ICARUS 84, 362-373 (1990)

Recent Water Release in the Tharsis Region of Mars

PETER J. MO UGINIS-MARK

Planetary Geosciences Division. Hawaii Institut e Geophysics, University of Hawaii, Honolulu, Hawaii 96822

Received January 28,1988 ; revised July 17, 1989

Numerous channels have been identified in northwestern Tharsis, Mars, at the base of tbe Olympus Mons escarpment and to the wcst or Ceraunius Fossae. Tbese cbanncls are anastomosing in (onn, and have associated streamlined islands, indicating that they were most likely formed by flowing water. The preferred model for channel formation involves either the tectonic release of water from a deep groundwater system or, less likely, the intrusive heating of deep-seated ice lenses. Because the channels are carved in some of tbe youngest lava flows on Mars, it is believed that water or ice existed at certain places at equatorial latitudes until the recent «1 byr B. P.?) past. li;J [990 A<od.mk .... .,." rile .

INTRODUCTJO

The interaction of magma with water and ground ice appears inescapable if the cli­matic regime of Mars during volcanic ep­ochs was similar to its current state (cf. Sod­erblom and Wenner 1978 , Hodges and Moore 1979, AJlen 1979, (980). Indeed, since the earLies t inves tigations of the Mari­ner 9 data set (Sharp 1973a,b), it has been recognized that the interaction between vol­canism and ground ice , and the subsequent generation of melt water, has probably been an important geologic process on Mars . E x­amples of this proposed interaction include the existence of pseudocraters (Frey et al. 1979, Allen 1980), tablemountaios (Hodges and Moore 1979, Allen 1979) , and several outflow channel systems close to volcanic constructs (Mouginis-Mark 1985, Squyres et al. 1987). Recentl y, Wilhelms and Bald­win (1 989) have also explored the possible role played by igne us sills in the shaping of the Martian uplands and concluded that the non concentrated thermal source provided by intrusions might explain a number of oth­erwise puzzl ing phenomena in this area.

In the case of the channels to the west of El ysium Mons, intrusive act ivity radial to the summit of the volcano was inferred to

362 00t9-1035/90 $3 .00 Copyright © 1990 by Academic Press , Inc. A ll rights of reproduct ion in any fo rm reserved

be the most likely ause of late-stage heating of the ground ice (Mouginis-Mark (985) . For Doa and Harmakhis Valles (to the south of Hadriaca Patera) , Squyres et al. (1 987) at ­tributed the format ion of the channels to the emplacement of lav' flows upon a layer of permafrost, with the release of melt water due to basal heating by the lavas and subse­quent collapse of the overl ying materials. Most of the volcanic centers with associated melt water features are relativel y old (Gree­ley and Spudis 1981. Neukum and Hiller 1981 , Wilhelms and Baldwin 1989), placing the possible melt water re lease at a rela­tivel y earl y time in Martian history. Such an observation supports the idea that vola tiles hav been cold trapped at high latit udes over the history of the planet (FanaJc et at. 1986). Howe ver, recen t studies of smaU channels in the Mangala (Masursky et al. (986) and Memnonia (Scott and Chapman , 1989) re ­gions of Mars have shown that in certain instances young chann Is of probable fluvial origin could form at equatorial latitudes .

This paper describes several additional exception to this hypothesized global trend of geologically early volatile migration to­ward the poles and documents examples of water release that took place upon some of the youngest lava flows within northwestern

363 WATER RELEASE IN THARSIS

Th arsis. The channels desc ribed here are located to the east of the Olympus Mons escarpment and to the west of Ceraunius Fossae, which are areas believed to com­prise Java flows that may be several kilome­ters in total thickness (Scott and Tanaka 1981, Solomon and Head 1982). The materi­als at the base of the Olympus Mons escarp­me nt have been ma pped as a series of young lava flows by Scott et al. (1981 ; U nit Aop) and form a relatively smoot h surface with a very low c rater density (~90 craters > I km d ia meter per 106 km2 ; Scott et al. 1981) . T hese lava fl ows are interpreted to be the youngest sequence in the T harsis region, fo rmed by eruptions from fault s and fis sures in the plains to the east of Ol ympus Mo ns. Because of their location in a young volcanic a rea , the channels described here may have bee n formed by water released by inlrusive events and, by virtue of their young age and near-equatorial locat ion, are likely 10 be of importance for interpreting the recent distri­blltion of volatiles on Mars.

NEW OBSERVATIONS

Analysis of Viking Orbiter images with 2S-21 0 m/pixe l resolution has revealed nu­merous channe ls (Fig . I and Table J) wit hi n

TABLE I

VIKING ORB ITE R IMAGES USED TO

I DENTl rv C H ANNELS

Channel Viking Latitude Longitude Resolution No. Pic. No. (m/pi xe l)

O lympus Mo ns esca rpment I 461S24 16' N I 29°W 51 I 468S53-56 16°N 129°W 24 2 695A1 7-24 13°N 129°W 25 2 890A42 , 71' 13' N 129°W IS7

890A3S" 200 N 12SoW 167 Wes te rn Ceraunius Fossae

4 S16A 14-16' 26°N II So W 20S 4 39847 25°N J IS'W 106 4 48B52 21 ' N 12 1' W 12 1 5 623A t3-20 25°N 122°W 7S 6 516A33-3S* 22°N 11 3°W 196 7 890A04 23°N 122'W 178

NOle. ,. Due to lim itati o ns in image resolution . these frames were used 10 give only a le ntativ e idenlifica lio n of Huid channels.

northwestern Thars is that are s imil ar in morphology to those observed in the distal reaches of the Elysium Planitia channels (Figure 12 of Mouginis-Mark 1985) and the channels north of Harma khis Vallis (Fig. IS of Squyres et al. 1987) . While a nonfluvia l origin (such as lava channels) for these fea­tures remain s poss ible, the lack of late ral levees, flow lobes, or other constructional feature s argues against a volcanic origin , while the existence of s treamlined isl and. and braided floors s uggest that the channel s were water carved . In the absence of st rong morphologic evidence to the contrary , I wiJl therefore assume that the c ha nnels de­scribed here were indeed fluvial in o rigin.

Two of the most intriguing channels iden­tified are located within 20 km of the base of the Ol ympus Mo ns escarpment, at I3°N, 129°W and at 16°N 129°W. A sma ll outflow channel (channel I, Table I) is associated with an arcuate graben located a t 16°N 129°W. which is ~ is km from (he base of the escarpment (Fig. 2). This channel is ~8S

km in length, is braided , and compri e s three main ,egme nt s , each with wid ths of 300- 1,600 m . N umerous st ream li ne Ll isl ands (Fig. 3), similar to the examples described by Baker and Kochel (1979) and by Koc hel (1981), a re fo und withi n a d istance of 20 km fro m the open end of the gra be n, and a series of cataracts can also be seen. Atmos pheric haze precludes the confi den t determination of depth s fo r all the channel segments, but digital measurement of shadow lengths from the ca librated version of Viking image 468S53 indicates tha t the c hannel sy ste m is qui te shallow . Five scarps were measured (see Fig. 3A), a nd gave a n average de pt h of ~23 m.

This ave rage depth pe rmit s a n estimate f the channel volume to be de rived. Measure­ments of all the segments of channel I give a surface area for the e ntire system of ~SO

km2• Assuming tha t all of the channels a re the same de pth (clearly only a generaliza­tion, s ince locally the cataracts and distribu­taries identified in Fig. 3B appear to have very different depth s), the volume of mate­

)~

)

", . 3 .o~· ',0', ~

~ ; ~:~:o;

-:~:.".:'>~ .. . " ~:' .

1 ,,'

17

o 0"

0°'<.

/'"~ (f U

"'-Ii ~lr @ JOVIS THOLUS

MERCATOR PROJECnON 300' -­ -­ - 200 300 400' 500 600 700 600 900 1000 KLOMElffiS

1O~_ /' 2 ,: ;> ; --,

w 0"­~

'1:1 tTl -l tTl ~

~

s:: o c o Z-Vl

~ >;;d ~

FIG.!. Distribution map of the fluvial channels identified within northwestern Tharsis in this analysis, with the channel numbers keyed to Table 1. Where it can be confide[]tly identified , each arrow indicates the direction of water flow. Random stipple areas mark the locations of Olympus Mons aureole deposits. The fracture s associated with Tharsis doming are also show[].

365 WATER RELEASE IN THARSIS

FJG. 2. High resolution (24 m/pixel) view of the arcua te graben located 15 km from the Olympus Mons escarpment that has probably ac ted as a conduit for the released water that formed channel I . Areas outlined denote locations of Figs. 3 and 4. Viking Orbiter frame 468S53. Image width is equivalent to ~23 km. North is to top left of image.

366 PETER J. MOUGINIS-MARK

FIG. 3. (A) Computer enha nce ment of subscene fro m Viking image 468S53 (see Fig. 2 for location). A 155 x 15 5 pi xe l hi gh-pass filt er, fo llowed by a linear s tre tch from DN = 65 to 165 were applied to the data after re mova l of pixel drop o uts. Arrows show the loca tions where s hadow length measurements ""'ere made. Image width equivalent to 13.4 km. (8) Morpho logic s ketch of the braid ed portion of channell (area is the same as Fig. 3A). Although the direction offlow cannot be persuasively identi fied , two features suggest s tha t the flow direction was from the no rtheast to southwest (top to botto m of image) . At "A " the shape of the segment s of the channe l a re more consis tent wi th a merging tlow th an divergent flow. The area of cata ract s (cross-hatc hed area to eas t of point "A" ) appears to deepen toward the south, aga in supporting the idea of flow from the northeas t. Ho wever, other landforms a re e ither nonin fo rma ti ve or could suggest flow to ward the northeas t : streamlined island s (d otted areas) do not show an y preferred orientation or flow st ructures, while severa l sink holes ("S") co uld eq uall y be interpreted as di s tributed source area s for water flowing northward. Barbs s how drop in to pography. dashed lines indica te areas where boundaries are onl y approximate. Note the va riable channel depth (rela tively shallow at "x" and deep a t "y").

367 WATER RELEASE IN THARSIS

B

N

\

A

FIG . 3-Conlinued

ri a l removed from channell is ~ 1.15 km3.

Ta king Komar's (1980) estimates for the erosional efficiency of surface water flow on Mars, a minimum estimate for the volume of water released to create channel I would be ~3 km 3 (erosion efficiency of 40%), while a more plausible volume estimate for the water released from channell would be ~ 12 km' (erosion efficiency of 10%).

Two alternative interpretations are possi­

ble for the origin of the water that is pre­sumed to have carved channel I, and as such have a bearing on the potential role of late­stage igneous activity producing this water release: (1) That the water was released from the graben, which acted as a point source; of (2) That there was a distributed set of sources for the water and that the graben acted as a major sink for the majority of the water. Unlike the Elysium Fossae

368 Pl':TER 1. MOUGINIS-MARK

graben and Dao Vallis (Mouginis-Mark 1985, Squyres et al. 1987), no landforms in­dicative of large-volume water release at a source, such as chaotic terrain, can be found at the inferred source of channell (Fig. 4). Nevertheless, the incision of cataracts on the channel floor which evidently became progressively deeper toward the south (Fig. 3), the streamlined shape of a major conflu­ence within the braided channel ("A" in Fig. 3B), and the observation that flow to­ward the south is consistent with the flow directions oflava flows in this area (Mougin­is-Mark ef al. 1982) all suggest that the gra­ben was the point source for the water.

The ultimate fate of the surface water is unclear, since many of the channel segments appear to braid into subdued deltas at their southern extents or may be covered by sub­sequent deposits at their distal ends. Several channel segments do, however, end in what appear to be sink holes or ponds (Fig. 3B), which could have allowed some of this water to have percolated back underground, or to have permitted evaporation from confined basins.

A second series of small outflow channels (channel 2, Table I) has also been identified to the south of this first locality, within 120 km of the Olympus Mons escarpment at l3°N, 129°W. This set of channels also ap­pears to originate from a fracture in the young lava flows surrounding Olympus Mons, and has produced a series of braided distributaries ~6S km in length. Further­more, although image resolution does not permit detailed analysis, there are several additional graben to the east of Olympus Mons that appear from IS0-220 m/pixel Viking images to be likely candidates for water outflow channels (Fig. 1). Channel 3 (Table I) which is associated with graben to the east of parts of the aureole material (at 20oN, 12S0W) is the most likely candidate for a water channel in this area.

Despite the lack of Viking images compa­rable in resolution to those used to investi­gate channell, additional examples of prob­able fluid flow at some time after the

eruption of the Aop lava flows in the Tharsis region (Scott et al. 1981) have also been studied here. Principal among the pre­viously undocumented channels is a 400­km-long braided system (channel 4, Table I) to the west of Ceraunius Fossae that in­cludes Olympica Fossae (Fig. S). Viking Or­biter images with a resolution of 7S-20S m/ pixel reveal that in addition to the main channel, an anastomosing network of sub­dued channels ~20 km wide parallel O lym­pica Fossae for over 300 km. Both channel systems originate in a fracture located at 26°N, I 11 oW that is part of the complex of graben associated with Ceraunius Fossae. For the main Olympica Fossae channel, the maximum width is ~4 km, and in several places the channel is divided into 3-S sub­channels. Streamlined islands also occur within Olympica Fossae, and the subdued channeling of the rim indicates that multiple episodes of bank overspiH have taken place. Unfortunately, the low resolution and high sun angle of the Viking images precludes an estimate of the depth of Olympica Fossae.

T he Olympica Fossae channel is but one of three subparallel channels that originate to the west of Ceraunius Fossae and flow in a generally west-southwest direction toward Olympus Mons (Fig. 1). Between Olympica Fossae and Halex Fossae, at 2soN, 122°W, additional channels can also be found that are a few hundred meters in width and ~4S0 km in length (channel S, Table I). These channels appear to be preferentially located within an area of hummocky terrain that is different in morphology from the adjacent lava flows. SmaJl channels (channels 6 and 7 in Table 1) have also been found to the south and southwest of Olympica Fossae.

MODELS FOR WATER RELEASE

Although the surface flow of relatively small volumes (probably a few cubic kilome­ters) of water is evident from the Viking images, it is not immediately clear by what mechanism the original water or ice could have remained at this latitude and proximity

369 WATER RELEASE IN THARSIS

FIG. 4. Upper reaches of arcuate graben (see Fig. 2 for location) that is interpreted to be the source for channel J. Note that no How features are visible on the graben Hoor and that there is no evidence of hummocky terrain on the Hoor that could be associated with the di sruption of surface material s by the release of ground water. Width of image is equivalent to - 12.7 km. Image processed in the same manner as Fig. 3A, and is also a subscene of Viking frame 468S53.

to the Tharsis volcanoes until relatively re­cent times . Five possible models for the source of the subsurface volatiles have been in vestigated:

Model 1. Igneous activity supplied heat which modified shallow (top few hundred meters?) ice lenses which were relic from early Martian history. In this instance, wa­ter release at this latitude would mean that either the existing theoretical models for volatile migration on Mars (Fanale el al. 1986) require modification or that, for some unknown reason (such as unusually com­pact soils or the intercalated nature of the lava flows), the expected kinetics of ice withdrawal did not occur effectively in all equatorial sites.

Model 2. Igneous activity supplied heat which mobilized more deeply buried equa­torial ice lenses which were relic from early Martian history. These lenses could origi­

nally have been at a shallow depth, but were subsequently more deeply buried (perhaps to more than a kilometer?) for most of their history and were exposed by the faulting associated with the graben in this area. In this instance , the general model for global volatile migration remai ns intact but also needs to be modified to consider local tem­poral variations of surficial geology due to the emplacement of younger lava flows and, possibly, aeolian materials.

Model 3. Relic ice lenses were heated by an igneous event, but these ice lenses were not truly remnants of the earliest Martian atmosphere (i.e., ice that was distributed randomly across the planet during the late heavy bombardment of Ma rs). Rat her , these ice lenses could have originated from, say, volatile deposition following more recent volcanism when certain Tharsis volcanoes such as Alba Patera were probably erupting

370 PETER 1. MOUGINIS-MARK

F IG. 5. Pa rt of the middle reaches of Olympica Fossae (cha nne I4 , Tab le I) , s howing the main bra id ed c ha nne l. Direction of wate r fi ow is fro m middle right to bottom left. Arrowed is a second , muc h m ore subdued c hannel sys te m (c ha nne l 5 , Ta ble l) that pa ra llels O lympica F ossae for over 300 km . Viking O rbiter frame 39B47, centered a t 25°N, I 15°W. Image width equivalent to 80 km , large arrow points towa rd no rth .

371 WATER RELEASE IN THARSIS

considerable volumes of water vapor (Wil­son and Mouginis-Mark 1987, Mouginis­Mark et al. 1988).

Model4 . The release ofground water may have been unrelated to an igneous event, but took place within northwestern Tharsis due to tectonic deformation of the area due to loading of the surrounding plains by Olympus Mons. Deformation of the plains surrounding Elysium Mons has been inves­tigated by Hall et af. (1986), and it is possible that loading of the adjacent plains by Olym­pus Mons and Alba Patera also led to frac­turing of the adjacent surface materials. Model 4 also draws upon the model of Clif­ford (1987) in that the observed water could have been associated with the release of wa­ter transported over regional distances.

Model 5 . For channels I and 2 the release of water may have come from a relic ice layer entrained within the basal materials that comprise the Olympus Mons escarp­ment. It has been hypothesized (Hodges and Moore 1979) that the basal layers of Olym­pus Mons may have had a subglacial origin, and that basal melting of ice may have pro­moted the gravity sliding that produced the Olympus Mons aureole materials (Lopes et al. 1982, Tanaka 1985). Water from this basal melting within Olympus Mons may thus have migrated to the surrounding ter­rain, where subsequent intrusions of fault­ing permitted the melt water to reach the surface.

Geomorphic analysis of the Viking image data permits these different models for ice storage and melt water release to be investi­gated. It is concluded (see below) that the most plausible models involve either heating of deep-seated ice lenses (Model 2), or re­lease from a deep ground water system (Model 4) , perhaps fed in the case of chan­nels I and 2 by subsUliace runoff from Olympus Mons (Model 5). Model 1 is be­lieved to be unlikely because, were such shallow ice lenses still widely preserved on Mars, it is expected that the emplacement of the larger individual lava flows upon this terrain would also have been accompanied

by the formation of pseudocraters or topo­graphic depressions, or that the flows would have liberated sizable volumes of melt water at the surface (Allen 1980, Squyres et al. 1987); such phenomena are not seen even in the highest resolution Viking images of lava flows located within Tharsis (Schaber 1980, Thelig and Greeley 1986). Model 3 (recent deposition of volcanic fume) is also consid­ered to be unlikely by virtue of the distance from the major vents on Olympus Mons (- 350 km, and more than 25 km variation in elevation) and the very localized nature of these outflow channels. Were fume-charged ice lenses (i.e., ice deposits accumulated downwind of volcanoes erupting water-rich magmas) to be produced in this manner late in Martian history , it is believed that they would be more widespread and concen­trated around the summits of the volcanoes rather than in specific locations at radial dis­tances of several hundred kilometers.

In Model 2, a dike intrusion along a sub­surface line of weakness, expressed at the sUliace as a graben, could be the cause of melt water generation and release at the sur­face. Alternatively, as is the case for certain other Martian channels (e.g., Mangala Vallis; Carr 1981), water release could be unrelated to igneous activity. 11 is thus pos­sible that the loading of the Martian crust by Olympus Mons could have allowed water from depth (Model 4), or from subsurface runoff from Olympus Mons (Model 5), to reach the surface. Support for Models 2 and 4 also comes from the analysis of the other channel systems described here . In the case of Olympica Fossae (Fig. 5), water appears to have been released from graben in west­ern Ceraunius Fossae, comparable to the source area for Mangala Vallis. Compared to channels 1 and 2, there is no direct equiva­lent to Model 5 (subsurface water flow off a volcanic construct) for Olympica Fossae. Either the channel systems have different origins or the released water that carved channels I and 2 did not originate from within the flanks of Olympus Mons.

Some consideration of the relative likeli­

372 PETER J. MOUGINIS-MARK

hood of Model 2 and Model 4 is also possi­ble. Although the geometry of Martian dikes and siJls can only be inferred from terrestrial analogs (Knight et al. 1988), it is likely that the horizontal (and, possibly, the vertical) interface between a Martian dike/sill and an ice lens would be locally extensive. Such an interaction should thus release melt water over a large area of the surface from multiple sources. Since channell appears to origi­nate from a point source (Fig. 2), and Olym­pica Fossae are located almost 1500 km from Alba Patera (the nearest volcanic con­struct), an intrusive event is considered less likely as the cause of the melt water release (Model 2) than tectonic release of deep­seated water (Model 4) . Image data do not permit Model2 to be entirely ruled out, how­ever, and the volume of water released at channel I (- 3-12 km}) is well within the estimates of melt water release calculated by Squyres et al. (1987) for comparable vol­cano-ground ice interactions south of Ha­driaca Patera.

CONCLUSIONS

This analysis of fluvial channels demon­strates that water release took place in iso­lated areas within northwestern Tharsis un­til the relatively recent geologic past , and included episodes that postdate the em­placement of some of the youngest lava flows on the planet. While the origin and physical setting of the water remains un­clear, the current observations lend addi­tional constraints to our knowledge of the volatile history of Mars: (I) that ice lenses or deep-seated aquifers were probably pre­served for extended periods of time at depths of perhaps a kilometer or more over much of the planet, including near-equato­rial latitudes; (2) that in certain uery specific localities geologic conditions were such that this water could be released to the surface even in geologically recent times; and (3) that while dike intrusions may have served as a catalyst for melt water release at the surface, there is no compelling reason to infer that intrusive events were a necessary

component in this process. Thus Martian dikes hundreds of kilometers in horizontal extent need not be hypothesized to exist on the basis of the locations of water release within the Tharsis region .

ACKNOWLEDGM E NTS

I thank Mark Robin son fo r making the shadow le ngth m easureme nts of cha nnel I and for the pre paration of Figs . 2-4; Cassandra Coombs, Fraser Fanale. a nd Su san Postawko for their comme nts on a n earl y draft of thi s manu script ; and David Pieri and Peter Schult z who provided helpful revie ws. This work was sup­ported by NASA Grant NAGW-437 from th e PlanetHry Geology Program. Thi s is Hawaii Institute ofGeo ph ys­ics Contribution No. 2211.

RE FERENCES

ALLEN , C. C. 1979 . Volca no-ice int e ractions on Mars . 1. Ceophys. R es. 84, 8048-8059.

ALLEN, C. C. 1980. Volcano-Ice Intera ctions 011 Earth nnd Mars . Ph .D . Thesis , Unive rs ity of Arizona , NASA Tech. Me m. 81979 , pp. 163- 264 .

BAK ER, V. R. , AND R. C. KOCH EL 1979. Martian cha n­ne l morpho logy: Maja aod Kasei Valles . 1. Ceophys. Res. 84, 7961-7983 .

CARR, M. H. 1981. The Slllj(JC e oIM"rs, Ya le Uni v. Press , New Ha ven.

CLI FFORD 1987. Pola r basal me lting on Mars. 1. Ceo­phys. R n. 92, 9135-9152.

FANALE, F. P . , J. R. SALVAIL, A. P. ZENT , AND S . E. POSTAWKO 1986. Global di s tribution a nd migra tioo of Slibslilface ice o n Mars. Icar/ls 67, 1-1 8 .

FREY , H ., B. L. LOWRY. ANDS. A. CHASE 1979. Pseu­docra te rs 00 Ma rs. 1. CeopiJys . R!'s . 84, 8075-8086.

GR EE LEY , R., AND P. D. SPUDIS 1981. Volcani sm on Mars. R ev . Ceophys . Sp ace Phys. 19, 13-41.

HALL , J. L., S . C. SOLOMON, AN D J . W . H EAD 1986 . Ely s ium region, Mars: Tests of lithosphe ric loading models for the fornl a tion of tectonic fea tures. 1. Ceo­phys. R es . 91, 11 ,377-ll ,392.

HODG ES , C. A., AND H . J. MOORE 1979. The subglac ia l birth ofOl ymplis Moos and it s aureoles . 1. Ceophys. Res. 84, 8061-8074.

K OCHEL , P . D. 198 1. Streamlined isl8nds: An 80al ys is of the ir minimum-drag s hape . Report Plul7et. Ceol . Pmg. -1981, NASA Tech . Mem . 8421 I, 266-268. [Abstract ]

KOMAR, P. D. 1980. Mode s of sediment tra nsport in cha nne lized water fl o ws with ra mifications to the erosion of Martia n outflow chaonels . Icarus 42, 3 17-329.

KNIGHT , M. D. , G. P. L. WALK ER, P. J. MOUGINI S­MARK , AND S. K. ROWLAND 1988. Rift s of dee ply eroded Ha wa iian basa ltic shie ld s : A s truc tural a nalog

373 WATER RELEASE IN THARSIS

for large martian volcanoes . N ASA MEVTV Pro­gram Working Group Meeting: Volcanism on Mars, Oahu, Hawaii, June 1988. Lunar Planetary Institute ContI'. 660, 23-25. (Abstract]

LOPES, R., J. E. GUEST, K. HILLER, AND G . NEUKUM . 1982. Further evidence for a mass movement origin of the Olympus Mons aureole. 1. Geophys. Res. 87, 9917-9928.

MASURSKY, H. , M. G. CHAPMAN, A. L. DIAL , AND M . E. STROBELL 1986. Ages of rocks and channels in prospective martian landing sites of Mangala Valles region LUllar Planer. Sci., xvn, 520-521. (Abstract]

MOUGINIS-MARK, P. J . 1985. Volcano/ground ice inter­actions in Elysium Planitia , Mars. Icarus 64, 265­284.

MOUGINIS-MARK, P. J., S. H . ZISK, AND G. S. DOWNS J982. Ancient and modern slopes in the T harsis region of Mars. Nature 297 , 546­

550. MOUGINIS-MARK , P. J . , L. WILSON, ANDJ. R . ZIMBEL­

MAN J988 . Polygenic eruptions on Alba Patera, Mars. Bull. Volcanol. 50, 361-379.

NEUKUM, G., AND K. HILLER 1981. Martian ages. 1. Geophys. Res. 86, 3097-3121.

SCHABER , G . G. 1980. Radar, visual a nd thermal char· acteristics of Mars: Rough planar surfaces. Icarus 42, 159- 184.

SCOTT, D. H., AND M. G. CHAPMAN 1989. Geologic setting of an unu sual martia n channel: Hypotheses on origin. Proc. Lunar Planer. Sci. ConI 19th, pp. 377-382. Lunar Planetary Institule, Houston,

TX. SCOTT, D. H. , G. G. SCHABER, K. C. HORSTMAN ,

A. L. DIAL, AND K. L. T ANAKA 1981. Map showing lava flow s in the northwest part of the T harsis quad­

rangle of Mars. U.S. Geological Survey Miscella­neous Map 1-267.

SCOTT, D. H., AND K . L. TANAKA 1981. Mars : Paleo­stratigraphic res tora tion of buried surfaces in Thars is Montes. Icarus 45, 304-319.

SHARP, R. P. 1973a. Mars: Troughed terrain. 1. Geo­phys. Res. 78,4063-4072.

SHARP, R. P. 1973b. Mars: Fretted and chaotic terrains. 1. Geophys. Res. 78, 4073-4083.

SODERBLOM, L. A. , AND D. A . W ENNER 1978. Possible fossil H20 liquid water interfaces in the Martian

crust. Icarus 34, 622-637. SOLOMON, S. c., AND J . W. HEAD 1982 . Evolution

of the Tharsis Province of Mars: The importance of heterogeneou s lithospheric thickness and vol­canic construction. 1. Geophys. Res. 87, 9755­

9774.

SQUYRES, S. W., D. E . WILHELMS , AND A. C. Moos­MAN 1987. Large-scale volcano-ground ice interac­tions on M ars. Icarus 70, 385-408 .

TANAKA, K. L. 1985. Ice-lubricated gravity spreading of the Olympus Mons aureole deposits . Icarus 62, 191 - 206.

T HEIUG , E. , AND R. GREELEY 1986. Lava flows on Ma rs: Analysis of s mall surface features and compar­isons with terrestrial a nalogs. Proc . LUllar Plaller. Sci. COllI 17th, ParI I , 1. Geophys. Res . 91, EI93- E206 .

WILHELMS , D. E., AND R. J. BALDWIN 1989. The role of igneous sills in shaping the martian uplands. Proc. LUllar Planer. Sci. COllI 191h , pp. 355-365 , Lunar Planetary Institute , Houston, TX.

WILSON, L. , AND P. J. MOUGINI S- MARK 1987. Volca­nic input to the atmosphere from Alba Patera on

Mars . Nature 330,354-357.