locally high relief and is inferred to be ancient in the near-vent area, contributing to smaller basal diameter highlandrocks and impact brecciaproduced during the period of (30-100 km) and steeper flank slopes (as high as 35ø) than heavy bombardment. In the Aeolis region, occurrencesof Nplh thosetypical of shieldvolcanoes [Walker, 2000; Davidsonand are surroundedby highly cratered, fractured, and channelized de Silva, 2000]. Terrestrial arc stratovolcanoesare typically 2- material mapped as Npl•, which is thought to consist of a mix- 2.5 km in height, while intraplate structurescan be as tall as ture of volcanic materials,erosional products, and impact brec- 3.7 km [Walker, 2000]. In profile, young stratovolcanoes cias of middle Noachian [Scott and Tanaka, 1986; Greeley havea simpleconical shape. Older stratovolcanoescommonly and Guest,1987]. FeatureA is thusinterpreted to be a an- haveconcave upward flank profilesas a result of fluvial dissec- cient structure, remarkably pristine for its age (Plate 1). tion or sector collapse [e.g., Reid et al., 2000]; truncation of Zephyria Tholus is located at the head of Durius Valles, which the upperslopes by calderaformation (by explosion or col- consists of several flat-floored tributaries incised into the lapse) also occurs. Stratovolcanoesare typically found in crateredhighlands of . These tributaries con- chains on the overriding plate of convergent plate margins, verge into a single channel before emptying into De with intervolcanospacings comparable to their diameters(30- Vaucolueurs, a-300 km diameter degraded impact basin 100 km). Exceptionsinclude some edifices found on Iceland [Nelsonet al., 2001a]o Zephyria Tholus may be relatedto the and some associated with the East African Rift. sourceand origin of Durius Valles [Nelson et al., 200lb].

Copyright2001 by the AmericanGeophysical Union. 3. and Stratovolcanoes on Mars

Papernumber 2000JE001322. Volcanic edifices on Mars include the giant central vent vol- 0148-0227/01/2000JE001322509.00 canoes (e.g. Olympus and Elysium Montes), paterae, tholi,

17,505 17,506 STEWART AND HEAD: ANCIENT MARTIAN VOLCANOES

3000 ,, 10949.1.]•. b. 2000- ? looo.j Verticalexaggeration 10x 0 I • ! ß 3000 11515.1 12433.1 2 000.

1000-

...... 0 ! -'- I I I 3000 12942.2 2000-

1000-

0 '1 ...... I I 3000 .•2000- 11691.1 • 1000-

0 I I ! ! 3000 ...... -•' 11848.2 10452.2 • 2000- o • 1000- 0 • • "/..... 186.75 .r '•::?•:': :71::-.:. ", ..... 184 ...... •::•.•...... •.•4•::•...... '":<.."::::... • ' ...... : .:-..•-m..•:.':••"•.•.•:•::•:•...... •. '".%:•...: •...... •...• •..:•......

• ß"• ...... '"• -":....' ...... L'4..':...'...:;2 •: ...... '...-' :::•..-.-'..--...... ß...... '•4•:•:• .....

o ---- ...... :.. o •::••"::'• ...... •::g&:.•...? •-"•'•:•/•-•g• F "' T" :.:'-::.'.'.:-"..•...... :.:."."i'•......

•;•:•::....:;::::..::?:.?•:•:•:•:.::.::..:::..:.:'...... :•:•:..•:•

185 -.- - 19.5 •fitude (de• -20.5 -17.25 Latitude (deg) - 18.25

Figure 1. MOLA altimetry data for several candidatefeatures: (a)MOLA profiles acrossFeature A and (b) MOLA profiles acrossFeature B. Here and in all otherfigures, zero referenceis MOLA-derived mean radius(see Smith et al. [1998, 1999] and Zuber et al. [2000] for details). Points are individual MOLA data points, and numbersrefer to profile number. North is to the left in both images.

cones, and small shields [Hodges and Moore, 1994; (e.g., see summaryby Wilson and Head [1994]). There is also Zimbelman, 2000; Greeley et al., 2000]. On the basis of flow extensive evidence for explosive or pyroclastic activity [e.g., morphology and other considerations,most of these features Mouginis-Mark et al., 1982; Greeley and Crown, 1990; Crown are interpretedto be the result of effusive basaltic volcanism and Greeley, 1993; Head and Wilson, 1998; Plescia, 2000]. STEWART AND HEAD: ANCIENT MARTIAN VOLCANOES 17,507

F ature A - Feature B

ß

ß t

.%

3O km ",,•.• -1000 0 rn 1000 2000 3000 4000 "•" '

Plate1. RegionalViking image mosaic overlain on MarsOrbiter Laser Altimeter (MOLA) topography, with superposedimages of thefeatures mapped as Nplh in theAeolis region of Mars.Feature A is ZephyriaTholus. Thisis Vikingimage F430S23 of thecandidate volcano located at 20øS,187øW. Image width is 60 km. Note the flat-flooredsummit crater and the lobatefan depositin the impactcrater to the northeast. FeatureB is a Vikingimage of materialmapped as Nplh located at 18øS,184øW. Note the several domical features with to- pographysimilar to FeaturesA andB locatedto theirsouth (annotated as C1-C5). 17,508 STEWART AND HEAD: ANCIENT MARTIAN VOLCANOES

Although Francis and Wood [1982] arguedthat few Martian edi- 4. Topographic Characteristics of Candidate fices could have a pyroclastic origin on the basis of their low Volcanic Features in the Aeolis Region topographicslopes, Wilson and Head [1994] demonstratedthat becauseof the difference in atmospheric pressure and gravity The recent acquisition of accuratetopography information on Mars relative to Earth, pyroclastic deposits should be very for Mars using the Mars Orbiter Laser Altimeter (MOLA) [Smith common on Mars. In addition, volcanic deposits of all kinds et al., 1998, 1999; Zuber et al., 2000] permits the accurate are expectedto travel farther from the source vent than they measurementof the slope and shape of the Aeolis features would on Earth. This would, in turn, result in volcanic con- (Plate 1, Figures 1-3). Here we describe their characteristics, structs on Mars a factor of 2 broader and a factor of 4 shorter similarities, and differences and assess evidence relevant to a than Earth constructs, calling into question the slope argu- possible volcanic origin. ments of Francis and Wood [1982]. Wilson and Head[1994] demonstrated that volatile nuclea- 4.1. Feature A (Zephyria Tholus) tion and the resultant disruption of can occur in liquids MOLA data show that FeatureA (Zephyria Tholus; Plate 1, of basaltic composition on Mars because of the low atmos- FeatureA)is a truncatedcone, characterizedby asymmetrical pheric pressure. They noted that elevation would have a strong shapeand a summit depression(Figures l a and 2a). The base effect on the presenceof pyroclastic activity, with greater dis- of the feature averages-30 km wide and lies at about 500-1000 ruption of magma occurring at higher elevations (when the at- m elevation, and the presenthighest point of the rim is -3000 mosphericpressure is lower). On the basis of these considera- m, yielding a height of about 2-2.5 km (Figure la). Flank tions, Head and Wilson [1998] inferred that plinian-style (ex- slopesapproach 14 ø near the summitand decreaselinearly with plosive) eruptions should have been common throughout decreasingelevation (Figures 2a, 3a, and 3b), producing a Martian volcanic history. Increased magmatic volatile con- concave-upwardflank shape that is similar to Earth stratovol- tent, more silicic , and addition of nonjuvenile vola- canoes that have been fluvially dissected [Davidson amt de tiles (e.g., ) would all increase the likelihood of Silva, 2000]. The north flank of the structureappears to have explosive pyroclastic activity. rougher topography (Plate 1, Figures l a, 2a, 3a, 3b, and 4)be- In summary,pyroclastic activity is predictedto be an impor- causethe MOLA track crossesa deeply incised valley. Evi- tant processfor Mars, even for magmas of basaltic composi- dence for dissection by channels is well illustrated on this tion. Therefore stratovolcano-like features could be common. flank in the perspective view (Figures 3a and 3b); two sinuous However, they are likely to have more subduedtopography channels,each over a kilometer in width, begin near the sum- (lower flank slopes) than Earth stratovolcanoesbecause of the mit and continue to the base, winding down the slope of the enhanceddispersal of pyroclasticair fall depositsand enhanced featureand producinga prominent planeze between them. A run-out distances typical of Martian pyroclastic flows [e.g., planeze is a triangular fiat-faced facet on the flanks of volca- Wilson and Head, 1994; Head and Wilson, 1998]. noes formed by the intersection of two master gullies in the upper reachesof a cone. At the northeastern base of the structure a lobate flow ex- 3 ' ' ' 30 tends down a crater rim and out onto the crater floor (Plate 1, Figure 3a and 3b). Although this feature appearsto be oriented •2 20 • radially to FeatureA and could be a volcanic flow emanating from it, the morphologicalfreshness of the lobe and its prox- imity to a younger suggestalternatively that it 10 • may have been initiated by lobate emplacement or seismic-energy-inducedfailure of the crater wall as a result of the . i ; I ; I : I ', I ', '1 "', o 3O The other flanks of the edifice alsoappear to be dissectedby channelsinto planezesand other large angularflank segments. This type of topography could form as a result of differential erosion of flank deposits [e.g., Davidson and de Silva, 2000] or, alternatively, as a result of sector collapse of part of the volcanoflank when flank depositsbecame oversteepened [e.g., • 1- van Wyk de Vries et al., 2000; Reid et al., 2000]. For the Aeolis featuresexamined in this paper, high-resolution Mars 0 ' ! ' I ' I ' ! ' I ' Orbiter Camera (MOC) data are available only for the lower part 0 10 20 30 40 50 60 of the westernflank of FeatureA (Figure 4d). These images re- N-S distance(km) veal that the edifice is heavily crateredand that the terrain of the lower flanks is characterizedby local developmentof linear Figure 2. MOLA summit profiles. (a) MOLA profile across and sinuous radial ridges composedof coherent material sur- Feature A (orbit 12263.2) and point-to-point slopes with 3 km roundedby heavily mass-wasteddebris deposits. The heavily (dashed)and 5 km (dotted) baselines. Vertical exaggeration is degradedflank texture couldbe relatedto radial volcanic flows 10x. The north flank has slopes as high as 14ø near the sum- and channels (Figure 4f) and is also reminiscent of channel mit, decreasingapproximately linearly with elevation, produc- structureson the flanks of [Mouginis-Mark et ing the concave-upwardshape typical of a fluvially dissected stratovolcano[Davidson amt de Silva, 2000]. (b) MOLA pass al., 1982]. across Feature B (orbit 12433.1). This also has a concave- The truncated summit of the feature consists of a circular de- upwardsouthern slope and a relatively flat summit, approxi- pression 8 km wide with a crater rim 200 m higher than the mating a truncatedcone in profile. floor (Figures l a, 2a, 3a, 4a, and 4b). The floor of the summit STEWART AND HEAD: ANCIENT MARTIAN VOLCANOES 17,509

E w

b•

....

Figure 3. (a andb) Perspectiveviews of FeatureA, the highly dissected,Noachian-aged candidate Martian stratovolcanofirst describedby Greeleyand Spudis[1978]; view is generatedby overlaying high-resolution Viking imageF430S23 onto a digitalelevation model (DEM) usingMOLA topography.Vertical exaggeration is -3x. The craterat the centerof the coneis slightlytilted to the northwest.Note the fan depositin the crater in the northeast. (c and d) Perspectiveview of FeatureB, the topographichigh to the northeastof the volcano in Figures3a and3b. High-resolutionimage data are not availablefor thisfeature. It is also heavily dissected by valleysand has the samesize and slope characteristics as the otherfeature. Although a summitcrater is not obviousfrom the Viking images,MOLA altimetry revealsa subduedring-shaped depression in the summitof this feature (Figures lb and 2b). crater is relatively smooth (slopes of <1ø; Figure 2a) and tilts would fall in the Olympus-typecaldera category of Crumpleret slightly to the northwest. The circularity of the crater, its dis- al. [1996], which consistsof simple, generally circular or arcu- tinctly centered summit location, its lack of impact-relatedfea- ate summitdepressions that are often nested. FeatureA is most tures, and the smoothnessof its interior all are broadly similar similar to the subtype, which consistsof a sin- to the characteristics of volcanic calderas on Mars [Wood, gle very circularcrater in plan form, floored by smooth plains. 1984; Crumpler et al., 1996]. The FeatureA summit structure The size of the Feature A summit crater (-•8 km wide and several 17,510 STEWARTAND HEAD: ANCIENT MARTIAN VOLCANOES

o

?,g..-

..

,-.

-.

Figure4. Morphologyandtopography ofFeature A(Zephyria Tholus). (a)Viking image (F430S23) (70 m/pixel).(b) Viking image with different illumination conditions (portion ofMDIM MI20S187; -230 m/pixel).(c)Topographic mapderived from MOLA DEM (16 pixels/degree). (d)Viking image with rectangle showinglocation ofMOC image MO3-07208. Portions ofMars Orbiter Camera (MOC) image MO3-07208 (-4 mresolution). Width of image is2.83 km. (e) Top: smooth plains (upper left) adjacent tothe base of the edi- fice(lower part), which issegmented andheavily cratered. Narrow ridges arearrayed generally radially tothe structureandconsist ofcoherent blocky linear scarps surrounded bymass-wasting fragmental material. (f) Middle:ridges and intervening fragmental material aremore equidimensional, perhapsrelated tothe large numberofsubdued craters. The linear scarp atthe bottom isthe northern margin ofthe large linear-textured flankdeposit seen inthe lower image. (g) Bottom' portion ofone of the major valleys that extend radially fromthe summit area down the flanks; narrow linear to sinuous ridges are arrayed generally parallel to the strikeof the valley and are surrounded bymass-wasted finerdebris. This portion ofthe deposit is heavily cratered,witha narrow channel (about 200-300 mwide) visible inthe upper middle part of the image. At the baseofthe image, adjacent surrounding smoothplains are observed. (h)Annotated MOLA profile along MOC imageshowing local topography over the flank of thefeature. hundredmeters deep) is at thelower end of therange of caldera thatthe heightand summit crater width/depth ratios are in the diametersand depths for Mars [Crumpier et al., 1996,Figures rangeof terrestrial values and that the edifice diameter is toward 23 and24], consistent with the small diameter of theedifice. thehigher end of theterrestrial values. However, pyroclastic Comparisonof the morphometry of Feature A and data for air falls and flows shouldbe more widely dispersedon Mars 216 terrestrialstratovolcanoes [Pike, 1978, Table2] shows [Wilsonand Head, 1994], and thus edifices built with pyroclas- STEWART AND HEAD: ANCIENT MARTIAN VOLCANOES 17,511

1500 -20.2 Viking Latitude (deg) ß

looo- ß ß oil

ß

500 - ß

vertical exaggeration5x 0

Figure 4. (continued)

tic eruption contributionsare predictedto be broader than simi- summit crater and younger dome, the Viking resolution is not lar structureson Earth, potentially accountingfor the diameter sufficient to assessthis possibility, and no MOC images of of Feature A relative to terrestrial edifices. this featureare currentlyavailable. The flanks of the featureare Among the severalplausible origins for high topography in characterizedby numeroustriangular-shaped blocks with one ancient terrain are (1) multiple overlapping impact crater rims tip of the triangle pointing up slope. Although the detailed consisting of topography due to structuraluplift and eraplace- morphology and relation to possible flank channelsare not ment of ejecta preferentially on the rims; (2) tectonic scarps completelyclear owing to relatively low image resolution, the causedby faulting and production of differential topography; similarity to featureson the flanks of FeatureA (Plate 1; Fig- (3) volcanic construction,caused by edifice-building processes ures 3 and 4) suggeststhe possibility that these featuresare (extrusiveand explosive eruption productsaround a vent), and planezesor sector collapseremnants. (4) preferential erosion, including exhumation of preexisting Viking images(Plate 1, FeatureB) show that the northeast- topographicfeatures, or simple productionof inselbergs dueto ern flank of this featureappears broadly linear, and the feature the regional stripping of units. itself, in contrastto Feature A, is elongated, trending in a gen- On the basis of the symmetry, truncatedcone shape, pres- eral NW-SE direction. If this linearity represents the presence ence of a summit crater that was apparently resurfacedby of a fault scarp, evidencefor it is not prominent, nor is there smooth plains, and the lack of adjacent impact craters that additional topography along the strike approachingthis alti- might have producedhigh topography through multiple over- tude that might be relatedto a fault scarp. Another possible lappingimpact crater rims, we find that MOLA data supportear- lier interpretations [e.g., Greeley and Spudis, 1978; Hodges and Moore, 1994] that Feature A is an edifice of volcanic ori- 4 gin. We find no compellingevidence for the previous presence and subsequentstripping of a layer on the order of 2 km in thickness to producethis feature, nor do we find evidence for regional faulting that might have producedthis feature. Adopting the interpretation that the structureis a volcano with a summitcaldera, we estimateda possible original precal- dera size of the volcano by fitting parabolic curvesto the to- pography of its flanks (Figure 5). The curves meet at a maximum peak elevation of •-3.9 km, giving the volcano a possible overall original height of-3 km, comparable to the height of terrestrial stratovolcanoes[Pike, 1978].

4.2. Feature B

To the northeast of Feature A (Plate 1) is a second dissected domical feature (Feature B; 18øS, 184øW) that lacks an obvious summit crater. Its base is approximately the same size (-30 km) as that of Feature A, and the two mountains have similar - 19.8 -20 -20.2 elevation ranges (Feature B is -2 km in height) (compareFea- Viking Latitude ture A in Figures l a, 2a, 3a, and 3b to Feature B in Figures 1b, 2b, 3c, and 3d). Its slope characteristicsand its cross-sectional Figure 5. MOLA profile across FeatureA, the candidatestra- tovolcano (orbit 12263.2). Parabolic curves were fit to the profile are strikingly similar to those of Feature A (compare MOLA laser shot points by a least squaresmethod to recon- Figures 1 and 2). Although no summit crater is obvious from struct its possible original elevation. Vertical exaggeration is the low-resolution Viking images, the MOLA profile reveals a 10x. The minimum reconstructedrelative height of the vol- flat but complex summit structureconsisting of a small summit cano is -3 km but could be higher dependingon the degreeto depression containing a central hill (Figure lb). Although which the flank elevation has been lowered by fluvial dissec- broadly similar to featuressuch as Mount St. Helens, with its tion and mass wasting. 17,512 STEWART AND HEAD: ANCIENT MARTIAN VOLCANOES

origin could be a degradedimpact crater rim. Broad depressions Some of the other diffuse high topography in Aeolis is more exist to the northwest and east, which may representtopogra- likely to be relatedto coalescedcrater rim deposits, but the un- phy associatedwith heavily degradedcraters, but the topogra- usual occurrencejust of Feature A may be ancient addi- phy is very subduedand no other evidence is seen of high tional volcanic constructs that have been heavily modified by topographythat might be due to coalescingrim deposits. subsequentimpact cratering. The similarity of this feature to Feature A in terms of ap- The topographic featuresillustrated in Figure 2 have a size proximate size, height, and flank slopes, its topographic dis- similar to large Earth stratovolcanoes, but Earth stratovolca- tinctivenessfrom the surroundingplains, and its lack of close noes have much steeperflank slopes [Walker, 2000] than do correlation with rugged multiple overlapping impact crater Martian features. This morphology,however, is similar to that rims and prominent fault scarps all support a volcanic origin. predictedfor pyroclastic eruptions under Martian conditions FeatureB differs from A in its more irregular and elongated na- [Wilson and Head, 1994]. We therefore conclude that Features ture, the lack of a distinctive summit crater (although the sum- A and B observedin the Aeolis region of Mars are likely candi- mit is relatively flat), and the more prominent planezes on its dates for a stratovolcanicorigin, with the reservation that their flanks. presently degradedstate (Figure 4) prevents the observation and documentationof features associatedwith pyroclastic de- 4.3. Other Features posits. Although the fYagmentalmaterials and evidence of mass wasting seen at very high resolutionon the flanks of Fea- A number of other topographic highs appear in the region ture A in MOC images(Figure 4) could be pyroclastic deposits, shown in Plate 1 (labeled C1-C5); these are similar to Features the age and degradedstate of the edifice precludesconfident in- A and B in that they are about 20-50 km in diameter and stand terpretationof the origin of these deposits. distinctly (-750-1250 m) higher than the surroundingplains. The apparent ancient age (Noachian) [Scott amt Tanaka, Could these be volcanic edifices or more degradedversions of 1986; Greeley amt Guest, 1987] of these features, if correct, FeaturesA and B? The broad high -180 km southwest of Fea- providesfurther evidencethat volcanic activity was an impor- ture A (Plate 1, C1, lower left) is topographically diffuse, with tant contribution to the formation and evolution of the low slopes and a close association with a large impact crater Noachian highlands. The relatively fresh appearanceof Feature -80 km in diameterjust to the west. This topography appears A couldmean that it is of slightly younger age, co- to be best explained by the additive ejecta and structuraluplift incident with other Hesperian-agedvolcanic and fluvial activ- of this crater, a very ancient one just to the east, and two ity in the region [e.g., Nelson, 2001a, 200lb]. In any case, youngercraters (about 20 and 25 km diameter)just to the south. A second smaller occurrence is seen --150 km east of Feature A the criteria outlined here for the origin of isolated knobs and topographichighs can be usedto help distinguish the range of (Plate 1, labeled C2 in eastern part of the image). This feature origins and the possibility of volcanic edifices in ancient de- is small and irregular in shape (about 10 by 20 km), is low posits elsewhere on Mars. These criteria include (1) size, (2) (-500 m high), and is not obviously associatedwith major im- slope characteristics,(3) evidence of planezesand other dissec- pact craters. Insufficient data are available to assessits origin. tion processescommon to, but not necessarilyuniquely asso- A third occurrenceis observedjust south of Feature A (Plate ciated with, pyroclasticallybuilt volcanoes,and (4) the inabil- 1; labeled C3-C5; central part of image). In this case, there are ity to explain by other hypothesessuch as coalescingimpact several coalescingpositive topographicfeatures forming an ir- crater rims, exhumation, etc. Although there is evidence for regularly shapedstructure. Part of the distinctive topography pyroclastic activity throughoutthe history of Mars, the style is due to the presenceand rim of a relatively fresh 30 km diame- of volcanism in earlier history (Noachian-Hesperian)clearly ter crater that has been superposedon what appearsto be preex- includedmore abundantexamples of pyroclastic activity than isting high topography. The characteristicsof the high topog- later examples [e.g., Greeley and Spudis, 1981; Greeley and raphy are similar in many ways to those of the flanks of Fea- tures A and B; there are channel-like features on the flanks, and Crown, 1990; Crown and Greeley, 1993]. The morphologyand morphometry of FeatureA are consistent with an origin from candidatesfor topographic blocks resembling planezes and depositsof mixed pyroclasticand effusive eruptions. Although possible sector collapse (Plate 1; Figures 2-4). Although sub- the featuresanalyzed in this study are broadly aligned, there is sequentimpact craters have considerably modified the original no evidencethat they are relatedto subductionzones proposed deposits, the remaining high topography stands prominently by Sleep [1994] for this time interval. above the surroundingplains at approximately the same level as the nearby Feature A. These terrain characteristics, and their Acknowledgments. We thank StephenPratt for preparationof data proximity to Feature A, lead us to suggest that these features for use in theseanalyses and for discussionof science resultsand Peter could be remnantsof volcanic edifices perhaps originally com- Neivert for help in preparationof the figures. Gil Ghatan and Brian parable to Feature A but highly modified by subsequentim- Nixon helpedin the preparationof the MOC imagesand in the discussion pacts. Present data are insufficient to addressthis issue; addi- of the results.Jim Zimbelmanprovided an extensiveand helpfulreview. This researchwas supportedby NASA PG & G Grant NAG5-4723 and tional data acquisition involving high-resolution mineralogi- MDAP "Martian Volcanic Edifice Characterization"grant NAG5-9780 cal and imaging information is requiredto clarify this impor- to J.W.H., which are gratefully acknowledged. tant issue. References

5. Discussion and Conclusions Bates, R., and J. Jackson,Dictionary of Geological Terms, Am. Geol. Inst., Washington,D.C., 1984. On the basis of the new MOLA data for individual features Cas,R., andJ. Wright, VolcanicSuccessions, Allen and Unwin, Concord, Mass., 1987. and the topography of the Aeolis region as a whole, we con- Crown,D. A., and R. Greeley,Volcanic of the Hadriaca Patera clude that FeatureA (Zephyria Tholus) is of volcanic origin and and the Eastern Hellas region of Mars, J. Geophys.Res., 98, 3431- that Feature B is also very likely to be of volcanic origin. 3451, 1993. STEWART AND HEAD: ANCIENT MARTIAN VOLCANOES 17,513

Crumpier,L. S., J. W. Head,and J. C. Aubele,Calderas on Mars:Char- Pike,R. J., Volcanoeson the innerplanets: Some preliminary compari- acteristics,structure, and associatedflank deformation,in Voh'ano sonsof grosstopography, Proc. Lunar Planet. Sci. Conf 9th, 3239- Instabilityon the Earth and Other Planets, edited by W. J. McGuireet 3273, 1978. al., Geol.Soc. Spec. Pttbl., 110, 307-348, 1996. Plescia,J. B., Geologyof theUranius Group volcanic constructs: Uranius Davidson,J., and S. de Silva, Co•npositevolcanoes, in Encyclopediaof Patera,, and UraniusTholus, Icarus, 143, 376-396, Volcanoes,edited by H. Sigurdsson,pp. 663-681, Acade•nic, San 2000. Diego, Calif., 2000. Reid,E., S. B. Christian,and D. L. Brien,Gravitational stability of three- Francis,P. W., and C. A. Wood, Absenceof silicic volcanismon Mars: dimensionalstratovolcano edifices, J. Geophys.Res., 105, 6043-6056, Implicationsfor crustalcomposition and volatile abundance, J. Geo- 2000. phys.Res., 87, 9881-9889,1982. Scott,D. H., andK. L. Tanaka,Geological map of thewestern equatorial Greeley,R., andD. A. Crown,Volcanic geology of the TyrrhenaPatera regionof Mars,U.S. Geol. Surv. Misc. Invest. Map, 1-1802A,1986. regionof Mars,J. Geophys.Res., 95, 7133-7149,1990. Sleep,N.H., Martianplate tectonics, J. Geophys. Res., 99, 5639-5655, Greeley,R., andJ. E. Guest,Geological map of the easternequatorial 1994. regionof Mars, U.S.Geol. Sttrv. Misc. Invest. Map, 1-1802B,1987. Smith,D. E., et al., Topographyof thenorthern hemisphere of Mars from Greeley,R., and P. D. Spudis,Volcanism in the crateredterrain henri- the Mars Orbiter Laser Altimeter (MOLA), Science,279, 1686-1692, sphereof Mars,Geophys. Res. Lett., 5, 453-455,1978. 1998. Greeley,R., and P. D. Spudis,, Rev. Geophys.,19, Smith,D. E., et al., The globaltopography of Marsand implications for 13-41, 1981. surfaceevolution, Science,284, 1495-1503, 1999. Greeley,R., N. T. Bridges,D. A. Crown,L. Crumpier,S. A. Fagents,P. J. vanWyk de Vries, B., N. Kerle,and D. Petley,Sector collapse forming Mouginis-Mark,and J. R. Zimbelman,Volcanism on the red planet, at Casitavolcano, Nicaragua, Geology, 28, 167-170,2000. Mars, in EnvironmentalEffects on VolcanicErttptions, edited by J. R. Walker,G. P. L., Basalticvolcanoes and volcanicsystems, in Encyclo- Zimbelmanand T. K. P. Gregg,pp. 113-142,Kluwer Acad., Norwell, pediaof Volcanoes,edited by H. Sigurdsson,pp.283-289, Academic, Mass., 2000. San Diego, Calif., 2000. Gregg,D. R., The geologyof TongariroSubdivision, N. Z. Geol. Surv. Wilson,L., and J. W. Head, Mars:Review and analysisof volcanic Bull., 40, 152 pp., 1960. eruptiontheory and relationships to observed , Rev. Geo- Head, J. W., and L. Wilson, TharsisMontes as compositevolcanoes?: 1. phys.,32, 221-263,1994. The roleof explosivevolcanism in edificeconstruction and implica- Wood,C. A., Calderas:A planetaryperspective, J. Geophys.Res., 89, tions for the volatile contents of edifice-forming magmas, Lunar 8391-8406, 1984. Planet. Sci. [CD-ROM], XXIX, abstract1127, 1998. Zimbelman,J. R., Volcanismon Mars. in Encyclopediaof Volcanoes, Hodges,C. A., andH. Moore,Atlas of volcaniclandforms on Mars, U.S. editedby H. Sigurdsson,pp. 771-783, Academic, San Diego, Calif., Geol.Surv. Prof Pap. 1534, 194 pp., 1994. 2000. Mouginis-Mark,P. J.,L. Wilson,and J. W. Head,Explosive volcanism of Zuber,M., et al., Internalstructure and early thermalevolution of Mars Hecates Tholus,Mars: Investigationof eruptionconditions, J. Geo- from Mars GlobalSurveyor topography and gravity,Science, 287, phys.Res., 87, 9890-9904,1982. 1788-1793, 2000. Nelson,D. M., J. D. Farmer,R. Greeley,R. O. Kuzmin,and H. P. Klein, Durius Valles outflow basin, Mars: Proposedsite for MER-A, LPI J. W. Head and E. M. Stewart,Department of GeologicalSciences, Contrib.1079, pp. 55-56, Lunar and Planet. Inst., Houston,Tex., Brown University, Box 1846, Providence, RI 02912. 2001 a. ([email protected]) Nelson,D. M., J. D. Farmer,and R. Greeley,Hydrologic history of south Elysiumbasin-north Terra Cimmeriaarea, Mars,Lunar Planet. Sci. (ReceivedJuly 18, 2000; revised February 26, 2001; [CD-ROM], XXXII, abstract2069, 200lb. acceptedMarch 9, 2001.)