sign in sign up
<<
Home , Arsia Mons, Ascraeus Mons, Atmosphere of Mars, Basalt, Ceraunius Tholus, Elysium Mons

ri

VOLCANISM ON

JAMES R. ZIMBELMAN Center fer aiiä PUmttary Studies, Smithsonian Institutimt

I. Introduction Fossae Descriptor applied to an aligned of fractures I. Background in a . III. Large Central Volcanoes The intennediate geologic on Mars iden- IV. Paterae and Tholi tified through geologic mapping of superposition rela- V. Highland Paterae tions, and the areal of impact craters. VI, Small Constructs Icelandite Volcanic with silica content comparable to VII. Volcanic an andésite, but which did not form as a result of subduc- VIII Compositional Constraints tion or orogenic processes, IX. Volcanic History of Mars Mons Descriptor applied to a large isolated on X. Future Studies a planetary surface, The oldest of the geologic epochs on Mars iden- tified through geologic mapping of superposition rela- tions, and the areal density of impact craters. Patera Descriptor applied to an irregular or complex crater GLOSSARY with scalloped edges, surrounded by shallow flank slopes. pseudocrater A "rootless vent" created by the interaction The youngest of the geologic epochs on Mars of with or wet sediments beneath identified through geoiogic mapping of superposition the flow. relations, and the area! density of impact craters. shield A broad volcanic construct consisting of An irregular collapse feature formed over the evac- multitudes of lava flows. Flank slopes are typically •5°, uated chamber within a volcano. or less than half as steep as the flanks on a composite central volcano Emplacement of volcanic materials from a volcano, centralized source vent rather than along a distributed SNC .A group of meteorites thought to originate line of vents, on Mars, due to a relatively young and trapped composite volcano A volcano consisting of intermixed effu- like the of Mars. The acronym sive lava flows and pyroclastic materials. Flank slopes is derived from the names of the three meteorites that are tyjjically >10'', or more than twice as steep as the define the major subdivisions identified within the group: flanks of a . S, Shergotty; N, Nakhta; C, .

Copyright © 200C by .\cüdem!C Press Encycloptdia of Vüktmots 771 All rights of reprixliittion in any form restrveii. 772 ON MARS

Tholus Descriptor applied to an isolated domical smali body in the solar , but this activity is powered mountain or hill, usually with slopes much steeper than by unique gravitational tidal forces within the Jovian that of a patera. system; see "Volcanism on lo"). Next we review some volcanic plains Planar mappable units interpreted to consist important background information before examining of volcanic materials, usually with individual flow mar- the types of volcanic features present on Mars. gins resolvable within the unit. yardang A rounded erosional landforiu produced by wind- driven . Some yardang fields on Alars are interpreted to be formed in wind-eroded pyroclastic deposits. 11. BACKGROUND

The first direct evidence of the importance of came during the 9 orbital mission. A I. INTRODUCTION massive global dust stonn ob.scured the entire surface when arrived at Mars on November The advent of exploration revealed that Mars 14, 1971, but as the weeks passed, the dust slowly began possesses some of the most dramatic volcanic features to settle. The first surface features at equatorial found any\\'here in the . How did a to become visible through the dust pall were four dark half the size of the Earth generate volcanoes like Olym- "spots" in the region of the planet that telescopic ob- pus Mons, which is several times larger than the largest servers called . Gradually these spots were re- volcanoes on Earth? This question is just one example vealed to each be topped by a complex crater assemblage, of the issues currently being investigated as part of the the first clue that four huge volcanoes were present in relatively recent scientific endeavor called "Comparative the Tharsis area. By chance, the three previous fly-by Planetology." This article will summarize the basic in- Manner missions failed to image this area in detail suffi- formation presently known about volcanism on Mars. cient to show these large volcanoes. Mariner 9 continued The reader can then compare the Martian volcanoes to to map the entire until October 12, the numerous volcanoes studied in great detail here on 1972, revealingthatTharsis was not alone in containing the Earth as well as to volcanoes observed on other impressive volcanic constructs. The Mariner 9 view of planetary bodies. Mars was updated significantly with the improved reso- Two generalizations can be made before we investi- lution of the cameras on the two Viking Orbiter space- gate the Martian volcanoes in greater detail. Volcanoes craft that operated at Mars from June 19,1976, to August on Mars appear to be broadly similar in overall morphol- 17, 1980. The -55,000 Viking Orbiter images led to ogy {although different in scale) to volcanic features on refined interpretations for many Martian features, but Earth, so there is no direct evidence for Martian eruptive they did not significantly alter the first global perspective processes that are somehow significantly different from of Mars revealed by the Mariner 9 ¡mages. the volcanic styles and processes known on Earth. This The distribution of volcanoes on Mars is not uniform; inference strengthens our confidence in using the basic there are four regions where central volcanoes are con- approach of attempting to understand the Martian vol- centrated (Fig, 1). Twenty-three volcanic constructs canoes in terms of traditional . Next, the have been given names (Table I) by the International Martian volcanoes occur on of various relative Astronomical Union, the only organization that can of- ages (discussed below), so that volcanism represents a ficially designate names for planetary features. The dominant geologic process throughout much of Martian largest concentration of Martian volcanoes occurs in the history. This is in contrast to volcanism on the smaller Tharsis region, where the volcanoes are distributed on body of Earth's , where volcanic activity was con- and around a 40i)(}-km-diameter bulge in the Martian fined to roughly the first third of lunar history (see , centered on theet]uatorat 11Ü°W . This ""). Comparative planetologj' bulge has lifted the crust to an elevation •10 km above thus supports the concept that volcanism is the primary the Martian datum, twice as high as the highest portions mechanism for a planetary body getting rid of its internal of the cratered highlands that dominate the southern heat, where smaller bodies tend to lose their internal hemisphere of Mars (shaded portion of Fig, 1), The heat more rapidly than larger bodies ('s moon lo "Tharsis Bulge" is surmounted by three volcanoes col- appears to contradict this trend because this object, the lectively called the (5, 6, and 16 in size of Earth's Moon, is the most volcanically active Fig. 1), aligned along a N40E trend that includes addi- VOLCANISM ON MARS 773

FIGURE Í Location map for named central volcanoes on Mars, The black areas Indicate various volcanic centers, v^ith numbers keyed to Table I. The shaded area represents densely cratered materials typical of the highlands that dominate the southern hemisphere. Letters indicate landing sites (VI, Viking I ; V2, , P, ) and the (MFF). Mercator projection.

tional volcanic centers northeast of the Tharsis Montes. late in Martian history, hut ages for documented samples The Tharsis Montes and nearby (15 in collected from regionally expansive terrains are needed Fig, I) are the four tallest features on Mars, the summit to constrain the precise time range covered by the crater- of which were first imaged by Mariner 9 above ing record on Mars. In spite of this uncertainty in abso- the dusty lower atmosphere. Outside of the Tharsis area, lute age, the geologic stratigraphy provides a robust Martian central volcanoes are confined to the Elysium, framework in which the relative ages of various materi- Syrtis Major, and Hellas regions (Fig. 1), all named for als, including both central volcanoes and regional volca- broad bright or dark surface markings visible with Earth- nic plains, can be evaluated. based telescopes. In addition to the named central volca- Remote-sensing data provide information about Mar- noes, broad expanses of regional plains exist across Mars, tian materials that is complementary to the photogeo- many of which include compelling evidence that they logic interpretation of spacecraft images. Reflected were formed by the emplacement of numerous individ- visual and light provides compositional infor- ual volcanic flows. mation by the wavelengths at which some of the light The Mariner 9 and Viking images showed the physio- is selectively absorbed. Earth-based telescopic studies graphic makeup of the entire surfaceof Mars. Through can not resolve individual volcanoes but reflectance careful observation of the superposition and cross-cut- spectra indicate that bright regions are covered with a ting relationships of distinctive material units, a global dust that includes considerable oxidÍ7,ed whereas geologic sequence has been developed. The areal density the dark regions are less dusty and may include pyrox- of impact craters present on the various material units ene-bearing material such as like allows the relative stratigraphy to be established even . Thermal infrared measurements by the Viking where units are not in contact with each other. The orbiters indicate that the Martian surface is coated with convention defined for three major divisions of Martian particulate materials, ranging from micron-sized dust geologic history are the Noachian (oldest), Hesperian covering practically all of surface in the bright regions (intermediate age), and Amazonian (youngest) epochs, to sand-sized particles mixed with larger unresolved providing a broad temporal sequence in which we can blocks in the dark regions. Significantly, the Viking ther- discuss the various Martian volcanoes. Unfortunately, mal measurements showed no "hot spots" attributable the absolute age of the,se Martian epochs is very uncon- to internally generated heat anywhere on the planet. strained at present; various models of the impact flu.x Radar signals transmitted from Earth and reflected from on Mars allow the Ama7,onian period to be from millions the Martian surface have shown that some individual to billions of in age. Most planetary scientists likely volcanoes, such as the Tharsis Montes, display strong would accept that Amazonian materials represent events scattering behavior. The scattered radar signals indicate 774 VOLCANISM ON MARS

TABLE t Named Volcanic Centers on Mars

Horizontal scale' Rei/ef* Number Nome Center (M (km)

1 Alba Paiera AQ'N. I09°W 450 ~4 2 Albor I^N, 2I0°W 160 -5 3 Amp h i tri tes Patera 59°S, 299°W 130

"Approximate horizontal dimension (maximum and minimum values for elongate features). 'Estimated relief from summit to the surrounding ; from U.S.G.S. (I99Í). ' Largest of several small features along (Fig. 4).

considerable surface roughness at the 10 cm to meter scale on the surface of the Tharsis Montes, something III. LARGE CENTRAL VOLCANOES that may likely also apply to other volcanoes but which are too small to be clearly resolved with Earth-based Olympus Mons and the three Tharsis Montes each have radar. West of the Tharsis Montes, a region of several shapes tyjiical of a basaltic shield volcano on Earth, but million square kilometers returns no radar signal back at non-Earth-like scales (Fig, 2). The flanks of the four to Earth; this area has been nicknamed "Stealth" after have maximum slopes of •5", with shallower its very low radar reflectance properties. Stealth could summit and basal slopes, much like the shape of the be a deposit sufficiently thick so that the radar signal is subaerial portion of the shield volcano. The fully absorbed, such as through deposition of pyroclastic great bulk ot the Mauna Loa volcano, which with its materials associated with the nearby volcanoes, or subaqueous mass comprises the largest mountain on Stealth also might have unusual relief that efficiently Earth, is still dwarfed by any one of these four moun- scatters the radar signals away from the antennas on tains, but particularly by Olympus Mons with its 500-km Earth, diameter and 25-km vertical relief. The four Tharsis VoLCANiSM ON MARS 775

"Mons" refers to a large isolated mountain, and all four Tharsis shield volcanoes clearly deser\'e this designation. The lower flanks of the Tharsis Montes have been buried by ]ilains-forjning Hows emanating from the vol- canoes themselves or vents ¡n between them. Only Olympus Mons is surrounded by an arcuate scarp nearly fi(K) km in diameter and with 3-6 km of vertical relief i (Eig. 2). The origin of the l)asal scarp at Olympus Mons remains controversial, but several enonnoiis fan-shaped deposits northwest of the volcano have been interpreted * to be gravitationally driven landslides, perhaps aided by volatile lubrication along the detachment surface. High- resolution Viking images show multitudes of interwoven Él lava flows extending down the flanks of Olympus Mons and over the basal scarp, spilling onto the surrounding 'í(pK "*. '^^^ji^^^j^^^,. plains. Many of the intlividual flows have medial chan- m kiit^ nels similar to channel-fed lava flows on terrestrial volca- ^^ ÍMLüI bpír>^ '^ÄBi noes. Thermal infrared data indicate that all four of the •-^líisHi ^'^ volcanoes are thoroughly coated with fine windblown * dust, severely limiting any infonnation about the volca- FIGURE 2 Olympus Mons, the largest central volcano on nic rocks that can be learnetl from most remote-sens- Mars (15 In Fig. I and Table I). Olympus Mons has shallow fiank ing techniques. slopes typical of a shield volcano. The voicano has 25 km of relief As noted earlier, all four shield volcanoes are sur- from its summit to the plains surrounding the basal scarp, which mounted by summit calderas. Olympus and .\seraeus Is •600 km in diameter and from 3 to 6 knn in height. (Viking .Montes bave complex calderas revealing multiple epi- Orbiter image M6A28, courtesy of NASA.) sodes of collapse w^hereas Arsia and Pavonis Montes have only a single collapse rim preserved at present. Geophysical modeling of fractures around the caldera on shield volcanoes are so large that their size is not fully Olympus Mons verifies that these features are consistent appreciated without considering the ciin'ature of the with collapse into evacuated magma 'oirs within Red planet; an astronaut on the surface anywhere near the shield construct. By analogy to terrestrial shield these mountains could not see their summits because volcanoes, these intermediate magma chambers may this would be beyond his (or her) local horizon! have been periodically replenished from a larger magma The great size and height of the four large Tharsis source located beneath the volcanic pile. volcanoes generated considerable discussion about their A separate volcanic center with its own crustal bulge potential source regions. If bouyanc\' is assumed to ¡le {hut one considerably smaller than the Tharsis Bulge) the sole driving force for a typical basaltic magma on is present in the Elysium region (11 in Eig. I). Elysium Mars, the summits of the four large Montes volcanoes Mons is the largest construct in the region, and it has indicate their sources lie at depths >I2() km, potentially some important distinctions from the four Tharsis shield implyinga thick lithosphère in the Tharsis region. Inter- volcanoes. has considerably steeper flank estingly, older Martian volcanoes generally display re- slopes (up to 12°) than the Tharsis shields, surmounted duced relief (when compared to the Tharsis shields), so with one major caldera (although the floor preserves lithospheric thickness may have increased throughout subtle details that suggest repeated collapse prior to Martian history, VVlien combined with the lack of any the last lava infilling). The shape of Elysium Mons is obvious nearby plate tectonic features such as subthic- somewhat reminiscent of conical composite volcanoes tion arcs or transform faults, the great bulk of these typical of margins such as the or the mountains may result from sources that released , but there are no associated features {such as their magma at one location without the lithospheric a nearby trench) to indicate that subduction is in any movement we associate with plate on Earth. way responsible for this Martian volcano. Following In this light, it is interesting that the volume of Olympus Mariner 9, Elysium Mons was compared to the volcano Mons is roughly comparable to the integrated volume Tibesti in Africa, where an intracontinental hot spot is of the Hawaiian-Emperor . The descriptor a more likely analog than an acdve plate margin. VVTiile 776 VoLCANiSM ON MARS there are some vocal supporters of possible subduction than Alba Patera discussed above) are generally smaller on Mars, the consensus impression since Mariner 9 has than 20(t km in diameter (Table 11). been the apparent dearth, if not total absence, of any Paterae in the Tharsis region (7, 21, and 22 in Fig. 1) classic plate tectonic features on Mars. tend to have caldera diameters more than half as broad AJha Patera, north of the three Tharsis Montes (1 in as the overall construct. This observation led some re- Fig. 1), represents another important exception to the searchers to conclude that such paterae may he the sum- broad shield morphology. This volcano is only slightly mits of shield volcanoes whose lower flanks have been smaller in horizontal dimension than Olympus Mons, buried by more recent volcanic plains, which seems quite but it lacks the dramatic relief present on the Tharsis reasonable for che Tharsis region, where volcanic plains shields and Elysium Mons (Table I). Early results from are abundant and likely quite thick. Tholi in the Tharsis the Mars Orbital Laser Altimeter on the Mars Global region (8, 12, 19, and 23 in Fig. 1) tend to have slopes Surveyor spacecraft indicate that this volcano has flank steeper than tho.se of the shields but usually not as steep slopes ^ 1°, much less than the already shallow 5° flanks as Elysium Mons. Some volcanoes (e.g., Ceraunius 7"ho- of the shield volcanoes. Even more unique is a !)imd of lus) have sinuous channels carved into their flanks, im- dense fractures that makes almost a complete ring plying significant effusive (and erosive) flow after the around this shallow-sloped construct. Some researchers hulk of the construct formed (Fig. 3), A few researchers have compared this tectonic pattern ro that which woulit have interpreted the erosive channels on tholi and pa- surround a core of competent rock that is subjected to terae flanks to he due to pyroclastic flov^^s. The buried- intense regional ; such an interpretation would imply that Alba Patera was completely solidified well before the regional stress pattern was imposed. Crater indicate that .Mba Putera has been exposed to the impacting bombardment longer than either the Tharsis shields or Elysium Mons, but some volcanic effusion may have postdated the stress that produced the distinctive arcuate hand. Recently, it was proposed that Alba Patera may be a Martian e.vample of a Venusian corona {see "Volcanism on "); such an interpreta- tion has profound implications for crustaî thicknesses on Mars during an epoch when other comparable thin- features are not readily apparent. As we shall see next, Alba Patera is not representative of the majority of features that have been given the name "patera."

IV. PATERAE AND THOLI

Alba Patera is the largest member of a class of volcanic construct on Mars that is quite distinct from the Mons- type volcanoes. The descriptor "Patera" refers to an irregular or complex crater with scalloped edges that is surrounded by shallow flank slopes. These features contrast greatly with impact craters or basins of compa- FIGURE 3 Ceraunius Tholus, a volcanic dome with an elon- rable size, and a volcanic caldera is the consensus inter- gate crater from an oblique impact at its northern base (8 in Fig, pretation. Another distinctive class of volcanic construct, I and Table I). The volcano is noncircular, with dimensions I SO X which is consistently smaller than the Mons type, has 100 km. At least one sinuous channel carved into the volcano the descriptor "Tholus" applied to an isolated domical was active after the oblique impact since a riverlike "delta" formed small mountain or hill, usually with slopes much steeper in the crater where its rim was breached by the lava channel. than those of a patera. Both Paterae and Tholi (other (Viking Orbiter image 5I6A24, courtesy of NASA.) VOLCANISM ON MARS 777 shield interpretation for paterae has difficulty account- ing for the close proximity of both broad patera and steep thotus in the Tharsis region (S, 22, and 23 in Fig. 1), where the thoh volcanoes may indicate either effusive or compositional properties quite different from those of paterae volcanoes. The Elysium region has paterae and thoH as well, although fewer than in the Tharsis region {2, 4, and 9 in P'ig. 1). These volcanoes have both scoured flanks and, on Recates Thoius, a portion of the summit area where very few impact craters are preserved. Both char- acteristics have been interpreted to result from pyroclas- tic activity: erosive pyroclastic flows to generate the flank scours and an inferred ash deposit following a plinian- stj'le eruption at the summit of Hecates Thoius, Neither of these interpretations should he considered proof of pyroclastic activity, but they do suggest an increased possibility for a pyroclastic component associated with paterae- and tholi-style eruptions, as compared to Mons-style volcanoes.

FIGURE 4 Tyrrhena Patera, one of the low-profile paterae in the southern highlands near the rim of the Hellas impact basin (20 in Fig. I and Table I). The radiating pattern of channels has V. HIGHLAND PATERAE eroded into material •180 km in diameter, interpreted to be ash deposited around the volcanic vent, (Viking Orbiter image Paterae located in the cratered highlands of Mars de- 87AI4, courtesy of NASA.) serve special treatment because they appear to have sev- eral characteristics distinct from those of the small volca- noes in the Tharsis and Elysium regions. The highland way to build up a broad low-profile deposit of friable paterae are concentrated in two regions: a linear trend materials around a volcanic vent is through deposition ot lour volcanoes on the outer rim of the Hellas impact of pyroclastics. Currently, it is not clear if pyroclastic basin in the southern hemisphere (3, 10, 17, and 20 in flows or falls are the dominant agent for emplacement Fig. 1), and two vents relatively recently identified of the ash, and both processes appear to be viable candi- within the classical dark region of Syrtis Major in the dates. VVIiat is apparendy rare at these volcanoes is a northern hemisphere ( 13 and 14 in Fig. 1 ). Both regions significant effusive lava flow component. The top of include evidence that suggests that pyroclastic materials the friable deposit at each volcano appears to be more comprise an important fraction of the products pro- competent, forming a caprock between channels eroded duced by these highland paterae, suggesting that volca- into the deposit; this resistant zone may be case-hard- nism in the highlands may have involved eruptions fun- ened or indurated friable materials, or it may involve damentally different from those that produced the broad some effusive lava flows, hut this is difficult to resolve Mons volcanoes, the shallow paterae, or the steep tholi. with current data. All four Hellas volcanoes represent The four paterae on the rim of the Hellas basin are a substantially different style of eruption and emplace- notable in that all possess very shallow flank slopes, ment than the other central volcanoes discussed above, likely much less than the shallow 1° flanks of :AJha Patera. suggestive that the magma involved may have somehow The Hellas volcanoes are more noteworthy lor their been different (increased volatile concent?) from the ef- intensely eroded appearance than for any obvious con- fusive that dominate the more recent Martian vol- struct around the central vent (Fig. 4), The channels canoes. cut deeply into the flanks of the Hellas paterae have Nili and Meroe Paterae were only recognized as being been interpreted to indicate removal of friable material volcanic vents in the 1980s, long after Mariner 9 and concentrated around the central vent, possibly through the early Viking images established the presence and either fluvial or volcanic processes, or both. The logical types of the other volcanic centers. Both paterae are 778 VOLCANISM ON M.\RS near the center of the Syrtis Major dark region that has been observed telescopically from Earth for several centuries, Remote-stnsidií data of the Syrtis Major re- gion indicate the dark materials are significantly less oxidized than the bright dusty portions of the planet, with strong indications of a mafic chemistr)' that likely includes a component. V\Tien the volcanic vents were identified, researchers proposed that the dark material in Syrtis Major may in fact be wind-transported mafic . V^iking images in Syrtis Major revea! abundant dark dunelike features, supporting the concept that wind action in crucial to the story of this area. The action of the wind on the individual ash particles may help to remove an oxidation coating as the grains bounce into each other.

VI. SMALL CONSTRUCTS

Mars has numerous occurrences of small domes, some FIGURE S Small volcano (between arrows) located along of which may have a volcanic origin. Unfortunately, Tempe Fossae (18 in Fig. I and Table I). Solar iilum (nation from these features are mostly small enough that Viking im- the left indicatesan elongate mound built uparound a very elliptical aging data are not sufficient to provide definitive evi- depression, probably analogous to a along a dence for their origin. One of these features that is big linear vent. The feature is --3 km in length and likely <300 m in enough to show some important physical characteristics vertical relief. (Viking Orbiter image 627A26, courtesy of NASA) is found in the Tempe region, within a zone of fractures called Tempe Fossae (Fig. 5). Here an oblong low mound is present around an elongated summit depres- careful study of the extensive Viking database. There sion (between the arrows in Fig. 5). The depression are no names given to these features, nor is there any follows the trend of the surrounding fractures, so that obvious pattern to their occurrence throughout the structural control likely played a role in determining highlands. The dominant characteristic is deep scours where vented materials could reach the surface. The va.st present on all of their flanks, similar to the scoured majority of these features are smaller than the Tempe exteriors of some paterae that are thought to have been example; typically, the individual domes are <1 km in subjected to pyroclastic flows. Assuming they are volca- diameter and at most a few hundred meters in height, nic in origin, they appear to represent yet another dis- A popular inteqiretation of their origin is that of tinct Style of volcanic deposit confined to the old densely cones, with the variation that some may he pseiidocraters cratered highlands of Mars. The eroded domical hills where a lava flow interacted with either a shallow cannot be dated by crater density due to their small layer or wet sediments. There are literally thousands of areal extent, hut their presence solely within the ancient these features in various places throughout the northern materials of the cratered highlands most research- lowlands of (Fig. 1), so that either a ers to assume an old age for these features. or pseudocrater origin would imply abun- dant water-lava interaction across the northern plains. An alternative nonvolcanic interpretation is a periglacial feature such as a pingo (an ice-cored mound). VII. VOLCANIC PLAINS The final category of distinct volcanic on Mars consistsof about tu'odoz.en deeply dissected coni- cal hills scattered throughout the highlands. These fea- Volcanic plains comprise some of the most extensive tures typically are <30 km in diameter and were only geologic units on Mars outside of the cratered highlands. recognized as potentially of volcanic origin through The plains tend to be quite featureless at moderate to VOLCANISM ON MARS 779 low resolution, but in high-resolution images, these The volcanic plains of Mars display flow features that plains display numerous loba te margins surrounding in- can he divided into at least two broad categories; simple dividual packages of material that appear to have been and complex. This distinction follows from an extension emplaced through flow of fluid materials (Fig, 6). The of George Walker's classification of terrestrial lava flows lobate margins are sufficiently thick (various measure- by the cooling units preserved in the volcanic sequence, ments indicate lobe thicknesses ot tens to hundreads of with simple flows representing a single cooling unit of meters) that the flowing material is widely accepted to great extent (and probable large volume of effusion) and be lava rather than something like a water-rich complex flows representing an intermixed sequence of flow. Volcanic plains surround both the Tharsis and discrete flow lobes that each comprise individual cooling Elysium volcanic centers, and a stacked sequence of units. Some plains units are nearly featureless in terms such flows likely accounts for a significant part of the of detail visible in even the highest resolution orbital topographic bulges surrounding these centers. Volcanic images, and many of these plains units may eventually plains also comprise regionally extensive units in loca- turn out to he volcanic in origin, even though at present tions that presently do not have visible volcanic con- they are not readily placed within the simple or complex structs. In particular, the plains of the Lunae Planum designation. Many of the volcanic plains that fill the area east of the Tharsis region and the floors of impact basins (like Hellas) or that lack clearly plains northeast of the Hellas basin extend over many resolvable individual flows over wide areas are consid- thousands of kilometers, and both of these plains have ered to be simple in the sense that the volcanic emplace- crater densities that indicate a Hesperian age that must ment most likely was sufficiently rapid or prolonged to be older than the volcanic plains around Tharsis and allow the plains to be considered as representing one Elysium. Thus substantial outpourings of plains-form- major cooling unit. The younger volcanic plains around ing lava flows, likely firom vents that are themselves Tharsis and Elysium are primarily complex (as in Fig. 6), buried beneath the thick pile of flows, were significant with numerous finger-like fltnv lobes traceable in some events during both the Hesperian and .^.mazonian cases for many hundreds of kilometers. As orbital ¡mage epochs. resolution improves, as with the on the spacecraft, some of the older volcanic plains likely may turn out to be made up of a series of complex flow lobes like the ones that surround Tharsis and Elysium. Not surprisingly, some plains units are problematic in origin. These units tend either to he relatively feature- less or to have such complicated surface exposures that their origin remains the subject nf considerable contro- versy. The Medusae Fossae Formation southw'est of the Tharsis region (MFF in Fig, 1) deserves a brief discus- sion here because although its origin remains controver- sial, one of the contending hypotheses is that this mate- rial is the result of massive pyroclastic eruptions. The Medusae Fossae Formation consists of several discrete layers nf material that is very friable (eroded by the wind into huge fields of yardangs) but locally prominent eaprock units help presei^'e the underlying materials. This observation has been compared to the welded and nonwelded zones present within many de- posits on Earth. .\ significant problem w^ith this hypothe- sis is that there is no indication of the po.ssible source vent or vents, which is troubling if volcanic eruptions laid down ash that in places is over 2 km thick and that extends over several million square kilometers along the FIGURE 6 Lava flow margins on the Elysium plains near Martian equator, following the boundar\' bet^veen the Mecates Tholus. Image width •54 km, centered on 32.2°N, southern cratered highlands and the northern lowland 2I3.5°W, (Viking Orbiter image 65IAI0, courtesy of NASA.) plains. The prospect of large ash - prod u ci ngerup dons in 780 VOLCANISM ON MARS

this part of the planet is Strengthened by recent remote- The first concrete compositional information ob- sensing observations that an area west of and tained for Martian materials came from an X-ray fluo- east ofthe majority of the Medusae Fossae Formation rescence instrument that was flown on both the Viking displays the unique property that Earth-based radar sig- 1 and Viking 2 landers. In spite of the fact that the nals are not reflected from this broad area. "Steahh" landers set down over 6000 km apart, the composition has been interpreted to be associated with pyroclastic ofthe Martian fines at both sites was remarkably uniform eruptions from vents west of the Arsia Mons volcano, (Table 11), which in itself was a strong indication of which lends some support to possible major ignimbnte the homogenizing influence of the regular global dust eruptions in the area. storms. Although repeated attempts were made to obtain a rock fragment with the Viking sample arm, no Martian rock was analyzed by the Viking instruments (centime- ter-sized fragments all turned out to be indurated clods of dust). This situation changed dramatically in July of VIM. COMPOSITIONAL CONSTRAINTS 1997 when the Mars Pathfinder mission successfijlly landed near the mouth of the .^res Vallis canyon and Compositional information for Mars is available from deployed the first mobile vehicle on Mars, the Sojoumer three different sources: remote-sensing data {from either rover. Sojoumer had a spectrometer that the rover was spacecraft instruments or Earth-based telescopes), on- able to place directly on several Martian rocks, as well site analyses by instruments on spacecraft that have re- as on the fine-grained materials between the rocks. The turned data from three landing sites on Mars, and a fines at the Pathfinder site were generally quite consis- special group of meteorites for which there is compelling tent with the fines observed at the Viking sites (Table evidence that these rocks were blasted from the surface II; note the differing sums for columns 2 and 3), but of Mars and eventually fell co Earth. Each of these data the rocks show distinct chemical differences from the sources has unique advantages and disadvantages as well fines ("Yogi" and "Barnacle Bill" in Table II), Images as various limitations to the utility of the data. Here we ofthe rocks at the Pathfinder site give a strong impres- will explore each data type for its implications for volca- sion of being volcanic in origin, with ubiquitous pits nic materials on Mars, that are likely vesicles. Assuming the rocks are volcanic, Remote-sensing studies of Mars began with telescopic the compositions obtained from the Sojoumer spec- measurements of various properties ofthe Martian sur- trometer imply they may be basaltic andésite. If the face and atmosphere, and since the 1960s these data anomalously large content for the rocks is as- have been augmented with higher spatial resolution data sumed to be contamination by dust on the rock surface, from a host of instruments on spacecraft that either flew the measured silica contents become lower limits and by or orbited the planet. Earth-based instruments can at least some of the rocks may turn out to be andésites achieve very high spectral resolution but only with rela- (e.g., "Barnacle Bill"), although recently the term Ice- tively low spatial resolution on the planet (covering an landite has been applied to these silica-enhanced rocks in area that may be hundreds of kilometers across). Space- an attempt to eliminate the subduction-re la ted processes craft instruments tend to achieve substantially improved usually associated with the term andésite. Indeed, there spatial resolution, but often this has come at the expense is no evidence in or near the Pathfinder site for orogenic of spectral resolution. The ubiquitous Martian dust generation of andésite lavas, but fractional iza tion complicates this effort at any spatial resolution, masking within isolated pockets of basaltic or basaltic andésite significant portions ofthe surface within the instrument could have generated some Icelandite-like field of view and generally decreasing the contrast of rocks. The Pathfinder results are undergoing ftirther the already subtle spectral features being sought. In spite refinement in association with the instrument calibra- of this handicap, remote-sensing data have shown that tion, so these results may be subject to revision at a the dark regions ofthe planet have strong mafic affini- later date. ties. Unfortunately, only the Syrtis Major volcanoes are The very best chemical information about Martian present within a dark region, and the spectral informa- materials comes from 12 very special meteorites. These tion of other volcanic centers is effectively masked (at meteorites are considered to be from Mars because gases visual and thermal wavelengths) by the dust. Tens of implanted in them during the shock of their ejection microns of dust can strongly affect visual reflectance, from their parent body is essentially identical to the and only 2 cm of dust can totally obscure even competent as measured by the Viking landers from thermal infrared observations. and unlike gases obtained from any other terrestrial or VOLCANISM ON MARS 781

TABLE 11 Chemical Compositions of Selected Martian Materials

Wiking ; IViking 2 Pathfinder Pathfinder Paàifinder Shergotty fines' fines' •7ogi"= "ßomocie ßi7i"' meterorite''

SiOi 43.3 49.0 55.5 58.6 51.4 AliO, 7.1 a4 9.1 lOM 7.1 FeO' 17.4 m ]3A 12.9 19.4 MgO 6.0 7* &9 3.0 n CaO 5.7 u &

° Average of two samples of fines from Chryse (Viking I ) and (our Fines samples from Utopia (Viking 2); from Table 6.1 of Frankel (1996). 'Average of three fines samples (A2, A4, and A5) from Pathfinder from Table I of Rieder et al. (1997), Note that results are reported as normalized to 98.0%. 'Results of two rocks (A7, Yogi; A3, Barnacle Bill) from Pathfinder from Table I of Rieder et al. (1997). Results normalized to 98.0%. 'Martian meteorice Shergocty, a basaltic ; from Table V. p. 605, in Kieffer el al. (1992). 'Total iron, regardless of oxidation state. Reported as FfijOj for column 2 and FeO for columns 3-6.

extraterrestrial sample. Eleven are igneous in and Mars, portions of which must date from practically the are only one-third the age (1,5 Ga to 180 Ma) of practi- formation of the planet. In spite of this wealth of chemi- cally all other meteorites, and the tft'elfth has an older cal detail, it will require the return of documented sam- age more typical of most meteorites. Collectively called ples from known localities on Mars before we can relate SNC meteorites, seven are or Iherzolites/harz- the rock histories to regional or local geologic materials. burgites like the Shergottj' (S) , three are clino- or wehrlites like Nakhla (N), Chassigny (C) is the sole , and ALH84001 is a --4.5-Ga orthopy- roxenice with veins of carbonate in which are controver- IX. VOLCANIC HISTORY OF MARS sial features that may (or may not) be evidence of primi- tive life from early in Martian history. The SNC meteorites have told us a lot about Mars The information presented in this article can be summa- in general, but we unfortunately have no evidence for rized into a general picture of the volcanic history of where on the Martian surface they came from. The Mars. Without constraints on the cratering record for abundance of volcanic rocks among the available group Mars, absolute ages cannot be associated with any of is consistent with the extensive coverage of Martian these stages. However, the stratigraphy of geologic units volcanic plains discussed above, but the young cr\'stalli- at least gives us a fairly clear picture of how volcanic zation ages cannot be u.sed to calibrate the geologic eruptions developed throughout Martian history. epochs on Mars without knowledge of what .specific unit The oldest terrain on Mars is the Noachian densely they came from. Some of the nonbasaltic rocks may cratered highlands, and if ALH84001 ¡s representative have come either from slowly cooled cores of thick lava of this , at least some magmatic activity took place flows where limited fractionation may have taken place during this time of intense cratering. Isolated volcanic or possibly even from a plutonic body. ALI 184001 likely centers developed within the cratered highlands, most came from somewhere in the cratered highlands of of which are preserved today only as deeply scoured 782 VOLCANISM ON MARS sniüll hüls scattered througthout the southern highlands. have an important influence on issues such as the volatile The Hesperiiin epoch witnessed massive eruptions of content of Martian magmas. The absolute age of all fluid lava that produced voicanic plains that covered Martian terrains, volcanic as well as sedimentary or extensive portions of the Martian surface. The composi- metamorphic in origin, remains a glaring uncertainty tion of these plains-form i nti materials was likely basaltic in all studies of the emplacement of volcanic materials on ill nature, although Pathfinder results suggest that more Alars, The return of documented samples from diverse evolved lavas {such as basaltic andésite) also may have localities on Mars would be the most definitive way been present. Several volcanic centers developed on the to answer these and other outstanding questions about rim of the Hellas impact basin and within what is now Martian volcanism, and fortunately NASA is actively the Syrtis Major dark region. These volcanoes involved planning toward that end. The Mars Surveyor program the generation of considerable amounts of ash, leading is an ambitious series of landers and orbiters targeted at to a very low overall protile and intense channelization addressing a host of science questions through repeated by either fluvial or magiuatic fluids. The unique Alba visits to Mars, culminating with the return of samples Patera feature ma)' represent a volcano caught in the cached by mobile rovers. In preparation for the transition from the more ash-rich eruptions of the high- missions, sophisticated remo te-sensing data will be col- lands to the more lava-rich eruptions tyjiical of most lected from new technology sensors in orbit around subsequent eruptions. Mars, as evidenced by the dramatic new data currently The late-llesperian and Amazonian epochs involved being returned by the Mars Global Surveyor spacecraft. voluminous eruptions that resulted in both broad volca- All of these new data will undoubtedly add considerable nic plains and numerous central volcanic constructs. detail to the volcanic history of Mars outlined in this ar- Volcanic activity concentrated around the Klysium and ticle. Tharsis regions; crater densities suggest that efïusive activity was most prolonged in the Tharsis region, where some volcanic surfaces are very sparsely cratered and thus may be tpiite young. Thousands of small domes in ACKNOWLEDGMENTS the northern lowlanti plains may be the result of either localized Strombolian activity' (cinder cones) or the in- teraction of lava with water or wet sediments (pseu- Reviewer comments and the advice of Editor H. Sigurdsson docraters). Not all of the volcanic activity in the Amazo- were ver)' helpful in revising the manuscript. This work was nian period may have been gentle efftision of lava; supporccU by NASA Grant NAG5-4IÚ4. significant ash deposits may be present west of the Tharsis Montes, although the true nature of these enig- matic materials remains controversial. Volcanic materi- als are likely still being redistributed by current winds, SEE ALSO THE FOLLOWING ARTICLES such as in the Syrtis Major dark region. Bawlric Volcanoes and Volcanic Systems • Calderas • Ignimlmtes and Block-and-Ash Flow Deposits • Lava Flows and Flow Fields • Pyroclast Transport and Deposition • X. FUTURE STUDIES Volcanism on lo • Volcanism oti the Moon •

Several issues about Martian volcanism remain unre- solved at present, and hopefully the series of missions NASA has planne

L. A. McFadden, and T, V, Johnson, eds.). Academic Press, Rieder, R,, Economou, T., Wanke, H., Turkevich, A,, Crisp, San Diego, J., Brückner,J., Dreibus, G-, and .\1cSween, II. Y•Jr. (1997). Francis, P. (1994). Pages 416-425 iii "Volcanoes: A P!anL't;m- The checnical composition of Martian and rocks re- P(;rs|>ective." Oxford Univ. Press, New York. turned by the mobile .Alpha Protun X-ray Spectrometer: Frankel, C. (19%). Pages 100-138 in "\'o!canoes of the Solar Preliininarj' results from the X-ray mode. Science 278, System." Cambridge Univ. Press, New York. 1771-1776 (Mars Pathfinder special issue). , R., and Spudis, P. D. (19fil), \'olcanism on Alars. Scott, D. H., Tananka, K. L., Creeley, R,, and Guest, J. E. Rci: Gtophys. Space Phys. 19, 13-4!. (1986). (îeologic maps of the western equatorial, eastern Kieffer, H. 11., Jakosky, B. .\!., Snyder, C W., and Matthews, equatorial and polar regions of Mars, Miscellaneous Investi- M. S. (1992). Pages 424-452 and 601-611 in ''Mars. '' Univ. gations Series .Maps [-1802-A, B, and C. U.S. Geological of ArÍKona Press, Tucson, Survey, scale 1:15,000,000. McSween, H. Y., Jr., (19y4). WTiat we have ¡earned about U.S. Cieologieal Survey (1991). Topographic maps of the po- Mars froni SNC meteorites, Mek'nritiis 29, 757-779. lar, western, and eastern regions of Mars. Miscellaneous Miirsky, Cj.(I996). Pages 219-243 /«"Introduction to Plane- Investig-ations Series Map 1-2160, U.S. Geological Survey, tary, Vokanism." Prentice-Hall, New York, scale 1: 15,000,000.