Lunar and Planetary Science XXXII (2001) 1509.pdf

TERRESTRIAL ANALOGS FOR MARTIAN VOLCANIC FEATURES SEEN IN MOC IMAGES. L. Keszthelyi and A. S. McEwen, Lunar and Planetary Laboratory, University of Arizona, Tucson, Arizona 85721

Introduction: Many of the Mars Orbital Camera lavas. For example, if basaltic lavas are erupted after images include volcanic features whose initially inter- significant cooling and crystallization, their rheology pretation is best made by examining terrestrial analogs. will mimic more evolved lavas [3]. But relatively Channel-fed aa, rubbly pahoehoe sheets, inflated pa- evolved compositions should be expected to cap the hoehoe, and lava tubes have now been observed and Martian shield volcanoes, just as evolved, alkalic lavas evidence for voluminous and widespread mafic pyro- cap the Hawaiian shield volcanoes [4]. Such lavas are clastics continues to mount. After these initial qualita- the natural products of a cooling mantle magma source tive interpretations are made, more quantitative mod- region or a crystallizing basaltic magma chamber. The eling can be effectively applied in future studies. presence of such lavas is not evidence for arc-type vol- Shield volcanoes: The five major shield volcanoes canism. are the most prominent volcanic features on Mars. While highly speculative at this point, the observa- Examination of ~200 MOC images of the Tharsis vol- tions to date suggest that the Martian shield volcanoes canoes, Olympus Mons, and Elysium Mons showed a may be dominantly basaltic but are capped by a layer terrain largely mantled by various types of cover. This of more evolved lavas. Moderate explosive activity is makes interpretation of the volcanic features challeng- usually associated with the eruption of such evolved ing. In fact, slightly over half the images showed lavas, but neither cinder cones nor spatter vents have dunes, craters, impact-generated regolith, and thick been imaged to date. This may be because such fea- mantling deposits that effectively covered the primary tures have been (1) missed in the released MOC im- volcanic morphologic features. However, about 30% ages, (2) covered or destroyed by secondary processes, of the images showed identifiable channelized lava (3) never formed because the Martian lavas are rela- flows with morphologies similar to aa. The flows tively volatile-poor, or (4) never formed because branch and overlap to produce an overall “feathery” Strombolian-type eruptions in the Martian environment texture. When these flows hit the flatter slopes of the would lead to much more extensive fragmentations and surrounding plains, they often transition into sheet dispersion of pyroclasts than on the Earth. The best flows that retain an aa-like morphology. hope for resolving this issue would be to target the Mt. Etna on the island of Sicily provides an excel- source regions of the most recent flows on these volca- lent terrestrial example of aa covered slopes with a noes. Since the freshest looking lava correlates well similar “feathery” morphology. Recent flows on other with the most radar bright surfaces [5], a relatively strato-volcanoes such as Arenal in Costa Rica, also small area needs to be searched. exhibit a similar morphology. However, those flows While the feathery slopes covered by channelized on are much better preserved than the typical Martian aa dominate the shield volcanoes, there are several shield volcano flows. During 2000 a search was con- noteworthy oddities. A few collapsed lava tubes can ducted for a better analog. Despite its diminutive size, be found in the images from Olympus Mons. While Crater Mountain, just south of Big Pine, CA in the similar pits are found on the other shield volcanoes, Owens Valley was most accessible example found, those pits appear to be related to tectonic graben. We with extensive eolian mantling and some mechanical speculate that the reason lava tubes have only been erosion. At this level of degradation, it is difficult to identified on Olympus Mons is the sparse sampling distinguish the margins of flows from channel levees, provided by MOC to date. challenging attempts to measure the parameters needed We note that the presence of lava tubes does not for quantitative modeling of the flows. require more fluid lavas or pahoehoe flows since tube- However, the types of terrestrial analogs that have fed basaltic-andesitic aa flows are common on Mt. been found make some interesting suggestions about Etna. Sinuous rilles, which may have formed as lava the type of activity that formed the Martian shield vol- tubes or large open channel flows, are seen on canoes. First, the compositions of the terrestrial exam- Ascraeus and Pavonis Mons. ples are basaltic-andesites and trachybasalts. Thus the Flood lavas: Extensive flood lavas cover much of overall morphology of the flows suggests that the the more recent surface of Mars. The best preserved shield volcanoes are likely to be composed of some- flood lavas are found in the Elysium and Amazonis what evolved rocks. This has been suggested much basins [6]. The age of some of these lavas might be earlier from Viking images [1,2]. There are ways in less than 10 Ma [7], but they have a wide range of which basaltic lavas can behave like more evolved ages. Erosional remnants of fine-layered sedimentary Lunar and Planetary Science XXXII (2001) 1509.pdf

MOC IMAGES AND TERRESTRIAL ANALOGS FOR MARTIAN VOLCANOES: L. P. Keszthelyi and A. S. McEwen

deposits associated with Medusae Fossae Formation on darker layers can be seen in the ash deposits from the top of the lavas in the southern Cerberus plains suggest 1790 Keanakakoi eruption of Kilauea Volcano. The that some of the lavas may be older than they appear. layers are the result of (a) different phases in the erup- These flood lavas are dominated by a platy-ridged tion, (b) secondary eolian sorting, and (c) variable in- texture suggesting both compressional folding of the duration by acidic syn- and post-eruptive precipitation. surface and rifting of plates. Flow modeling suggested This terrestrial example emphasizes the point that the that the lava compositions should be roughly basaltic, presence of extensive ash deposits is not an indicator of though compositions between komatiitic basalt and andesitic or more evolved compositions. basaltic andesite are possible [8]. The most explosive mafic eruptions on the Earth Rubbly Pahoehoe. Part of the 1783-1784 Laki flow are phreatomagmatic. However, on Mars, purely field in Iceland exhibits platy-ridged surface morphol- magmatic eruptions could plausibly disperse ash far ogy. Field examination of the Laki flow field in the enough to produce the extensive layers seen in the summer of 2000 showed that this lava was composed MOC images. This is because the lower atmospheric of “rubbly” pahoehoe. Rubbly pahoehoe is a lava type pressure enhances exsolution of dissolved magmatic transitional between pahoehoe and aa and is character- gasses and increases gas-driven fragmentation while ized by a brecciated flow top composed of broken the lower gravity and atmospheric drag allow particles pieces of pahoehoe rather than the spinose clinker of aa to fly farther. The volume of pyroclastic materials ex- flows [9,10]. Rubbly pahoehoe is thought to be em- pected from the eruption of the recent flood lavas placed in a manner different than either aa or pahoe- themselves should be sufficient to produce the Medu- hoe. Rubbly pahoehoe seems to involve the transport of sae Fossae Formation [8]. Using the gasses released by lava underneath a thick, insulating, but mobile and dis- the 1783-1784 Laki eruption as an analog, we compute rupted crust. This mode of emplacement is similar to that typical Martian flood lava is expected to release the "thick crust" open sheet flow model presented in the equivalent of several hundred cubic kilometers of [8] and bolsters the calculations suggesting that many extremely acidic aqueous fluid over a period of years. of the Martian flood lavas were emplaced with average This is a more conservative calculation than earlier effusion rates of ~104 m3/s with surges ~105 m3/s. estimates of volatile release associated with Martian Inflated Pahoehoe. One of the puzzles posed by flood basalt eruptions [13]. Combined with eolian the early MOC images was why we could find no evi- reworking, these volcanic products can readily explain dence for inflated pahoehoe flows on Mars, despite the the observed layered terrains. fact that the environment was conducive to the forma- Conclusions: Basaltic lavas are likely to dominate tion of inflated pahoehoe and the distinctive morphol- the plains of Mars and probably the deeper parts of the ogy of the flows even in 10 m/pixel images [8]. More major shield volcanoes. However, more evolved, recent MOC images do show several excellent exam- probably alkalic, compositions appear to cap the shield ples of inflated pahoehoe flows, particularly in western volcanoes. Circumstantial evidence also points to large Cerberus Planitia. volumes of mafic pyroclastic materials being reworked Layered terrain: Several different types of lay- to produce many of the "sedimentary" layered terrains. ered terrains have been found in the MOC images. References: [1] Hulme G (1976) Icarus, 27, 207- Some of these, particularly in the walls of the large 213. [2] Zimbelman J. R. (1985) JGR, 90, D157-D162. canyons, show largely continuous layers a few tens of [3] Gregg T. K. P. and Sakimoto S. E. H. (1996) LPS meters thick that can easily be interpreted as stacks of XXVII, 459-460. [4] Peterson D. W. and Moore R. B. sheet-like flood lavas [11]. Excellent terrestrial ana- (1987) USGS Prof. Pap. 1350, 149-189. [5] Harmon logs can be found in most terrestrial flood basalt prov- J. K. et al. (1999) JGR, 104, 14065-14089. [6] inces, but the Columbia River Basalts provide the most McEwen A. S. et al. (1999) LPS XXX, abstract #1829. recent and best preserved examples. In the case of the [7] Hartman W. K. and Berman D. C. (2000) JGR, scarps in the shield volcanoes, the layers are somewhat 105, 15011-15025. [8] Keszthelyi L. et al. (2000) more discontinuous, as would be expected since the JGR, 105, 15027-15049. [9] Keszthelyi L. and flows on the volcanoes are less laterally extensive than Thorvaldur Th. (2000) GSA Annu. Meeting, abstract the flood lavas on the plains. #52593 online. [10] Keszthelyi L. (2000) AGU Spring More controversial are the layered terrains within Meeting, abstract #V41A-09 online. [11] McEwen et some of the canyons, craters, and plains. While the al. (1999) Nature, 397, 584-586. [12] Malin M. C. and idea that they are water-lain sediments has been re- Edgett K. S. (2000) Science 290, 1927-1937. [13] Plescia J cently favored [12], such features can also be formed (1993) Icarus, 104, 20-32. as variably indurated ash deposits. On a scale much smaller than the Martian examples, fine brighter and