Creation of the

Valles Marineris

I. Introduction
The Valles Marineris is a vast system of canyons on the surface of the planet Mars, just south of its equator at 11.6° south and 70.7° west. It is easily one of the most recognizable features to a planetary geologist and within it lie the largest canyons known (so far) by mankind. With a minimum depth of over six kilometers, this canyon is sure to contain vast amounts of geological data.

One of the key ingredients required to harbour life is the presence of water, a concept that is the driving force behind many planetary studies. "Follow the water" is a saying often used by NASA in their education outreach programs. The idea being that where there is water, there might be life.

Therefore an ideal location to analyze this paradigm would be one that contains land-forms and features typically seen in sub-aquatic, fluvial environments, which as it just so happens, perfectly characterizes the appearance of some of the features in the Valles Marineris. Many discussions throughout the years have generally concluded water existed on Mars at one time or another. So while it is generally accepted that there is water, where this water is and where it has been remains hotly debated. By examining possible scenarios for how this massive canyon system was created and changed over time, more accurate hypotheses can be drawn as to the fate of Martian water.

II. Remote Sensing
In an age where probes and people can't yet go to Mars to drill for themselves, the more that can be inferred from current data, the more focused future studies can be, thus maximizing scientific return. Data from Mars is collected in one of three ways: Earth-based observations, observations from Martian orbit and observations from the surface of Mars itself. The first method of observation unfortunately does not provide good imaging capabilities, and in fact is not able to detect Valles Marineris. The latter two however are, and can do so in amazing detail.
IIa. Probes
Image credit: NationalSpace ScienceDataCenter
Figure 1: Image of Noctis Labyrinthus,
west most part of Valles Marineris, captured by Mariner 9
Credit: NationalSpace ScienceDataCenter
The first probe to detect the Valles Marineris was Mariner 9, when it snapped photos of the feature in question in 1971 (Figure 1) as part of its normal mapping operations. As a tribute to its discoverer, this newly-found canyon system was called the Valles Marineris (Mariner Valley). Two Viking probes which followed took extensive photographs of this region and is the basis for many Mars studies today since these data sets are more mature than recent sets from current missions. The Viking I and II probes consisted of each an orbiter and a lander, and between them managed to image the entire planet. Of relevance to the Valles Marineris, the Viking I lander touched down on July 20, 1976 in Chryse Planitia, an outflow plain north-east of the canyon. The Mars Pathfinder mission in 1998 also touched down in Chryse Planitia, however since there was no orbiter component to the Pathfinder mission, the data obtained is very localized to its landing site.

There are two probes actively studying Mars: Mars Global Surveyor (MGS) and Mars Odyssey. Both of which have numerous instruments on board to help study Martian geology, and both of which target the Valles Marineris in the course of their normal operation. It is from these probes that the most detailed observations are made, however much of the information collected was only very recently, or still is, in peer review and not yet widely available for interpretation.
IIb. Instrumentation
Image credit: NASAs Jet Propulsion Laboratory
Figure 2: Diagram showing source energy and radiation
for the operation of a gamma-ray spectrometer in space
Credit: NASAs Jet Propulsion Laboratory
Among these various probes were carried four main instruments of interest. The first is a Gamma Ray Spectrometer (GRS), which operates and obtains data similarly to spectrometers here on Earth, and resides on the Mars Odyssey craft. With this device, it is possible to measure the abundance and distribution of around 20 different elements, including silicon, oxygen, iron, magnesium, potassium, aluminum, calcium, sulfur and carbon. This device is also sensitive to chlorine, a key element that in significant amounts may suggest the presence of ancient lakebeds and/or oceans (University of Arizona, 2002). The GRS differs from spectrometers on Earth primarily in the energy source used to excite atoms in the sample being viewed. The feasibility of using an on-board power source prohibit the instrument from actively causing excitation, so instead it observes gamma rays omitted by the Martian surface in response to being struck by ambient cosmic rays, most of which originate from the Sun (Figure 2). The GRS was only fully deployed on June 4, 2002, so data returned is still rather preliminary.

Image credit: NASA/JPL/Arizona State University
Figure 3: False colour infrared image from Mars Odyssey
draped over elevation data from Mars Global
Surveyor. Different colours represent different
minerals (e.g., dark purple represents olivine).
Credit: NASA/JPL/Arizona State University
The second instrument of interest is the Thermal Emission Imaging System (THEMIS), also aboard Mars Odyssey. This device primarily measures infrared (IR) light to detect mineral compositions (Figure 3). Infrared light is emitted by any object with a temperature greater than 0K, allowing observations of temperature variances throughout day-night cycles. In addition, every mineral, gas and compound has a unique spectral signature (Arizona State University, 2001) that can be compared with known pure signatures on Earth.

The third instrument is the Mars Orbiter Laser Altimeter (MOLA) on-board MGS. Its purpose is to measure the precise altitude and topography of Mars by firing IR laser pulses at the Martian surface and measuring the amount of time passed for the reflected beam to be received by the MOLA.

The fourth set of instruments, and perhaps the most important, are the optical (visible light) cameras aboard every probe to Mars. These range in resolution capabilities from 1km per pixel for the Mariner 9 probe, 8m, 150m and 300m per pixel for the Viking Orbiters, 100m per pixel for THEMIS aboard Mars Odyssey, to 1m per pixel for the Mars Orbital Camera (MOC) on MGS (Arizona State University, 2002; Williams, D.R., 2001; Williams, D.R., 2002).

III. Geography
The terrain in and around the Valles Marineris varies greatly, and the basic geography can be inferred from numerous photographs of this area.

Image credit: NASA/JPL/Arizona State University
Figure 4: Map of Valles Marineris and surrounding area

Credit: NASA/JPL/Arizona State University
Generally, the area surrounding Valles Marineris comprises ancient plains and is relatively smooth. To the west of the Valles Marineris is the Tharsis region, an area of unusually high elevation and home to some of the largest volcanoes in the solar system. Also in this area is the Noctis Labyrinthus, a large, complex depression with smaller inter-connecting canyons and valleys that forms the 'start' of the Valles Marineris. To the north and south lie the Lunae, Ophir, Sinai, Solis and Thaumaia plains, each spanning tens of degrees of latitude (Figure 4). In the east, at the 'end' of the Valles Marineris is the Chryse Planitia, an area typified by chaotic terrain and likely an outflow area.

The canyon itself is made of up several smaller canyons, or "troughs" (Figure 4) and somewhat decrease gradually in altitude from west to east. The major canyons, or "chasmata" as they are referred to, are the Tithonium, Ius, Ophir, Candor, Melas and Coprates chamsa, with the largest being Melas followed by (in no particular order) Coprates, Candor and Ius.

Some of these troughs are 50 - 100km in width, several hundred kilometers in length and about two to eight kilometers in depth (Schultz, R.A., 1991).

Closer inspections reveal a complex system of faults and grabens, macroscopically parallel to the predominant laterality of the trough walls (Schultz, R.A., 1991). In the troughs not of the "rectangular" type (e.g. Ophir Chasma), the orientations of the slip faults of grabens in these areas are not parallel to the chaotic canyon margin, instead they parallel those faults seen in the rectangular troughs, thus implying that this area underwent deformation early on, before the canyon walls had much chance to erode (Schultz, R.A., 1991).

Some of the walls in the troughs have circular or semi-circular erosional patterns, which are believed to be the result of old craters (on the surface or buried) (Malin, M.C. and Edgett, K.S., 2000).

Another interesting point to note is that in both Coprates and Melas chasmata, the material of the canyon floor appears similar in morphology and structure to the rock from the Ophir Planum.

IV. The Present
In order to determine the history of the Valles Marineris, the current-day features must be noted and carefully examined, with attention paid to how they may relate to possible previous processes.

Copyright Calvin J. Hamilton
Figure 5: Mass wasting in the Valles Marineris.
Evidence exists of recent geological activity, primarily in the way of mass-wasting processes, as seen in Figure 5. These along with patches of chaotic and knobby terrain are indicative of repeated, continual collapsing of wall-rock debris (Tanaka, K.L, 1997).

Images from other parts of the Valles Marineris show that surface outflow channels are apparent and connect most of the canyons (Figure 4) (Tanaka, K.L., 1997). Eastward-heading debris flows are also observed covering parts of Ius and Coprates Chasma (Hauber, E., 2000). Additionally, it can be seen that floor material from the Coprates, Melas and Candor chasmata are relatively flat with a moderate crater density. Orbital observations also show that Melas Chasma has thin eolian deposits.
IVa. Gravimetric and Magnetic Data
Image credit: NASA/GSFC/Laboratory for Terrestrial Physics
Figure 6: Gravity anomaly map of both Martian hemispheres
Credit: NASA/GSFC/Laboratory for Terrestrial Physics
Using non-optical instruments, Mars Global Surveyor has made detailed gravity maps of Mars, which under analysis, show an obvious negative gravity anomaly at the Valles Marineris (Figure 6). This data coupled with altimeter readings from the MOLA tend to indicate crustal thinning due to lithospheric stretching (Hauber, E., 2000). Magnetic field observations are outlined on a planetary scale in Figure 7.

Image credit: NASA/GSFC/Laboratory for Terrestrial Physics
Figure 7: Radial magnetic field map of both Martian hemispheres
Credit: NASA/GSFC/Laboratory for Terrestrial Physics
The research discerning the geology of modern-day Valles Marineris has made several key findings. Dark deposits are common throughout the troughs and can possibly be associated with volcanic vents (Lucchitta, B.K., 2002), although for unknown reasons these deposits are "geochemically depleted" when compared to other Martian volcanics (Scott, E.D., 2000). The process of distinguishing contact relations between trough walls and the floor proved difficult in Malin and Edgett's (2000) study of Martian sedimentary rocks, since most of the floor contacts the wall with these dark-deposit volcanic materials.

Another significant feature seen in the Valles Marineris are Interior Layered Deposits (ILDs). The nature of these mesas is not readily apparent, although some researchers believe they are volcanic in origin (Lucchitta, B.K., 2002), and that they are definitely younger than floor areas without dark deposits (Malin, M.C. and Edgett, K.S., 2000).

In a 2001 study by Christensen, P.R. et al attempting to map possible water outflow regions, the global mapping of Mars for the mineral hematite revealed significantly elevated levels of iron oxides in localized areas of the Valles Marineris floor. They noticed that they were related with the same dark deposits observed by Lucchitta, Malin and Scott, and that no hematite-rich materials were evident in areas that did not have ILDs (e.g., Ius Chasma and Coprates Chasma).

Tanaka (1997) noted that Martian crust is mostly made up of poorly consolidated material and easily triggered by nearby seismic or volcanic events. This may have served a role in the Valles Marineris' creation if either of these types of events were present.
IVb. Rock Sequences
The outcroppings observed in the Valles Marineris are very continuous and invariable over larger distances. Indeed, the rock sequences in this area are similar to those found in craters of western Arabia Terra (and elsewhere) (Malin, M.C. and Edgett, K.S., 2000).
IVc. Fault Systems
Extensive fault systems are present around the Valles Marineris (Schultz, R.A., 1991), radiating outwards from the Tharsis region to extend parallel to the trough systems. Grabens are prevalent throughout the canyon system, and those structures observed in Coprates Chasma point to scenarios analogous to continental rifting on Earth, indicating rifting of the Martian lithosphere (Schultz, R.A., 1991).

V. The Past
Numerous hypotheses and theories exist as to how the huge, 4000 kilometer-long scar that is Valles Marineris was created. Three of which are easily and continually supported by observed evidence, and thus designated "major" theories. The others are either a combination of major theories or are more radical and thus designated as "minor" theories. However, before each theory can be presented, the constituent processes must first be induced and examined.
Va. Geologic Record
Dating of Martian surfaces is done by way of measured crater flux rates using the moon as a known baseline. While it may be possible to obtain very approximate absolute age ranges on Martian surfaces by extending lunar parameters to Mars, there are too many unknowns and large uncertainty values that would make these numbers almost useless (Malin, M.C. and Edgett, K.S., 2000). However, relative ages can be determined as long as sufficient craters are available, otherwise the data again becomes unreliable.

Dating of layered outcrops is unsurprisingly determined through stratigraphic and crosscutting relationships (Malin, M.C. and Edgett, K.S., 2000).
Vb. Tectonic Events
Image credit: NASA/GSFC/Laboratory for Terrestrial Physics
Figure 8: Relative altitude of Tharsis - Valles Marineris area.
Red represents high altitudes, blue represents low altitudes.
Credit: NASA/GSFC/Laboratory for Terrestrial Physics
As mentioned earlier, the Tharsis region (Figure 8) has a very high altitude in comparison with the rest of Mars. Gravimetric data shows a general, dome-shaped gravity anomaly in this area, as seen in Figure 6, bound to the same geographical area labeled the "Tharsis region. This (generally agreed) tectonic event crosses over 90 of latitude and deserves admiration in its own right. This region was the source of the most recent large volcanic event on Mars (Malin, M.C. and Edgett, K.S., 2000), and is where most tectonic activity on Mars took place (Harder, K. and Christensen, U.R., 1996).

One possible explanation for the upwelling in the Tharsis area (the relevance of which will later become evident) is that there may be a single, large mantle plume upwelling into this area, fitting gravimetric and crustal thinning observations (Harder, K. and Christensen, U.R., 1996). Harder and Christensen's modeling has shown in two dimensions that an endothermic phase boundary, caused by chemical fractionation, in the lower mantle and above the core would greatly alter convection patterns, possibly resulting in one or two plumes given sufficient time. Their recent three-dimensional modeling of this same process confirmed their earlier results, and additionally noted that the remaining plumes remain stationary and do not migrate. However, their simulations note that the single plume is much larger size at 7.8 billion years after planet accretion than it is at 4.5 billion years, therefore implying the stresses and volcanics in this area should only increase. Alas, a possible explanation for this is the cooling of the Martian mantle over time.

There are three sets of faulting seen created by the Tharsis stress field (Schultz, R.A, 1991). One set is geometrically centered in Syria Planum, and the other two sets (Pavonis I and II) are centered near Pavonis Mons. Schultz (1991) determined that faults seen in and around the Valles Marineris are the same as those as Pavonis I.
Vc. Major Theories
From the observations at hand, it is now possible to derive possible major theories for the creation of Valles Marineris. The first of which, is the erosional theory.

The voids left behind in the canyon system represents huge amounts of mass that would need to be removed (Malin, M.C. and Edgett, K.S. 2000). By performing some calculations, Schultz (1991) determined the approximate amount of material unaccounted for by the voids represented by the main troughs is close to 3 x 10^6 km3. He is also able to conclude the maximum age for a large erosion event to be in the Late Hesperian. Pure erosional theory states that all this missing material had to be removed by wind erosion, surface water flow or mass wasting (typically landslides).

An obvious problem with this theory is it provides no details on an initiating event or set of circumstances, and does not explain why this phenomenon is not seen elsewhere on Mars to this vast degree.

The collapse model for the creation of Valles Marineris however does provide an initiating event. In this process best described by Schultz (1991), fractures attributed to the Tharsis stress field unsteadies the area in question. Subsidence and collapse then occur from ground water/ice retreating and/or magma withdrawal. Macroscopic observations would show this starting out with the formation of pit craters, resulting from internal collapsing, and having them slowly grow, merging with other pit chains, eventually forming troughs as they are seen today. However, even Schultz makes the point that pit craters and chains are structurally and morphologically different from large troughs, implying that other processes may be at work.

A variation of the collapse theory is the Keystone collapse model (Schultz, R.A., 1991), based upon the idea of lithospheric stretching directly underneath Valles Marineris and the large amounts of normal faulting and graben formation that would be associated with such an event.

In a refreshingly succinct analysis, McKenzie and Nimmo (1999) hypothesized on the effect of large dyke intrusions. With such large amounts of volcanics associated at one time in this region, coupled with the fact that the Valles Marineris is radial to the Pavonis I Tharsis fault system, their assumption is a plausible one. In their study, they indicated such dykes could provide enough heat to melt ground ice, and over a large enough area to mobilize enough water to account for the present-day topography. These flow rates were calculated to be around 20,000,000 cubic meters per second, with a total water volume of 2,000 cubic kilometers. By making further assumptions, McKenzie and Nimmo calculated that the cooling of one kilogram of basalt from 1,500K to 200K would release enough heat to melt five kilograms of ice. Therefore, their whole system of events would arrange like thus. A dyke intrusion would cause structural collapse and faulting at the surface, exposing the permafrost layer. By heat released by the dyke as it cools, permafrost would melt and either leech out the slipped wall, or be trapped underneath existing permafrost. As more ice melts, more water is mobilized and trapped, eventually until a breaking point is reached and the water is released catastrophically.

The tectonic theory of Valles Marineris creation appears to be the most viable case for the production of the troughs because of observed indisputable evidence for down faulting along trough margins (Schultz, R.A., 1991). Schultz (1991) puts forth three possible causes of tectonic activity: regional stresses from the Tharsis stress field, localized uplift and doming around the Valles Marineris, or polar wander. Because of the link between the troughs and the faulting induced by the Tharsis region, a tectonic hypothesis is favoured over erosion and collapse hypotheses (Schultz, R.A., 1991).
Ve. Alternative Theories
The Convective Removal of a Lithospheric Root theory, put forth by Scott (2000), is an alternative initiating event to the three major theories already given. She hypothesizes the existence of a lithospheric root that at one time anchored the Syria Planum area in buoyant equilibrium. Once the root was removed via convective eddies in the mantle, the area began to rebound, after which one of the three major theories takes over.

VI. Conclusion
To quote 18th century geologist James Hutton, "the present is the key to the past." A paradigm that proves itself effective, even in cases of extraterrestrial geology. Exclusively through remote sensing techniques, scientists have begun unraveling Valles Marineris' past and are one step closer to finding possible signs of life on other planets, however remote they may seem. Probably the most likely creation scenario is a combination of all three of the major theories present, in which case future researchers will certainly have their work cut out for them.

Suggested areas for future study would include a reexamination of all results already obtained, with an effort to enhance and improve them with new visual, mineralogical, and altitudinal data streaming in from the Mars Odyssey and Mars Global Surveyor probes. For example, the ability to correlate rock assemblages based on mineralogical similarities instead of purely visual ones should open up a whole new realm for observation.

Arizona State University. 2002 "Mars 2001 Odyssey THEMIS" Mars 2001 Odyssey Themis 2002. Retrieved March 8, 2003 from the World Wide Web:

Arizona State University. 2001 "What is TES?" ASU Planetary Exploration Laboratory 2001. Retrieved March 8, 2003 from the World Wide Web:

Christensen, P.R., Morris, V.R., Lane, M.D., Bandfield, J.L., and Malin, M.C. 2001. "Global Mapping of Martian Hematite Mineral Deposits: Remnants of Water-Driven Processes on Early Mars" Journal of Geophysical Research 106: 23,873-23,885

Harder, H., and Christensen, U.R. 1996. "A One-Plume Model of Martian Mantle Convection" Nature 380: 507-509.

Hauber, E. 2000 "The Structure of Valles Marineris, Mars: A Combined Analysis of Viking and MOC Images and MOLA Elevation Data" Geophysical Research Abstracts 2000. Retrieved February 19, 2003 from the World Wide Web:

Lucchitta, B. K. 2002 "Are the Valles marineris Giant Volcano-Tectonic Depressions?" Lunar and Planetary Science XXXIII 2002. Retrieved February 10, 2003 from the World Wide Web:

Malin, M. C. and Edgett, K. S. 2000. "Sedimentary Rocks of Early Mars" Science December 8: 290.

McKenzie, D. and Nimmo, F. 1999. "The Generation of Martian Floods by the Melting of Ground Ice Above Dykes" Nature 397: 231-233

National Air and Space Museum. 2000 "ICA Multilingual Glossary" International Cartographic Association Commission on Planetary Cartography 1997. Retrieved March 8, 2003 from the World Wide Web:

Schultz, R.A. 1991. "Structural Development of Coprates Chasma and Western Ophir Planum, Valles Marineris Rift, Mars" Journal of Geophysical Research 96: 22,777-22,792.

Scott, E. D. 2000 "Sub-lithospheric 'Subduction' on Mars: Convective Removal of a Lithospheric Root. III: Syria Planum Region" Lunar and Planetary Science XXXI 2000. Retrieved February 10, 2003 from the World Wide Web:

Tanaka, K. L. 1997 "Origin of Valles Marineris and Noctis Labyrinthus, Mars, By Structurally Controlled Collapse and Erosion of Crustal Materials" Lunar and Planetary Science XXVIII 1997. Retrieved February 10, 2003 from the World Wide Web:

University of Arizona. 2002 "2001 Mars Odyssey Gamma Ray Spectrometer" Department of Planetary Sciences Lunar and Planetary Laboratory 2002. Retrieved March 23, 2003 from the World Wide Web:

Williams, D.R. 2001 "Mars Global Surveyor" National Space Science Data Center 2001. Retrieved March 23, 2003 from the World Wide Web:

Williams, D.R. 2002 "Viking Project Information" National Space Science Data Center 2002. Retrieved March 8, 2003 from the World Wide Web:

Williams, R.M., Phillips, R.J. and Malin, M.C. 1999 "Flow Rates and Duration Within Kasei Valles, Mars: Implications for the Formation of a Martian Ocean" NASA Goddard Space Flight Center's Labratory for Terrestrial Physics 2000. Retrieved February 10, 2003 from the World Wide Web: