Comparison of Volcanic Features of Elysium (Mars) and Tibesti (Earth)

Comparison of volcanic features of Elysium (Mars) and Tibesti (Earth)

MICHAEL C. MALIN* Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, California 91125

ABSTRACT the Tharsis region, which represent the Emi Koussi (19.7°N, 18.5°E; Fig. 1), largest and most conspicuous examples of situated at the extreme southern portion of The Elysium volcanic province on Mars martian volcanism (McCauley and others, the volcanic region, ranges from 60 to 80 and the Tibesti volcanic province in Chad, 1972; Carr, 1973, 1974). Comparison with km across and consists of 2,000 m of vol- Africa, were studied using Mariner 9, terrestrial volcanoes, especially those on the canics resting on Paleozoic and Cretaceous Landsat and Apollo photography. Elysium island of Hawaii, has been most fruitful sandstones which have been uplifted 1,500 Mons on Mars and Emi Koussi on Earth (Greeley, 1973). m. The original cone may have reached as show remarkable similarities in summit In this paper, another martian volcanic much as 4,000 m above sea level but is now caldera and flank morphologies. Each has a province will be studied. The Elysium re- only 3,415 m high, with a large (15-km), large central caldera —12 km in diameter gion has several structures which are not multiple-crater caldera some 500 m deep at and from 500 to 1,000 m deep; both cal- found elsewhere on Mars, as well as some the summit. Much of the volcanism oc- deras contain numerous craters and large, which are similar to those of Tharsis. Com- curred during the middle and late Tertiary, irregular pits. Channel-like features which parison will be made with terrestrial fea- with only limited activity during the head at the calderas and taper downslope tures of the Tibesti region of northern Holocene. Current activity is restricted to a show evidence of collapse and possible lava Chad, Africa, which displays a variety of few fumaroles near the southern base of the erosion. Elysium Mons rises some 14 ± 1.5 volcanic forms similar to those of Elysium. volcano. Most of the erosional dissection of km above its base, and the summit is about After establishing the regional setting of the volcano's flanks (Fig. 1) occurred dur- 20 km above the 6.1-mbar mean martian the volcanic fields, two volcanoes, Emi ing the early Pleistocene. pressure surface. Crater size/frequency Koussi on Earth and Elysium Mons on The Tibesti region has experienced ex- analysis indicates most of the craters are of Mars, will be compared in detail. Three tensive tectonic and erosional activity. endogenic origin. The subdued, hummocky main topics will then be addressed: (1) evi- Where present, Paleozoic (Cambro-Or- terrain on the flanks are distinctly different dence for silicic volcanism on Mars, (2) the dovician) sandstones rest unconform- from the slopes of the younger Tharsis tectonic situation of volcanoes situated on ably on an eroded Precambrian surface Ridge volcanoes, showing little if any sign stable (that is, nonmoving) crust, and (3) and dip away from the uplifted massif. of recent material flow. the evidence against fluvial erosion in the Paleozoic formations are in general totally The lack of aqueous erosional forms on relatively recent past on Mars. missing north of the east-west branch and Elysium Mons argues strongly against re- immediately west of the lower portion of cent (~ 105 to 106 yr) pluvial episodes. The REGIONAL SETTING the north-south branch of the province, forms and associations of features through- OF TIBESTI (EARTH) the result of post-Carboniferous, pre- out the Elysium region suggest that central Cretaceous warping, uplift, and subse- volcanism started earlier in Elysium than in Several reports (Geze and others, 1959; quent erosion. Prior to the Lutetian (lower Tharsis and that the source of the Elysium Vincent, 1960, 1970) have presented de- Eocene), the massif was cut by northeast- volcanics has been chemically evolved, with tailed descriptions and interpretations of and north-northeast-trending faults which evidence of silicic magma. Finally, the data the Tibesti region of northern Chad, Africa. follow Precambrian trends. Subsequent up- are consistent with the view that the mar- In this section, only a brief review of some lift along a north-northwest axis was fol- tian crust has been stable and essentially of their results will be presented. lowed by the overall tilting of the entire motionless for an extended period of mar- The Tibesti volcanic province occupies block toward the north-northeast. Sig- tian geologic time. approximately one-third of a triangular up- nificant erosion then occurred, forming land region some 100,000 km2 in area, ap- much of the present topographic relief and INTRODUCTION proximately 1,100 km south of the establishing the drainage system still Mediterranean Sea and 2,000 km west of preserved throughout most of the area The discovery of large volcanoes on the the Red Sea. The highest point is Emi (Geze and others, 1959; Vincent, 1960, planet Mars was one of the most important Koussi at 3,415 m; numerous areas are 1970; Hagedorn, 1971). This erosion has and exciting results of the Mariner 9 mis- more than 2,000 m high. been placed roughly at the boundary be- sion (Masursky and others, 1972; Most volcanoes in Tibesti exhibit two tween the Pliocene and Pleistocene. McCauley and others, 1972; Masursky, distinctive characteristics: wide, low-slope, The volcanism is believed to have started 1973). Much attention has been focused on shield-like profiles and large, central cal- during the very early Tertiary with the ex- Olympus Mons and the other volcanoes of deras (s 10-km diam). Unlike Hawaii, in trusion of basalts and andesites in floods of some places the Tibesti shields appear to be probable fissural origin. This phase was fol- " Present address: Planetology and Oceanography Section, Jet Propulsion Laboratory, Pasadena, Califor- formed of an ignimbritic mass covering lowed by the initial formation of large, cen- nia 91103. older volcanics. tral volcanoes through the emission of a se-

Geological Society of America Bulletin, v. 88, p. 908-919, 7 figs., July 1977, Doc. no. 70702.

908

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/88/7/908/3429828/i0016-7606-88-7-908.pdf by guest on 28 September 2021 COMPARISON OF VOLCANIC FEATURES: ELYSIUM (MARS) AND TIBESTI (EARTH) 909

Figure 1. Landsat photograph: Emi Koussi and surroundings, Earth. Earth Resources Technology Satellite (Landsat) multispectral scanner photograph of southeastern Tibesti showing Emi Koussi in lower right. Volcano is about 70 km in diameter with 12 x 15 km caldera. Recent basalts appear very dark in this infrared photograph. Massive valley incision occurred at the beginning of the Pleistocene. Portions of the Yega, Oyoye, Toon, and Tieroko volcanoes can be seen in the upper left section of the image. For photographic support data, see Table 1.

quence of trachyandesites, trachyphono- basaltic flows on the slopes of, and within dera), and the deposition of the sodium lites, and some rhyolites and of essentially the old valleys around, the caldera. Vol- carbonates within these craters occurred tholeiitic basalts — labradorite-rich, canic activity within the province has con- during the Holocene (Quaternary), and olivine-poor porphyries with abundant tinued throughout the Quaternary primar- fumarolic activity is still present throughout magnetite and ilmenite. Several of the ig- ily in the form of a hybrid volcanism in the the province. nimbrite sheets postdate the erosion of the west [potassic trachyandesites preceding older shields but are older than the massive fluid doreites (an olivine-bearing sub- REGIONAL SETTING valley incision which occurred at the begin- alkaline trachyandesite) and andesites] and OF ELYSIUM (MARS) ning of the Pleistocene. The deposition of in the east by the formation of a second ignimbrites at Koussi was apparently as- caldera at Koussi. The explosions which The Elysium Planitia is dominated by a sociated with the collapse formation of the formed the craters Trau au natron and Era group of three volcanoes located near caldera and was followed by a period of Kohor (the latter within the Koussi cal- 210°W long between 18° and 32°N lat (Fig.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/88/7/908/3429828/i0016-7606-88-7-908.pdf by guest on 28 September 2021 910 M. C. MALIN

2). The region consists of a broad, elliptical TABLE 1. PHOTOGRAPHIC DATA dome (2,400 x 1,700 km, major axis north-northeast), similar to the Tharsis Spacecraft Frame Lat Long Picture Picture Sun ele- Negative Ridge, with the three volcanoes occupying height width vation generation (km) (km) angle the triangular summit. The crest of this broad dome would lie approximately 6 km Apollo 7 5-1621 (1) 19.92°N 18.54°E 120 150 14° 5 above the mean martian pressure surface, Landsat 1 were it not for the presence of the large vol- RBV 1010-08425-1 (1) 20.18°N 18.20°E 185 185 31° 4 canic constructs. MSS 1064-08425-6 (1) 20.15°N 18.25°E 185 185 36° 4 The northernmost volcano, Hecates f 32° Tholus, is about 180 km in diameter. It is a Mariner 9 6391738 (9, ) 21.90°N 213.90°W 439 897 3 8982649 29.43°N 221.09°W 565 721 32° flat-topped, dome-like volcano with rela- 9054539 (2) 29.50°N 211.67°W 561 718 32° tively steep slopes that are convex in profile. 9126429 29.70°N 202.46°W 560 719 33° A UVS altimetry scan which, fortuitously, 7651453 (3,*) 24.54°N 216.94°W 528 684 27° traced across the dome showed its height to 7651593 (3) 23.58°N 204.95°W 632 732 17° be about 6 km (Hord and others, 1974). 7651103 17.05°N 216.45°W 430 554 31° The dome is in sharp and abrupt contact 7722993 (4) 17.35°N 208.07°W 430 555 32° with plains to the south and west. An em- 7723278 (9,f) 16.90°N 204.30°W 570 651 28° bayment into the dome, flooded by plains 6391803 (5,*") 27.30°N 213.20°W 480 1007 28° material (but not to the level of the sur- 13496013 (6) 24.18°N 213.06°W 61 76 9° 13496083 (7) 24.97°N 213.19°W 64 79 9° rounding plain), appears in the northwest, 13496118 (7) 24.53°N 213.02°W 641 825 10° and may have originated through erosion of 13496293 (8) 31.61°N 209.84°W 69 87 15° the slope of the volcano, or through 13496363 (8) 31.89°N 211.04°W 72 90 15° modification of a pre-existing depression immediately interior to the dome boundary (1) Emi Koussi, Figures 1, 7 (6) Elysium Mons, Figure 4 (Fig. 3). The northern terminus is indistinct. (2) Hecates Tholus, Figure 2 (7) Elysium Mons, Figures 5, 7 (8) Hecates Tholus, Figure 3 The eastern edge is marked by a narrow, (3) Elysium Mons, Figures 2, 4 (4) Albor Tholus, Figure 2 (9) Albor Tholus sinuous channel-like form 60 km long, run- (5) Elysium Mons ('f) Photogrammetric pairs ning roughly north-south. This in turn heads in the south into a graben (45 km long, 7 km wide) which delineates the rest of the eastern boundary of the dome. ments radial to the caldera and some faint coalescing chains of craters. However, the The surface of Hecates Tholus appears channel-like forms on the slopes. majority of the depressions which give rise rough in both high and low resolution im- The third volcano, Elysium Mons (Fig. to the lineated appearance are areas be- ages. The surface shows numerous 4 to 4), is 170 km in diameter and has a single tween closely spaced, elongate, linear 10-km-diam craters but is deficient in larger central caldera about 14 km across. mounds of low relief, typically 1 km in craters which are prevalent on the sur- Elysium Mons is 14 ± 1.5 km high (as de- length and 200 to 400 m across. rounding plain. Surface lineations appear termined by photogrammetric techniques The Elysium Plain, which surrounds the radial from a 12-km complex caldera which described by Blasius, 1973) and topograph- three volcanoes, is moderately cratered; it is offset to the southwest from the geomet- ically asymmetric. The asymmetry is in the has abundant large (>20 km) craters. It is ric center of the dome by 30 km. The UVS form of two distinct ridges which meet at believed to be representative of the second data suggest that the summit of the dome is the caldera, the shorter trending west and oldest plains unit on Mars (Carr and others, also offset to the southwest. Numerous the longer southeast. The boundary be- 1973; Soderblom and others, 1974). A channels are seen on the slopes, many tween the volcano and surrounding plain is major structural trend (northwest- formed by coalescing craters. A number of indistinct on all sides, although an apparent southeast, aligned with the minor axis of graben, concentric about the caldera, are fault scarp may delineate the northeast por- the regional dome) appears to postdate this also seen on the slopes of the dome. At high tion of the perimeter. oldest unit. Similarly aligned are several resolution, features reminiscent of the In low-resolution images, the surface of large, closed, linear depressions, the vol- Olympus Mons flows are interspersed Elysium Mons is roughly textured by what canoes Elysium Mons and Albor Tholus, through a rough, knobby terrain (Fig. 3) appear to be large hummocks with ~ 10-km and the elongate form of Elysium Mons. Albor Tholus, the southern member of spacing arranged in a crudely concentric Immediately north of Elysium Mons, and the group, is about 130 km in diameter and pattern. Craters in the size range 4 to 10 km extending some 200 km, is a region of 3 km in height (determined photogrammet- are more sparse than on Hecates Tholus but hummocky terrain of slight relief, showing rically). It has a complex central caldera, more abundant than on Albor Tholus. numerous features suggestive of long, low with the largest crater about 30 km in Larger craters are totally absent. Concen- ridges which trend north and north- diameter. A small, 8-km crater is nested tric graben occur as far as 170 km from the northeast toward Hecates Tholus. The cra- within and indents the northern rim of the base. ter population on this plains unit is some- larger crater. The periphery of Albor At high resolution, Elysium Mons ap- what lower than on the surrounding sur- Tholus is marked by a circumferential gra- pears significantly different from other mar- face. The graben concentric to Elysium ben to the south and west and by an es- tian volcanoes (Fig. 5). No flow features Mons cut both plains units. carpment in the north. The eastern margin such as those on Olympus Mons can be dis- The age relationships between the vol- of the volcano is indistinct. Concentric gra- tinguished; rather, the surface appears canoes and their surroundings are not easily ben are seen to the west and south to a dis- hummocky and extremely subdued. Long, determined. In the very few cases where tance of at least 70 km. shallow, linear depressions, with a distinct high resolution photography was available, The surface of Albor Tholus was photo- trend radial from the caldera, dominate the evidence for plains both younger and older graphed only at low resolution. It has the visual appearance of the summit region. than the volcanoes was found within the lowest density of 4- to 10-km craters of the Some of the depressions are steep-walled, same frame. In most areas, no superposi- Elysium volcanoes. There are faint linea- channel-like forms, apparently the result of tion relationships could be discerned. It is

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/88/7/908/3429828/i0016-7606-88-7-908.pdf by guest on 28 September 2021 Figure 2B. Photomosaic of Elysium volcanic province, Mars. Mercator-projected photomosaic of Mariner 9 frames showing Elysium volcanic province. Images processed by the Image Processing Lab- oratory of the Jet Propulsion Laboratory. Note distinct boundary of Hecates Tholus and diffuse boundaries of the other two volcanoes.

Figure 2A. Geologic sketch map: Elysium volcanic province, Mars. Regional setting for Elysium others (1974) and from photogrammetry based on the method described by Blasius (1973). Consistent Mons, Hecates Tholus, and Albor Tholus. Note location of volcanoes on the crest of a broad, topo- with previous mappers (for example, Carr and others, 1973) and for the purposes of this simplified graphic high. Hecates Tholus, with a form reminiscent of a classic volcanic dome, is probably 8 to 10 sketch of the regional setting of the Elysium volcanic field two types of tectonically related depressions km above the 6.1-mbar mean martian pressure surface; Elysium Mons is about 20 km above that have been indicated: Large, broad, irregular depressions have been termed "depressions" and narrow, surface. General topography from Christensen (1975), with detailed topography from Hord and regularly linear or arcuate depressions were termed "graben."

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/88/7/908/3429828/i0016-7606-88-7-908.pdf by guest on 28 September 2021 Figure 3 A. Photomosaic of Mariner 9 high-resolution frames showing part of the Hecates Tholus (Mars) volcanic dome (left) and Elysium Plains (right). Illumination is from the left.

Figure 3B. Sketch map of Hecates Tholus (Mars) region. Large embayment of dome in lower center of map may be a secondary caldera or impact crater modified by erosional processes. Crater at far right is believed to be of impact origin because of its circularity, concentric rim terracing, and the radial texture of the surface exterior to the rim. Note differences between crater and central caldera at left center. Volcanic flows near the crater may have been formed from fissure eruption provoked by the impact event, but it is more likely that they were formed before the impact occurred and that they are covered by impact ejecta distinguished from the volcanic material by the radial texture. Slight textural differences in the plains surface at the volcano's edge may indicate late channel activity and deposition at channel mouths. The channeling is believed to be volcanic.

Figure 4. Elysium Mons: low resolution. Left: Mercator projection of Elysium Mons with topographic point network and contours. Contour interval is 2 km (resolution of photogrammetric data = ±1.5 km). Summit arbitrarily chosen as zero altitude. Data may contain residual overall tilt of no more than 2 km across entire frame. Sun from lower left. Right: near vertical view of Elysium Mons, with lighting from upper right. Note subdued ridged plain immediately north of volcano.

inferred that plains materials were probably suggested. At present, no way has been de- Caldera Features emplaced both before and after the massive vised to distinguish between the alterna- central volcanoes were formed. tives. Since Hecates Tholus has nearly twice The Elysium Mons caldera is approxi- Crater statistics are equally ambiguous. as many 4 to 10-km craters as Albor mately 14 km in diameter, with a rim height The large number of smaller craters (<10 Tholus, this may indicate the northern in the northwest of about 1 km, but it is km) on the volcanoes would indicate great dome was formed somewhat earlier than rimless in the south and southeast. The age (as old as the oldest, most heavily wa- the southern volcano (if the mechanism of south rim may have been removed by up- tered terrain on Mars) if formed by impact. crater formation was in any way a time- welling and overflow of material from However, many of the craters are clearly of dependent function). Alternatively, it may within the caldera, through burial by ash endogenic origin (for example, coalescing indicate substantially different forms of deposits, or by explosion. The caldera rim chains of craters), and the cumulative size- volcanism. Finally, all of the Elysium vol- is marked by two concentric escarpments, frequency distribution of the singly occur- canoes display more cratering than the old- separated by about 200 m, with the outer- ring craters is similar to the distribution for est Tharsis plains (Fig. 6) and than the most scarp at least 100 m higher than the craters of obvious volcanic origin, suggest- Tharsis volcanoes (Blasius, 1976). This innermost. The multiple escarpments may ing that many of the craters smaller than 1 suggests that the Elysium volcanoes may be be the result of two or more periods of km were probably not formed by impact older than the volcanoes of Tharsis. caldera collapse or of differential faulting (Fig. 6). On the other hand, the craters be- and terracing accompanying a single tween 4 and 10 km do not follow the same COMPARISONS OF episode of collapse. The Emi Koussi caldera statistical trend as the smaller craters. In EMI KOUSSI is very similar to the Elysium Mons caldera fact, they seem to follow the trend of the AND ELYSIUM MONS on Mars. The crater is about 12 to 15 km craters on the surrounding plains, but they across; it consists of two nested calderas have a normalized density greater than that Comparisons of Emi Koussi and Elysium with a maximum depth of about 500 m on the plains. If the volcanoes are much Mons can be made in two areas — the (Geze and others, 1959). The collapse of older than the plains, one must then explain summit calderas and on the flanks of the the inner caldera occurred some time after the paucity of larger craters (>20 km) on volcanoes (Fig. 7). In discussing the cal- the formation of the outer caldera. The rim the volcanic flanks. If the excess craters are deras, both the large-scale features of the of the Elysium Mons caldera is essentially not of impact origin, then a second popula- craters and the structures on their floors smoothly curving, but that of Emi Koussi is tion of large volcanic craters would seem will be compared. occasionally indented by channels and mass

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/88/7/908/3429828/i0016-7606-88-7-908.pdf by guest on 28 September 2021 914 M. C. MALIN

shallow (a morphology typically endogenic; for example, maars, or collapse craters), others appear to be bowl shaped and to have raised rims (characteristic of impact craters), and a very few appear to have slightly raised but extended rims and shallow interiors (a morphology suggested by McGetchin and others, 1974, to be the martian equivalent of terrestrial cinder cones). Emi Koussi also has numerous craters on its caldera's floor. The largest, called Era Kohor, is 3 km across and 350 m deep. It is accompanied by at least two and possibly more craters, and all are probably of phreatic origin (Geze and others, 1959). Crater morphologies are remarkably similar to some of those on Mars. Numerous basalt flows and cinder cones are visible on the floor of Emi Koussi caldera, but none are identifiable in Mariner photos of Elysium Mons.

Flank Features

The most distinctive, and probably most coincidental, features of similarity on the two volcanoes are the channel-like forms which emanate from the calderas. In both cases, a larger channel breaches only the outer caldera wall, while a second, smaller channel transects both walls. In addition, both larger channels head in roughly circular depressions. The smaller channels are somewhat more sinuous. All channels taper downslope. The terrestrial examples connect to the regional hydrological drainage system; the martian examples become a discontinuous series of elliptical depressions and eventually disappear. The Elysium Mons channels trend north, the same direction as the structural trend reflected in the ridged plains and in the po- sition of Hecates Tholus, relative to Elysium Mons. The channels on Emi Koussi follow the regional tectonic trend, and they have been flooded by fluid basalts from within the calderas. Owing to their alignment along a structural trend and to their distal termination in a chain of rimless craters, the Elysium channels are interpreted here as the result of tectonic fracturing and collapse of the volcanic flank associated with caldera collapse and subsequent volcanism, and possibly of collapse of lava tubes. Reports of field ob- servations (Geze and others, 1959; Vincent, 1960, 1970) do not discuss the origin of the channels on Emi Koussi. However, consid- Figure S. Elysium Mons: high resolution. Two-frame mosaic, sun from left. Caldera —15 km in ering (1) the difficulty of breaching the diameter, 1 km deep in northwest. Large (5-km) crater at southeast caldera rim is believed of impact caldera rim, (2) the field observation that origin. Channels connected to caldera are thought to have a volcano-tectonic origin. Large black line recent basalts floor and wall the larger at top of mosaic is a data drop. For support information, see Table 1. channel, (3) the alignment along the pre- movements precipitated by aqueous ero- are near the limit of resolution and cannot existing structural trend, and (4) the head sion. be drawn without some uncertainty. The depressions, an origin similar to that pro- The floor of the Elysium Mons caldera floor is also pocked by at least 21 craters posed for the channels on Elysium Mons contains numerous large, coalesced de- larger than 200 m. Size and resolution limi- does not seem unreasonable. pressions bounded by escarpments of subtle tations prevent unique categorization of Other channel-like features mark both relief. At least four levels (perhaps as many these craters as endogenic or exogenic. volcanoes. The lower flanks of Emi Koussi as six) are present, although the boundaries However, some craters appear rimless and are deeply dissected, evidence of a major

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/88/7/908/3429828/i0016-7606-88-7-908.pdf by guest on 28 September 2021 icr

ELYSIUM VOLCANIC PROVINCE PLAINS 5 2 • 92 craters > 2km In 4.6x10 km 42craters>.4km in2.1 xlO km 40 HECATES THOLUS I 4 2 v 9 craters > 2 km in 2.5x10 km ELYSIUM MÖNS 95craters>.4km in 8.5x10 km

OLDEST THARSIS PLAIN g 2 • TOTAL • 89craters>2kmin1.35xl0 km o SINGLY OCCURRING io- OLYMPUS MONS s 2 A CRATER CHAINS » 3 craters > 5kmin2£x 10 km 5 2 4-6 CRATERS >4km In 2.3xl04kmZ 1km <3 craters < 5kminE.8x10 km • TOTAL

* HECATES THOLUS a o E o ELYSIUM MONS JC a ALBOR THOLUS ID • ELYSIUM PLAINS TT O • THARSIS PLAINS 10

100 10'

10 0.1 1.0 10.

DIAMETER, D (km) D( km) Figure 6A. Crater counts on high-resolution photographs of Elysium Mons reveal extremely large 10 slope for craters smaller than 1.5 km diam, corresponding to craters of volcanic origin. Correlation 10 10 10* 10° between slope for larger craters (seen at both high and low resolution) and other cratered surfaces (compare Soderblom and others, 1974) suggests that they are of impact origin. Diameter, D (km) Figure 6B. Crater populations on the Elysium and Tharsis plains, on Hecates Tholus, and on Olympus Mons illustrate the range in relative ages of these volcanic landforms. The insert shows the normalized population of 4 to 10-km craters, which have been used to assign relative ages to surfaces over the entire planet (Soderblom and others, 1974). The error bars in both figures are statistical rep- resentations of the random nature of crater counts (square root of N for small N), and are slightly overestimated.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/88/7/908/3429828/i0016-7606-88-7-908.pdf by guest on 28 September 2021 916 M. C. MALIN

erosional period during the late Tertiary- early Quaternary (Geze and others, 1959). Numerous small channels reflect Pleis- tocene erosion (as many as five episodes are noted elsewhere in Tibesti), and some erosion is still taking place, at reduced rates. Only a few channels occur on Elysium Mons and these are invariably formed by chains of collapsed depressions. Greeley (1973) has interpreted these as partially collapsed lava tubes, which is entirely consistent with the available data. There appears to be no evidence of erosional channels of aqueous origin on Elysium Mons. The slopes of Elysium Mons are heavily cratered. The largest crater seen in high resolution is about 5 km in diameter and located just south of the Elysium Mons caldera (Fig. 5). It is essentially bowl shaped, but it has a small, flat floor. Low- sun shadow analysis places its depth at about 500 m. It is believed to be of impact LEGEND origin. However, as noted earlier, many ICMJTTGMRR other craters are probably of endogenic origin. This is especially true for craters which occur in chains distinct from those men- tioned above. These chains are formed co*rjtcr primarily of circular craters and extend east 5KM from the caldera. Greeley (1973) proposed that these features represented pit craters aligned over rift zones. The fact that the topography of Elysium Mons indicates a structural trend in that direction supports the hypothesis of rift zones. The larger, northwest-southeast ridge may also overlie a rift zone. Emi Koussi has a few craters on its slopes, all of endogenic origin. The most likely explanation for these craters is phreatic eruption during the late Quater- nary. The greatest distinction between Emi Koussi and Elysium Mons can be made on the basis of overall flank morphology. The terrestrial volcano has numerous flows and cinder cones marking its slopes. These are discernible in both Landsat and Apollo photography at all wavelengths and sun angles for which imaging data were available (although flows were more visible in IR photos, and cones in the film and low-sun frames). Emi Koussi has a form typical of a broad shield composed of numerous flows. Elysium Mons, on the other hand, has a surface comprised of many subtle, irregular mounds of low relief, giving the slopes a hummocky appearance. The largest mounds seen are about 5 km across, but most are significantly smaller. Flows as seen on Olympus Mons and the other Tharsis Figure 7A. Emi Koussi. Top: sketch map of Emi Koussi made from enlargements of Landsat RBV volcanoes are not visible, although many of and MSS and Apollo 7 photographs. Physical features from low-sun film product; geologic contacts the mounds appear to have short, stubby from Landsat MSS infrared and color-differenced data. Identification of rock types based on tentative appendages which point downslope. The correlation with field map of Geze and others (1959). Bottom: enlargement of Apollo 7 frame 5-1621 observed radial pattern of features resolves showing region mapped. Note deeply dissected flanks of volcano, craters, and cinder cones.

LEGEND to elongate hummocks and narrow valleys; higfmr some of these are simply the depression between two primary ridges, and others ap- ELYSIUM MONS pear to be unassociated with positive relief CALDERA features. Some of the mounds have subdued DEPRESSIONS craters at their summits; most do not. The Aititmrr depth Kquenc« entire surface appears subdued, with little detail at high resolution. The slopes of lOWMt Elysium Mons (~10° to 12°) are sig- eOn — InTerkor Cratsrt O — Ennnwl Craur» nificantly steeper than those of Emi Koussi @ — possible Cm tor (-2° to 5°) and Olympus Mons (-2° to 5°) Cofws S1»»p Escarpmem (Blasius, 1973; Wu and others, 1973). T * ' basi oftcarp > [Wiiil downslope Chonnvl CONCLUSIONS {Linn doïhed "'îic inferred J AND SPECULATIONS

Without ground-truth from the surface of Mars, any comparison between the origin and evolution of terrestrial and martian volcanoes is necessarily speculative. The following interpretations are presented as plausible possibilities, not as unique ex- planations. The following discussion will ) center on the nature and origin of the O ( Elysium volcanics and on the implications of surface features for past environmental 6 f conditions on Mars. EVIDENCE OF MAGMATIC SOURCE EVOLUTION

The Elysium volcanic province contains a wide variety of forms which are similar to those seen in Tibesti. Fissure eruption is likely the source of the plains units (both cratered and ridged plains) on Mars and of the plateaus in Tibesti on Earth. Central volcanism is present in both regions in the form of large, shield-like constructs and domes. Hecates Tholus has the classic form of a volcanic dome formed on a slope, although it is extremely large by terrestrial standards. The presence at its summit of knobby structures enhances the similarity to terrestial volcanoes with highly viscous extrusions. Elysium Mons is most likely a composite volcano. This conclusion is reached through consideration of three observa- tions. First, the highly subdued appearance of the flanks of the volcano suggests the presence of a blanketing material which conforms to, rather than obscures, the irregular mounds (probably flows of viscous lava). This blanket has a lower albedo than the blankets which cover much of Mars poleward of 40° latitude, and may be comprised of pyroclastic materials. Second, the Figure 7B. Elysium Mons. Top: sketch map of Elysium Mons made from enlargements of Mariner presence of numerous craters (other than 9 photography. No distinct units were mappable, but several distinguishable elevation levels within those associated with rifts) suggests explo- caldera may correlate to phases of activity. "Ringed" craters may have extended, raised rims and shal- sive ejection of magmatic materials, likely low interiors, suggestive of the form predicted by McGetchin and others (1974) for martian cinder cones. Bottom: enlargement of Mariner 9 frame DAS13496083 showing region mapped. Surface of to produce ash deposits. Finally, the steep- volcanic flank shows low, subdued mounds and radial pattern, possibly caused by volcanic ash man- ness of the average slope (~12°), nearly tling flows of viscous lava, and numerous craters and crater chains. twice that of any other martian volcano yet

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/88/7/908/3429828/i0016-7606-88-7-908.pdf by guest on 28 September 2021 918 M. C. MALIN

studied, indicates the presence of materials source). This interpretation supports previ- oldest Tharsis surface), then the last time other than fluid lavas. Whether comprised ous speculations that the martian crust has Mars could have had a pluvial erosional of viscous lavas, vicsous and fluid lavas, or not undergone plate-tectonic motion (Carr, environment, resembling that of Pleistocene viscous and fluid lavas with ash deposits, 1973). Tibesti, must have been an extremely long Elysium Mons is not simply a shield of the The age of the Elysium volcanoes is not time ago (compare Soderblom and others, Tharsis type. only crucial to the above discussion but also 1974). to one concerning climatic change. In this Sagan and others (1973) proposed that EVOLUTION OF ELYSIUM latter case, however, it is not the total age the surface of Mars is periodically subjected VOLCANIC PROVINCE but the age of the present surface which is to high atmospheric pressure environments important. As has been pointed out pre- (approaching 1 bar) because of "runaway" In both Elysium and Tibesti, volcanism is viously, cratering statistics are ambiguous. advective planetary heating. Significant associated with the doming of pre-existing However, by comparing the trends on the amounts of water are necessarily included materials. It is localized within a relatively Tharsis volcanoes (Blasius, 1976) with in such high-pressure atmospheres (to pro- small area and follows the tectonic trends those on Elysium Mons, it can be shown mote planetary heating), and the advective established during the uplift. Differences in that Elysium Mons has a significantly great- model is geared toward producing an envi- the 4- to 10-km-diam crater density on the er proportion of raised-rimmed, bowl- ronment conducive to abundant running three Elysium volcanoes may indicate either shaped craters (diameters between 200 and water. To initiate this condition, an in- a slight age difference or a difference in the 1,000 m), probably of impact origin. The crease in polar insolation of approximately type of volcanism, with Hecates Tholus the presence of many very subdued circular 15 percent is required. Ward (1974) has oldest and (or) most evolved and Albor depressions on Elysium Mons may indicate shown that astronomical perturbations Tholus the youngest and (or) least evolved. a previous surface, since blanketed and which can change the polar insolation by It is tempting to cite this trend as evidence recratered. Because the present crater pop- over 100 percent occur on a 105-yr time for the southward migration of the magma ulation is essentially fresh in appearance, it scale, and so runaway advection (if it oc- source. However, since Hecates Tholus is concluded that erosional and (or) blan- curs at all) should have occurred during the shows no sign of earlier, less silicic vol- keting activity on the Elysium volcanoes last hundred thousand years. The absence canism, the observed trend would seem to has been minimal for some period of time of aqueous erosional features on ancient indicate that the magmatic source region greater than that recorded by the oldest volcanic surfaces which should record such became less silicic as it evolved toward the surfaces of the Tharsis volcanoes. activity greatly limits the amount of water south. This seems to contradict the terres- and other volatiles that could have been re- trial sequence of magmatic evolution where CONSTRAINTS ON MARTIAN leased to the atmosphere. Murray and magmas from a single source become more ENVIRONMENTAL CHANGES Malin (1973) and Ward and others (1974) silicic with time (as they do, for the most have proposed that the maximum possible part, in the Tibesti province). Although As can be seen by examining the satel- atmospheric pressure is limited by the multiple magmatic sources is a possible lite photography (Fig. 7), the major mor- amount of volatiles stored at the poles alternative, the local grouping of the phologic difference between Emi Koussi (around 50 mbar). The evidence on Elysium Elysium volcanoes and the single, broad and Elysium Mons appears to be due to Mons is consistent with this or slightly dome on which they are built suggest a the presence of aqueous erosion on Earth. higher pressures but not with the environ- common origin. In lieu of unambiguous age Emi Koussi and the entire Tibesti region ment proposed by Sagan. relationships, and considering the lack of record and preserve evidence of an ero- any other evidence for the migration of the sional period which occurred during the ACKNOWLEDGMENTS volcanic center with time, it seems most early Pleistocene (although subsequent ero- prudent to suggest that the source of sion is also responsible for some of the fea- This paper represents the first of three magma was essentially stationary with re- tures seen; Geze and others, 1959). works comprising my doctoral dissertation spect to the surface. Additionally, the gen- Although this erosion may seem minor at the California Institute of Technology. I eral topography of Mars (Christensen, when compared to that which occurred am pleased to acknowledge the guidance, 1975) shows the Elysium uplift to be bor- elsewhere during the Pleistocene, this is support, and intellectual stimulation which dered by two regional lows to the east and most likely explained by Tibesti's environ- Bruce Murray and Robert Sharp have given west. The implications of basin and swell mental isolation deep within the African me throughout my years at Caltech. I am topography in Africa have been examined continent. However, even the "limited" grateful to reviewers John McCauley, Keith (Burke and Wilson, 1972; Wilson, 1973; erosion produced deeply incised canyons. Howard, and William Muehlberger for Burke and Whiteman, 1973) and related to Current conditions produce only ephemeral their helpful comments. This research was mantle plume or "hot spot" activity be- gullies and a few small streams. Elysium supported in part by National Aeronautics neath a stationary crustal plate. The ab- Mons shows no trace of deep valley inci- and Space Administration Grants NGR sence of crustal rifting (often found in areas sion, nor do either of the other two vol- 05-002-117 and NGR 05-002-305. interpreted as uplifts above plumes) in canoes. Either martian volcanic materials Tibesti is attributed by these authors to do not record aqueous erosion as well as REFERENCES CITED melting and incorporation of crustal mate- terrestrial volcanic materials, or aqueous rials above the plume. The suggestion erosion has not occurred since the last activ- Blasius, K., 1973, A study of martian topography made here is that the martian crust has been ity in Elysium. It seems unlikely that the by analytic photogrammetry: Jour. essentially stationary with respect to the rock types are so different as to erode on Geophys. Research, v. 78, p. 4411. 1976, The record of impact cratering on the Elysium magma source, and that the ab- one planet and not on the other. If the sur- sence of tectonic rifting is similarly the re- great volcanic shields of the Tharsis region faces of the Elysium volcanoes are as old as of Mars: Icarus, v. 29, p. 343. sult of crustal melting (which may also be the relative cratering data suggest (that is, Burke, K„ and Whiteman, A. J., 1973, Uplift, responsible for the evolution of the magma the youngest Elysium surface as old as the rifting and the breakup of Africa, in Impli-

cations of continental drift to the earth sci- sults from Mariner 9: Jour. Geophys. Re- matism and tectonics, Clifford, T. N., and ences, Tarling, D. H., and Runcorn, S. K., search, v. 78, p. 4009. Glass, I. G., eds.: Darien, Conn., Hafner eds.: London and New York, Academic Masursky, H., and others, 1972, Television re- Publ. Co., p. 301-319. Press, v. 2, chap. 7.2, p. 735-755. connaissance of Mars and its satellites: Sci- Ward, W. R., 1974, Climatic variations on Mars: Burke, K., and Wilson, J. T., 1972, Is the African ence, v. 175, p. 395. I. Astronomical theory of insolation: Jour. plate stationary?: Nature, v. 239, p. 387. McCauley, J., Carr, M., Cutts, J., Hartmann, Geophys. Research, v. 79, p. 3375. Carr, M., 1973, Volcanism on Mars: Jour. W., Masursky, H., Milton, D., Sharp, R., Ward, W. R., Murray, B. C., and Malin, M. C., Geophys. Research, v. 78, p. 4049. and Wilhelms, D., 1972, Preliminary 1974, Climatic variations on Mars: II. Carr, M., Masursky, H., and Saunders, R., 1973, Mariner 9 report on the geology of Mars: Evolution of carbon dioxide atmosphere A generalized geologic map of Mars: Jour. Icarus, v. 17, p. 289. and polar caps: Jour. Geophys. Research, v. Geophys. Research, v. 78, p. 4031. McGetchin, T. R., Settle, M., and Chouet, B. A., 79, p. 3387. Christensen, E. J., 1975, Martian topography de- 1974, Cinder cone growth modeled after Wilson, J. T., 1973, Mantle plumes and plate rived from occultation, radar, spectral, and Northeast Crater, Mount Etna, Sicily: Jour. motions: Tectonophysics, v. 19, p. 149. optical measurements: Jour. Geophys. Re- Geophys. Research, v. 79, p. 3257. Wu, S.S.C., Shafer, F. J., Makata, G. M., Jordan, search, v. 80, p. 2909. Murray, B. C., and Malin, M. C., 1973, Polar R., and Blasius, K. R., 1973, Photogram- Geze, B., Hudeley, H., Vincent, P. M., and Wa- volatiles on Mars: Theory vs. observation: metic evaluation of Mariner 9 photo- crenier, P., 1959, Les volcans du Tibesti Science, v. 182, p. 437. graphy: Jour. Geophys. Research, v. 78, p. (Sahara du Tchad): Bull. Volcanol., v. 22, Sagan, C., Toon, O., and Gierasch, P., 1973, 4405. p. 135. Climatic change on Mars: Science, v. 181, Greeley, R., 1973, Mariner 9 photographs of p. 1045. small volcanic structures on Mars: Geol- Soderblom, L. A., West, R. A., Herman, B. M„ ogy, v. 1, p. 175. Kreidler, T. J., and Condit, C. D., 1974, Hagedorn, H., 1971, Untersuchen über Re- Martian planetwide crater distributions: lieftypen arider Räume an Beispielen aus Implications for geologic history and sur- dem Tibesti-Gebirge und seiner Umgebung: face processes: Icarus, v. 22, p. 239. MANUSCRIPT RECEIVED BY THE SOCIETY SEP- Zeitschr. Geomorphologie, supp. 11, p. 1. Vincent, P. M., 1960, Les volcans tertiaires et TEMBER 10, 1975 Hord, C. W., Simmons, K. E., and McLaughlin, quaternaires du Tibesti occidental et central REVISED MANUSCRIPT RECEIVED AUGUST 26, L. K., 1974, Pressure-altitude mea- (Sahara du Tchd) [Ph.D. thesis]; Faculty of 1976 surements on Mars — An atlas of Mars Sciences, Univ. Paris, Ser. A, no. 3617, MANUSCRIPT ACCEPTED OCTOBER 26, 1976 local topography: Boulder, Colo., Lab. At- 307 p. CALIFORNIA INSTITUTE OF TECHNOLOGY, DIVI- mos. and Space Physics, Univ. Colorado. 1970, The evolution of the Tibesti volcanic SION OF GEOLOGICAL AND PLANETARY Masursky, H., 1973, An overview of geologic re- province, eastern Sahara, in African mag- SCIENCES CONTRIBUTION 2545

Printed in U.S.A.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/88/7/908/3429828/i0016-7606-88-7-908.pdf by guest on 28 September 2021