P34A-01
Martian Valleys, Channels and Gullies at High-resolution
Instruments on the Mars Reconnaissance Orbiter are providing new perspectives on fluvial processes on Mars. As the various datasets have started to mature, we are able to begin to synthesize data from multiple instruments to gain unified insights into the fluvial history of the planet. I will focus on the synthesis of CTX, HiRISE, and CRISM data at several high interest regions on the planet. Both CTX and HiRISE images reveal that post-formation processes heavily modify the valley networks and outflow channels. Although erosional features are present in some, channel bedforms and deposits are largely absent. Many valleys and channels (as well as surrounding impact craters) are filled by aeolian transported material, as indicated by the formation of dunes along the floors of these fluvial landforms. The valley networks and channels themselves may have provided an additional source material for the fine-grained infill material. Dunes are frequently found on the floors of valleys and channels, but often not on the adjacent surrounding plains. One possibility is that fine-grained material forming these dunes may be a combination of aeolian infill and reworked fluvial deposits. Another high priority target for the MRO instruments has been the intriguing, relatively young gully features. While HiRISE has sampled a variety of gully types around the planet, we have focused on a few study areas to systematically understand gully formation in a single environment. One of these study areas is the Noachis ring trough (or eroded pit crater) in the southern highlands (47S, 5E). Gullies are located around much of the perimeter of the ring-shaped depression as well as along the interior walls of the central plateau and within the innermost central pit. Debris aprons from those gullies found along the walls of the central pit overlap and coalesce on the floor. Channel forms are located on the debris aprons. Generally at Noachis, north-facing slopes exhibit light-toned layers near the gully source regions at the top of the gullied slopes. Bright deposits are also seen along some debris fans. South-facing slopes seem to lack these light-toned layers along the upper walls and instead gullies seemed to have formed on jumbled, bouldery slopes. In fact, a sharp divide formed by a small crater intersecting the larger trough feature on the east side of Noachis shows a similar relation with light-toned layers on the gullied north-facing slope and a jumbled, bouldery, undercut gullied slope on the south-facing wall. Light-tone layers are again absent along the south-facing wall. CTX images also support the observation of light-toned layers on the upper north- facing slopes. This suggests that while slope aspect does not seem to control the actual formation of gullies, it does appear to affect the morphology as well as the presence or absence of the light-toned deposits. HiRISE images do show that frosts tend to remain longer on the poleward facing slopes and CRISM data do confirm the present of CO2 and water-ice frost in these regions which perhaps may contribute to the accelerated breakdown of the light-toned deposits on the poleward (south) facing slopes.
P34A-02 INVITED
The Key is in the Bend: How Sinuous Gullies can Improve the Understanding of Their Flow Properties
Recent gullies are erosional landforms developed on hillslopes in recent Martian time (less than 10 My). The nature of the fluid that created the erosion is still under discussion. On one hand, debris flows containing both liquid water and rocks are frequently invoked, but classic river stream activity dominated by liquid water is also considered. On the other hand, hypotheses invoking only dry granular slides explain many characteristics of the gullies channels, their size, length, presence of levees and terminal deposits. In addition, exotic fluids such as gaseous CO2 flows are still under consideration. The physics of these different fluids can lead to very similar flows, thus, a detailed observation of their morphology is required. For this reason, we have focused our interest to the presence of sinuous gullies, which channels present bends similar to meanders of rivers, with the difference that they occur into a steep slope for a channel, i.e. typically 10 to 20 degrees. The cause of the sinuous shape is usually thought to be due to a perturbation on the flow or a variation of discharge rate. We present models of dry granular flows disturbed on their travel: they display a bend, but no sinuosities are formed subsequently. The sinuous shape over such slopes is typical of a fluid flow that is not reproducible neither in dry granular flows, nor in a classical river stream. Several characteristics can be used to estimate the stream velocity and viscosity from these bends. Especially, the presence of levees that are dissymmetric in the bend is used to measure that the flow is relatively slow (1-3 m/s) with high viscosity (about 10,000 Pa.s). This viscosity is typical of debris flows containing minor proportions of liquid water compared to solid debris, but it is far above the viscosity of river streams, even with high sedimentary flux. In summary, the sinuous shape over slopes higher than 10 degrees requires material consistent with debris flows (rock/water mixture) whereas dry flows or river stream are unlikely. A question still under consideration would be to know if another material such as gaseous CO2 mixtures could mimic these properties or not.
P34A-03
The Formation and Evolution of Youthful Gullies on Mars: Surface Melting at the End of Mars' Most Recent Ice Age
Martian gullies provide evidence for surface flow of liquid water in the last several million years, which has been difficult to reconcile with the low-temperature/low-pressure conditions observed today. Several global, hemispheric, and regional surveys of Mars Global Surveyor, Mars Odyssey, and Mars Express data have revealed the three-dimensional distribution of these features across the planet and have shown that gullies only form in regions conducive to the deposition, accumulation and preservation of atmospherically-emplaced water in the form of snow, frost, or ground-ice. Two primary hypotheses have persisted: that ground ice serves as a dam for a sequestered aquifer at depth which is released suddenly, or that ground ice, snow, and surface frost melt at the surface or within the active layer. Multi-spectral data from Mars Reconnaissance Orbiter reveal that (1) gullies are episodic systems, which demands that any model for gully formation provides a recharge mechanism; (2) cold-trapping of condensed volatiles occurs at the alcove/channel level; (3) gullies form at the same locations as features plausibly interpreted to represent previous episodes of ice-related flow; and (4) gully channels frequently emanate from the crest of gully alcoves instead of the base, showing that alcove generation is not necessarily a product of undermining and collapse at these locations. These new data change the scope of the gully debate: any model for gully formation must also account for the evolution of gully systems over time. We argue that gullies occur within the temporal context of orbitally-induced transport of ice from the poles to the mid-latitudes and back, and that gullies represent the final stages of this recession as water ice became less stable in the mid-latitudes. This provides an end-to-end model for gully initiation and reactivation that bears little dependence on subsurface stratigraphy, and is instead governed by localized cold traps within micro-climates that provide the only places on Mars conducive to the preservation and melting of snow/frost within the last few million years. This is consistent with the majority of gully systems on Earth, with the low-temperature and low- pressure climate of Mars generating short channel systems that terminate on steep slopes, consistent with observations of martian gullies to date.
P34A-04 INVITED
Diverse Alteration Minerals Around Martian Impact Craters Revealed by MRO-CRISM: Indicators of Hydrothermal Activity or Subsurface Aqueous Alteration?
Hydrated silicates indicating aqueous alteration of mafic rocks have been identified in central peak, wall, and ejecta rocks of many impact craters. CRISM has revealed 1000s of exposures in the Southern Highlands, often associated with craters, with Fe/Mg smectite-bearing rocks, and a small percentage with more diverse alteration minerals. Chlorite is not uncommon and kaolinite, illite or muscovite, hydrated silica, and zeolites are sometimes present, albeit infrequently. Understanding the geologic setting and formation processes of aqueous mineral deposits is important for determining whether alteration pre-dated the impact crater or whether mineralogic products are more recent and represent evidence for post-impact alteration, perhaps in hydrothermal systems. Using combined CRISM-CTX-HiRISE observations, we assess this in detail for craters of unusually diverse alteration mineralogy in the region west of the Isidis basin. Three types of information are considered: (1) Stratigraphy and geomorphic setting of crater hydrated silicates, (2) constraints on the temperature, pressure, and geochemical conditions required for the formation and stability of each alteration mineral, and (3) models of the temperatures experienced in impact processes. Two main classes of craters with diverse alteration minerals exist between the Nili Fossae and Antoniadi basin: (1) craters dominated by chlorite with illite and smectite also present and (2) craters with Fe/Mg smectite, chlorite, and the zeolite analcime in the central peaks. In (2), materials in sands ringing the peak are hydrated and Si-OH bearing, which may indicate either hydrated silica (e.g. chalcedony) or an aqueously altered basaltic glass. These sands also correspond to the unique units identified by TES with elevated quartz and alkali feldspars and interpreted to be quartzofeldspathic (QF)/granitoid material. Fe/Mg-rich smectites or chlorites with accessory zeolite, silica, quartz, and K-feldspar result from hydrothermal alteration in terrestrial craters (e.g. Allen et al., 1982; Naumov, 2005) and suggest a possible reinterpretation of the QF material in (2) as hydrothermal in origin rather than igneous. Experimental data show analcime-smectite assemblages result from aqueously altered glasses and basalt powders at temperatures T < 200 C (Robert and Goffe, 1993). Illite-chlorite assemblages are stable at T~200- 260 C. Variation in the temperature, water-rock ratio, and fluid chemistry of a hydrothermal system may result in these distinctive mineral assemblages observed in Nili Fossae craters. An alternative hypothesis is that alteration materials associated with craters reflect changes in underlying crustal materials west of Isidis and that aqueous alteration pre-dated formation of the impact structures in which these minerals are mapped. Smectite transforms to illite upon burial at T>50-80 C, and illite and chlorite altered at depth may have been excavated by impact. Both hypotheses--subsurface pre-impact alteration and post-impact hydrothermal alteration--will be examined with new MRO data.
P34A-05 INVITED
Spring Deposits and Mud Volcanoes on Mars
We report evidence for spring deposits in Vernal Crater, Arabia Terra. The Vernal structures are low
mounds, each approximately 250 m by 500 m in extent, with terraced flanks, apical depressions, river-like
channels, concentric fractures, and elliptical tonal anomalies. All of these features are common in terrestrial
springs such as the Dalhousie complex in Australia. The structures occur in an apparent unit of interdune,
water-laid sediment and are associated with evidence of subsurface fluid flow in sets of aligned outcrops. The
Vernal springs may be part of a larger complex of spring deposits and lineations, suggesting that fluid flow in
this region was relatively extensive.
The structures in Vernal Crater are coated with a thin layer of dust, which prevents mineral identification from
orbit. In an attempt to find evidence for additional spring mounds, we conducted a survey of nearly 2,000
locations for which CRISM spectral images are available. We used CRISM data to identify dust-free, hydrated
areas and HiRISE, CTX, and MOC images to evaluate morphology. This study covered all longitudes and
latitudes from 50N to 70S, except near Tharsis where data were analyzed south of 15S. No location exhibited
morphological features that closely resembled those in Vernal Crater, suggesting that these putative spring
mounds are not common in the highlands of Mars.
Our search led us to concentrate on a dust-free area, centered at 41.8N, 332.5E in Acidalia Planitia where
Farrand et al. (2005) identified features resembling spring mounds or mud volcanoes. Tanaka et al. (2005)
mapped this region as part of the Early Amazonian Vastitas Borealis Unit, interpreted as reworked sediments
from outflow channels and highland sources.
We mapped over 20 high-albedo pitted domes in the area covered by one HiRISE frame, with dome
diameters ranging from 350 m to 1 km. Nearby, similar domes have measured heights ranging from 36 to 65
m. The dome material is darker in THEMIS nighttime IR than the surrounding plains, indicating that the domes
have relatively lower thermal inertia. The dome material is also very smooth, and appears smeared across
the textured plains and in local depressions as if it were emplaced by low viscosity flows.
CRISM spectral data (Leah Roach, Brown Univ.) were used to assess the mineralogy of selected features.
Dome spectra exhibit a steep shoulder from 0.4 to 0.6 microns and are generally featureless in the near IR,
overall resembling the spectra of bright Martian soil. The steep shoulder is due to the oxidization of iron to
Fe3+. Dome spectra do not exhibit evidence for hydrated minerals or precipitates such as carbonates or
silica.
While a range of origins has been suggested for the Acidalia domes, we believe that they are most consistent
with a mud volcano analog. A volcano or pseudocrater origin is unlikely, as no lava flows or volcanic features
are observed in the vicinity. A spring mound origin is also unlikely, due to the absence of hydrated or spring
precipitate mineral signatures in the CRISM spectral data as well as the absence of terracing, channels, and
circumferential faults that typify the Vernal springs.
In addition to the Vernal springs and the Acidalia mud volcanoes, several other recent studies have pointed
to evidence of subsurface fluid flow. These include the large spring deposits proposed by Rossi et al. (2008),
the resistant knobs in Candor Chasma (Chan, 2008), and the bleached zones along faults in Valles
Marineris (Treiman, 2008). Together, these features are changing our understanding of the hydrologic
history of Mars.
class="ab'>
P34A-06
High resolution imaging of the outflow channels on Mars
We report observations of the outflow channels on Mars from HiRISE images in MRO's first Martian year. Several hundred images of the outflow channels on Mars have been collected to date from HiRISE, as well as coordinated images with CTX and CRISM. Depositional features, such as slackwater deposits and small bedforms that are expected to be visible at the resolution of HiRISE have not yet been observed, largely due to post-fluvial modification of the channels. Many of the channels have been subsequently covered by a thin layer of lava, ash, dust, or lineated valley fill. Although altered slightly by later aeolian modification, Ares Valles and Kasei Valles preserve much of the original fluvial erosional forms, particularly cataracts and longitudinal grooves that can be used to infer the mechanics of the flow. Cataracts, steep knickpoints in the large outflow channels, were once large waterfalls on the Martian surface. These have been observed in all of the larger outflow systems, including Kasai, Athabasca, Mangala, and Reull Valles. High resolution imaging shows that all of the cataract systems have multiple generations of erosion, with smaller subchannels within the cataract system. Based on the length of the recession and the morphological evidence most of the large channels experienced multiple flooding events or pulses. The tectonically sourced outflow channels, such as Athabasca and Mangala Valles, show sourcing at regions of complex fault geometries, specifically at fault relays. In terrestrial systems, relays tend to be regions of concentrated stress that can produce dilation manifested as high joint density, as well as point sources for hydrothermal outflow on Earth. Athabasca and Mangala Valles, sourced proximal to large volcanic centers, may have been regions of major hydrothermal activity in the past.
P34A-07
Waterfall erosion, rock toppling, and the formation of amphitheater-headed canyons in fractured rock
Networks of valleys with amphitheater-shaped headwalls are prominent features on the surface of Mars. These landforms are commonly used as diagnostic indicators of undermining and headwall retreat by groundwater-seepage erosion. Of perhaps any canyon ever studied, Box Canyon, Idaho, most strongly meets the proposed morphologic criteria for groundwater sapping because it is incised into a basaltic plain with no drainage network development upstream, and approximately 10 m3/s of seepage emanates from its vertical headwall. However, we have found strong evidence that this canyon was carved during large-scale flooding about 45,000 years ago. Such evidence includes 4He and 14C dates, plunge pools, large boulders, and scoured rock along the rim of the canyon headwall. To explain the formation of the amphitheater headwall of Box Canyon, we propose that near vertical knickpoints can persist during retreat due to waterfall- induced toppling in fractured rock (e.g., columnar basalt). At a waterfall, rock columns are affected by shear and drag from the overflowing water, buoyancy from the plunge pool at the foot of the waterfall, and gravity. A torque balance is used to determine the stability of a rock column and any individual blocks that compose the column. Model results and flume experiments indicate that rotational toppling failure should occur about the base of a headwall (and therefore preserve its form during upstream propagation) where columns are tilted in the downstream direction, or slightly tilted in the upstream direction depending on the plunge pool height. We propose that such conditions are probably common in columnar-basalt bedrock. Thus, our toppling model might explain the origin of steep amphitheater headwalls in volcanic terrains on Earth and Mars by overflowing water and in the absence of seepage.
P34A-08
EPISODICITY IN THE GEOLOGICAL EVOLUTION OF MARS: ANALYSIS OF AGES FROM CRATER COUNTS ON IMAGE DATA AND CORRELATION WITH RADIOMETRIC AGES OF MARTIAN METEORITES.
In early attempts to quantify the geologic history of Mars by crater counting techniques, most of the geological units and structures came out as being rather old, in the range of billions of years with an exception of the Tharsis province, whose volcanic constructs were found to already have existed more than 3.5 Ga ago, but which showed, at least partly, signatures of relatively young (hundreds of millions of years) volcanic activity. On the other hand, most of the ages of the martian meteorites cluster at relatively young values of around 175 m.y., 300-600 m.y. and ~ 1.3 Ga, whereas very few old ages >3 Ga had also been found. The early cratering age determinations were based on the Viking image data analysis. With the new data from MGS (MOC), MEX (HRSC), and Mars Odys-sey (THEMIS), it has become clear that the apparent discrepancy between the two age sets was a se-lection effect due to the limited Viking resolution forcing to study predominantly large, old features. Significantly younger ages have been determined since on the basis of the new high-resolution im-agery with spatial resolutions in the meter to a few tens-of-meters range. In this work we report on results from investigation of a combination of HRSC, MOC and THEMIS imagery in ten regions of Mars, such as the regions of Mangala Valles, Kasei Valles, Libya Montes, Iani/Tiu/AresValles, Medusae Fossae and five more. We have in particular mapped out and analyzed for their geologic evolution and cratering ages the two large outflow channel areas, Mangala Valles and Echus Chasma/Kasei Valles. In both areas we have found multistage geological histories with mixed volcanic, fluvial, glacial, and hydrothermal activity. The new data in combination with the previous data have been analyzed by way of a refined method of cratering age extraction also giv-ing fine details of periods of resurfacing from the characteristics of the measured crater size-frequency distributions as they deviate from the production size-frequency distributions due to resur-facing effects. Our data show, in confirmation of more subtle indications over the past two years of research, that there has been volcanic and fluvial geologic activity on the martian surface at all times from >4 Ga ago until today. This activity was rather continual if to consider all ten studied regions together but with episodic temporary increases of intensity of both volcanic and fluvial and/or glacial processes at periods ~ 3.5 Ga, 1 to 1.5 Ga, 300 to 800 m.y., ~ 200 m.y. ago. The episodes we find on the martian surface in the crater size frequency analyses are well coincident with the age groups of the martian meteorites found from radiometric dating. Episodes of volcanic activity were probably initi-ating the fluvial and/or glacial activity. This explains their correlation in time.