P23B-0187
Stability and evolution of ice cored glacial deposits on the flanks of volcanoes
There are several units on Mars which have been interpreted as glacial deposits based on morphological evidences. Many of these units are found on the flanks of volcanic edifices. Especially the recent imaging by the HRSC camera on Mars Express has provided us with a wealth of new examples and with much more morphological details. These include deposits as young as 5-14 Ma for example on the flanks of Hecates Tholus. We have recently studied this deposit and found a high likelihood that it is still ice cored. Based on our cautious estimates this ice would be protected by an only 10-50m thick sublimation till. We will report more detailed on these findings and compare it with comparable deposits at the flanks of other volcanoes. This deposit is especially interesting because there are indications for very recent activity of Hecates Tholus, including lava flows potentially as young as 2Ma. Assuming that we are today in a quieter but still active phase of the volcano we will discuss the implications for the internal structure and especially the redistribution of ice in such a deposit assuming episodic changes in the internal heat flow. There are indications for the formation of ice lenses potentially sealing environments at the base of this and similar deposits on other volcanic edifices. There is a wide variety of implications ranging from storage and delayed release of volcanic gases (incl. methane) up to potentially forming niches for biological activities.
P23B-0188
Tharsis Montes Mars: Evidence of Massive Recent Erosion Caused by Volcano-Ice Interactions
The Tharsis Montes volcanoes, Arsia, Pavonis and Ascreus, share many volcano-tectonic features, including fissure swarms that extend NE and SW away from their centers. Lavas and sedimentary deposits have built up fan shaped aprons that overlie or partly bury the fissures swarms to a distance of over 500 km from the volcano massifs. The pattern with depressions observed on the NE and SW sides of the Tharsis Montes suggests a common origin. The fissure swarms coincide with large-scale negative landforms, some 25 km wide and over 2 km deep. These negative landforms can further be divided into lineated channels and isolated circular depressions or pit craters. While tectonic movements have surely been a major factor in forming the depressions the overall character of this network resembles erosional channels. Volcanism culminated most likely within the fissures swarms where surface lavas are visibly draped by a cover of sedimentary origin. Teardrop shaped crater remnants with erosional collar grooves are found in these deep depressions, e.g. Pavonis Mons. The derived teardrop orientation is downstream within the depressions and oblique to local wind directional pattern. The volcano-tectonic pattern, erosional morphology and pit craters all conform to an origin where volcanism has produced meltwater that in turn has led to jokulhlaups, both within the Montes calderas, the fissure swarms and even on the flank volcano slopes. Research has shown that the age of volcano units, based on crater counting, is lowest on the fissure swarm floors. Volcano-ice interactions within the calderas are likely to have forced copious quantities of melt water into the fissures swarms and caused large-scale erosion there. A genetic relationship is suggested between the erosional valley systems and the locus of maximum volcanism within the fissures swarms. The process of jokulhlaups resulting from magma-ice interactions can account for the generation of this erosional pattern. Large scale melting of ice within Tharsis calderas and leakage of meltwater from there through fissure swarms can account for the erosional phenomena observed within the NE-SW Montes depressions.
P23B-0189
A Protracted, Complex, Period of Valley Formation on Hecates Tholus Volcano, Mars
Analysis of THEMIS VIS, HRSC and MOC images of Hecates Tholus volcano, Mars, reveals an extended time period over which the volcano's extensive valley network must have evolved, and hence constrains the origin and longevity of the hydrogeologic conditions under which the valleys formed. These image data, along with MOLA topographic data, permit the following valley attributes to be identified: (1) Some valleys on the southern flank of the volcano pre-date the emplacement of the lava flows from Elysium Mons, while other valleys post-date this volcanism; (2) Braided valleys on the E. flank formed on the shallower slopes, while deeper valleys formed on steeper slopes; (3) Valleys that flooded impact craters by breaching their rims did not create any deltaic deposits within the craters, indicating low sediment loads for the water; (4) Valleys are most well developed on the NW flank of the volcano; (5) Water discharge took place within the floor of at least one impact crater, while the rim of this crater was not eroded, indicating a ground water source; (6) Valleys originated within a few kilometers of the caldera rim; and (7) No valleys formed within the summit caldera. No source for the water that carved the valleys has yet been identified, but snow melt [Fassett and Head, 2004; Lunar Planet. Sci. XXXV, no. 1113] or the remobilization of volatiles released from the degassing volcano [Scott and Wilson, 1999; JGR 104, 27,079 - 27,089] appear to be the most likely. In either model, a hydrothermal system such as the one proposed by Gulick [1998; JGR 103, 19,365 - 19,389] could have remobilized the volatiles and carved the valleys. Explosive volcanism is, therefore, still a possible style of activity for Hecates Tholus (Mouginis-Mark et al., 1982; JGR 87, 9890 - 9904), but alternative explanations for the non-Uniform distribution of sub-kilometer diameter impact craters on the volcano, including glacial erosion or the rotting of surface lava flows by an active hydrothermal system, need to be investigated.
P23B-0190
Volcanism and Fluvio-Glacial Processes on the Interior Layered Deposits of Valles Marineris, Mars?
The Interior Layered Deposits (ILDs) in Valles Marineris have been suggested to be possible sub-ice volcanoes. Recent images also show evidence of possible fluvio-glacial processes on the ILDs and hence volcano/ice/water interaction. For example, Mars Express Mission anaglyph from Orbit 334 of central Ophir and Candor Chasmata, THEMIS image V10551002, and MOC images E1700142 and E190020 show 2 ILD mounds in central Candor Chasma that have been sheared off at approximately equal elevations by some material that has been subsequently removed. Level shearing of ILD rock materials and subsequent removal of the abrasive material, suggest ice erosion and glacial processes because glacial ice is mobile enough to grind the rock and can melt away. Another adjacent ILD mound in Central Candor shows an abrupt flank termination and damming of material, rather than flank scour. The dammed material appears to be layers piled up in a ridge at the ILD base. This relation is observed on the HRSC anaglyph and MOC images E0101343 and E201146. Another ILD in Melas Chasma, seen on MOC image M0804981, shows lobes of flank material that terminate along a lineation; possibly suggesting lobe confinement against subsequently removed material. This morphology can also be observed on the flank of the Gangis Chasma ILD in MOC image M0705587. A possible terrestrial volcanic analog for this ILD flank morphology is the Helgafell hyaloclasitic ridge (tindar) in Iceland (Chapman et al., 2004), the eastern flank of which has a linear termination interpreted as largely unmodified and caused by hyalotuff material banked against a former ice wall that has since melted away (Schopka et al., 2003). Glacial shearing of some ILDs and confined banking of other ILDs suggest that these mounds formed at different times, as the sheared ILD likely predated ice and the confined ILD may have formed concurrently with ice. Alternatively, the banking may have been due to lack of shear forces (static ice) and confined post-depositional avalanche deposits. However, exposed in the banked cliff faces are near horizontal bedding planes that can be traced upslope into angled flank layers; a relation that may suggest ice concurrent with volcanic ILD formation (Chapman and Smellie, in press). In addition to glacial processes, many Mars ILDs show fluvial gullies cut into mostly low lying flank deposits. Gullies are eroded into all sides of the ILDs including their north-facing slopes, so solar heating likely did not generate the gullies. Although formal work on the subject is lacking, ongoing terrestrial observation by the author (on an edifice north of Helgafell and in Gj lp eruption films) indicate fluvial erosion of subglacial volcanoes on Earth may be concurrent with their formation, occurring after edifices rise above their surrounding ice-confined meltwater lake. Remnant ice on the top of the edifices can melt to generate streams that erode the growing volcanic flanks.
P23B-0191 INVITED
Interrelated Glacial, Volcanic and Hydrologic Processes on the Tharsis Montes, Olympus Mons and Hecates Tholus, Mars
Volcanic, hydrological and glacial processes are prominent in the geological history of Mars and assessment of areas where their relationships can be established provides important information on their nature and intensity. An effort to examine these interrelationships is motivated by the uncertainties that exist in atmospheric general circulation models concerning the homogeneous or heterogeneous distribution of water ice during periods of high obliquity and resulting transport of polar volatiles equatorward. Geological evidence for distinctive and very large tropical mountain glacial deposits on the NW flanks of the Tharsis Montes shows that the emplacement of ice in the equatorial regions is heterogeneous and is intimately linked to the presence of large volcanic edifices on the broad Tharsis rise. Accumulation of ice and formation of glaciers is likely caused by adiabatic cooling of water-laden polar air masses and precipitation and accumulation on the NW volcano flanks (F. Forget, personal communication). The resulting glacial deposits show interesting relationships to volcanic deposits formed prior to, during and subsequent to glaciation. Evidence that the glacial deposits formed from cold-based glaciation comes from the lack of modification of delicate structures associated with underlying lava flows (seen in detrended altimetry data). Evidence that volcanism occurred during glaciation is five-fold: 1) narrow linear ridges in the glacial deposits radial to the volcano are interpreted as dikes that were intruded into the glacier, rapidly melting the adjacent ice and collapsing, as has been proposed for englacial dike intrusions in Iceland; 2) broad, steep-sided, thick lobate flow-like features are interpreted to represent sill-like subglacial lava flows at the volcano-glacial interface; 3) circular donut-like annuli surround vent-like craters suggesting localized subglacial explosive eruptions; 4) steep asymmetric lava flows at glacial deposit margins are interpreted to be formed by the cooling and banking up of lava against the glacier terminus; 5) collapsed lava flows and local chaotic terrain are interpreted to be places where lava flows descended directly over glacial accumulation zones. Evidence for post-glacial volcanic emplacement is seen in: 1) edifice and flanking rift zone flows that invade and are superposed on the fan-shaped deposits, but show no evidence for associated melting and 2) the surface manifestation of a dike in excess of a hundred km in length that formed rows of tephra and spatter cones and localized lava flows a few to more than 10 km in length, all superposed on the Arsia fan-shaped glacial deposit. Emplacement of tropical mountain glacial deposits on the flanks of the Tharsis volcanoes suggests that there might be meltwater produced from one of several sources: 1) climate-induced (top-down) melting of surface snow in the tropics, 2) thickening of the ice sufficiently to produce basal melting, 3) conduction of magmatic heat to the glacier on the edifice flanks, 4) direct contact of the magma/lava with glacial ice. We find only limited evidence for meltwater processes at the Tharsis Montes, most likely related to dike emplacement and subglacial intrusion/flow events. In contrast, Hecates Tholus may provide an example of a volcanic edifice where summit ice accumulating occurred and basal melting was sufficient to produce the radial valleys that are observed. Thus, comparison of closely associated volcanic, fluvial and glacial deposits can be synergistic and provide new insight into the processes shaping the surface of Mars.
P23B-0192
Stratigraphy of Ganges Mensa, Mars
The slopes and surfaces of Ganges Mensa display a variety of morphologies which indicate that units of variable lithology constitute the stratigraphic stack of the mensa as observed by previous studies. Excellent coverage by the MOC, MOLA, and THEMIS instruments allows correlation between those data sets in order to make improved distinctions between morphologic units. Meter-scale images and detailed elevation data about the extent of these units and their boundaries allows for a better understanding of their associations with one another, and their relation to Ganges Mensa as a whole. This study identifies several thick units (several hundred meters) of erosionally resistant layered materials (some identified by previous studies based on lower-resolution data) and associated land-forms suggestive of a volcanic origin. One of these units is at the highest elevations on the mensa and appears to have acted as a resistant cap. There is evidence that it was once more areally extensive, however the unit which underlies it is more erodible and has been undermining the cap unit causing it to shrink. This morphologic evidence strengthens the hypothesis that Ganges Mensa formed in a fashion similar to terrestrial tuyas. In fact, the presence of these resistant layered units at different elevations within the stack may be markers of different water/ice levels through time, as they are in terrestrial tuyas. However, the true discriminator for the tuya origin hypothesis is the composition of the very thick (some up to 1 km), friable, finely layered units which compose the majority of Ganges Mensa. The tuya origin hypothesis requires that they be hyaloclastites, which has been difficult to discern from orbit. Unfortunately, the unit morphologies do not particularly eliminate any of the other origin hypotheses for Ganges Mensa, but these observations do have implications for them.
P23B-0193
Generation of meltwater by dike intrusion on Mars
The morphology and location of outflow channels on Mars indicate a subsurface origin of the water. However, the origin and processes, that can provide the large inferred discharge, is not well understood. Following previous studies, we hypothesize that dike intrusion played a significant role. The near-simultaneous discharge of magma and water, e.g., at Athabasca Valles, suggest a causal relationship between magmatism and flooding at least in some setting. We thus quantify, using numerical models the conditions, the conditions under which dikes may permit large discharges of water. We perform 2-D numerical simulations of the amount and distribution of meltwater adjacent to a cooling dike to test the hypothesis. We extend HYDROTHERM code developed by the USGS (Hayba and Ingebritsen, 1997) and include the effects of the phase transition between ice and water. We thus account for both boiling and freezing. The interaction between magma and frozen ground is assumed to occur at depths between several kilometers and a few hundred meters depth,. At the depths, boiling may occur without direct sublimation. We evaluate the melting process of the permafrost layer by estimating the following: (1) How much meltwater can be generated? (2) What is the pattern of hydrothermal circulation?, and (3) How fast does the dike cool? We vary the volume an thickness of the dike, and the permeability of the crust. We find that to generate large floods requires a large subsurface reservoir of pressurized water - melting by dikes alone is insufficient to create large floods.
P23B-0194
A New System of Tectonic Outflow Channels in the Memnonia Region of Mars
Using a combination of MOLA, THEMIS, and MOC data, a previously unrecognized set of tectonic outflow channels has been identified in the Memnonia region of Mars. Together with Mangala Valles to the north, these valleys constitute a continuous hydrologic system flowing northwards within the western Tharsis trough for a distance of 1800 km. The overall topographic signatures of the valleys are similar to the broad valley that encompasses the main Mangala Valles channel, with broad flat-floored valleys narrowing in an upstream direction towards their assumed sources within the graben of the Memnonia and Sirenum Fossae graben swarms. The actual channels within the newly identified valleys have been masked through subsequent volcanic resurfacing and tectonic modification, though evidence for a fluvial origin has been preserved in the form of tributaries perched above the level of resurfacing in the main valleys. The valleys must have formed prior to emplacement of the valley fill, which has been mapped as Late Noachian to Early Hesperian in age [Craddock and Greeley, 1994], thus significantly preceding the Late Hesperian to Early Amazonian age of the Mangala Valles channel [Tanaka and Chapman, 1990]. Both these new valleys and Mangala Valles, which together we have termed the "Memnonia-Sirenum outflow complex" after the graben swarms at their sources, likely originated through a mechanism of tectonic aquifer pressurization and subsequent drainage to the surface. This pressurization mechanism relies upon the stress change in the crust resulting from both the injection of a dike at depth and the tectonic release of the gradually accumulated extensional stress generated by Tharsis loading, as described by Hanna and Phillips [2005]. A significant fraction of this tectonic stress change within the aquifer is borne by the water in the pore spaces, so that a tectonic event results in a rapid pressurization of the aquifers surrounding the graben. This pressurized water would then drain to the surface, resulting in the flooding responsible for the observed erosion. The Memnonia-Sirenum outflow complex records an extended period of fluvial activity from the Late Noachian through the early Amazonian, driven by this mechanism of tectonic aquifer pressurization.
P23B-0195
Martian and Ionian Analogs of Permafrost-Volcano Interactions in Alaskan Permafrost
Volcanic landforms in Alaskan lowland permafrost exhibit several unique morphological attributes, as described in a companion AGU abstract (Beget et al.). These features include (1) giant maar sizes (in Bering Land Bridge National Preserve) an order of magnitude larger than common in non-permafrost terrains, (2) composite volcanic forms produced by repeated maar-forming explosions (the novel Ingakslugwat-type volcano in Yukon Delta), and (3) super-inflated lava flows with marginal thermokarst pits (Lost Jim flow, Imuruk Lake Volcanic Field, Bering Land Bridge area). We have identified on Mars, in areas not indicating glaciation, several landforms and on Io an active volcanic process that might be analogs of these in Alaska. On Mars, within and near Elysium (Galaxias Fossae and Hrad Vallis region) multiple crater-like depressions occur with other volcanic features. Their characteristics suggest that the depressions are maars. The composite structures suggest similarities to Ingakslugwat volcanoes. Possible analogs of giant oversize maars also have been identified on Mars. In addition to surface gravitational differences between Earth and Mars, it seems likely that volatile composition is a key aspect controlling the explosivity and sizes of maars on both planets. In Alaska, we suspect that volcanic interactions with methane clathrate hydrate-rich permafrost tends to yield larger maar sizes than with ice-rich permafrost or ground water. This working hypothesis fits well with observations that the giant maars formed during the climatically coldest periods (Beget et al., 2005, this conference). During those periods, permafrost was thick, strong, and unpunctured by lakes and rivers, and so it could have trapped clathrate-forming gases. During interglacials, thinner permafrost and the widespread occurrence of thaw lakes and surface streams may cause the permafrost to be ineffective in confining ascending gases, and so clathrates were absent or not abundant, and volcanic interactions involved mainly weak explosions with H2O. Mars is today and through its history has primarily been in a condition similar to (or colder than) severe full-glacial permafrost conditions on Earth, and so any mantle or crustal sources of methane or CO2 could produce clathrate phases; hence, on Mars, we predict (1) an abudnance of clathrate hydrates and (2) where volcanism occurs in clathrate-rich permafrost, large maar sizes are common. On Io, the Galileo orbiter obtained images of an actively advancing, hot lava flow that was over-riding a volatile-rich substrate, with consequences for gas venting along the flow margins. The geyser-like venting phenomenon is inferred from the occurrence of fresh bright streaks that appeared to have been erupted from points along the margins of the lava flow. In this case, the volatile material is thought to be mainly sulfur dioxide, which exists as solid ice in the shallow crust and as a liquid at greater depths, much like the relationship of H2O phases in terrestrial permafrost. Flow-margin thermokarst in the Alaskan flow may have involved similar volcano-volatile interaction.
P23B-0196
The Martian Surface is old and so are Shergottites
We report new Sm-Nd, Lu-Hf, and Pb-Pb mineral and whole-rock (WR) isotope data for the basaltic shergottite (BS) Zagami (Zag), as well as Pb-Pb WR isotope data for the BS Los Angeles (LA). The isotopic analyses were carried out by MC-ICP-MS at ENSL. The Sm-Nd and Lu-Hf data for Zag yield internal isochron ages of 155±9 Ma (MSWD=0.45) and 185±36 Ma (MSWD=1.2), respectively. While these young ages overlap with earlier Rb-Sr, Sm-Nd, and U-Pb ages (2), the Pb-Pb age does not. Our Pb isotope data on Zag and LA lie on the same Pb-Pb array as previous analyses of BS by (1), which, if interpreted as an isochron, indicate an age of ~4 Ga. The range of δ18O (3.9-5.2 permil) observed in shergottites (3, 4) is too broad to be accounted for by igneous processes only and attests to low-T interaction with fluids. The Martian surface appears to be covered with sulfates, while essentially lacking carbonates (5, 6), implying that the surface of Mars was once covered with acidic water bodies of unknown depths (7). An important observation is that apatite is a common phase in Zag and LA, as in all the shergottites (8), and explains why most of the REE, Th, U, and some fraction of Pb can be removed by leaching (9). The main inventory of Pb, however, resides in maskelynite. The Pb isotope data on shergottites, in conjunction with the existing body of geochemical and geophysical evidence, have important implications for the history of the Martian surface and lithosphere. A fundamental problem with the young crystallization ages for the Martian meteorites has been that these ages are difficult to reconcile with the large $^{182}$W and $^{142}$Nd isotopic anomalies present in these meteorites. On one hand, the anomalies from the extinct radionuclides appear to require a static, non-convecting mantle, whereas widespread volcanism on Mars as young as 150 Ma seems to require an actively convecting mantle. We suggest, based on the Pb isotope systematics of shergottites, that the Martian surface is very old and formed mostly over the first one billion years of the planet's history, thus eliminating the above paradox. We further interpret the young shergottite Rb-Sr, Sm-Nd, and Lu-Hf ages to be the result of the resetting of these isotopic systems by acidic groundwater percolation through the Martian crust, ending approximately 150-300 My ago. We argue that throughout much of Martian history, large acid lakes of regional extent collected and mixed groundwaters and redistributed $^{142}$Nd and $^{182}$W between rocks of different ages, some of them nearly as old as the planet itself and carrying strong isotopic anomalies. From an interpretation of satellite images, it has been argued (10) that over the planet's first billion years of evolution, one third of its surface was covered by bodies of standing water and ice floodwaters derived from a subpermafrost aquifer. The last pools of liquid water occupying various spots on the Martian surface may have disappeared either by evaporation or by retreating into a permafrost layer now buried beneath thick wind-blown deposits. 1. Chen and Wasserburg, GCA 50, 955 (1986). 2. Nyquist et al., Space Sci. Rev. 96, 105 (2001). 3. Clayton and Mayeda, GCA 60, 1999 (1996). 4. Franchi et al., Phil. Trans. R. Soc. London, 359, 2019 (2001). 5. Squyres et al., Science 306, 1698 (2004). 6. Gendrin et al., Science 307, 1587 (2005). 7. Fairen et al., Nature 431, 423 (2004). 8. McCoy et al., GCA 63, 1249 (1999). 9. Dreibus et al., LPS 27, 323 (1996). 10. Clifford, Icarus 154, 40 (2001).