Evidence for Persistent Flow and Aqueous Sedimentation on Early Mars

Science

Volumn 302, Issue 5652, 2003, Pages 1931-1934

  Malin, Michael C ^a   Edgett, Kenneth S ^a  

^a MALIN SPACE SCIENCE SYSTEMS   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

DEBRIS; FLOW OF FLUIDS; LANDFORMS; PLANETS;

CHANNELIZED FLOW;

SPACE RESEARCH;

FLOW; MARS; PLANETARY SURFACE; SEDIMENTATION;

ARTICLE; ASTRONOMY; ENVIRONMENT; FLOW; IMAGING SYSTEM; PRIORITY JOURNAL; SEDIMENTATION;

EXOBIOLOGY; EXTRATERRESTRIAL ENVIRONMENT; GEOLOGIC SEDIMENTS; MARS; TIME; WATER;

EID: 0346243807     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.1090544     Document Type: Article

References (36)

1. The debate over whether liquid water or some other fluid (including gas-fluidized dry flow) is responsible for erosional landforms on Mars (which include but are not limited to the martian outflow channels, valley networks, and gullies) is not specifically addressed in this work. Most geoscientists familiar with martian landforms agree that water is the most likely candidate fluid.
2. M. H. Carr, Water on Mars (Oxford Univ. Press, New York, 1996).
3. V. R. Baker et al., in Mars, H. H. Kieffer, B. M. Jakosky, C. W. Snyder, M. S. Matthews, Eds. (Univ. of Arizona Press, Tucson, AZ, 1992), pp. 493-522.
4. V. R. Baker, Nature 412, 228 (2001).
5. M. C. Malin et al., J. Geophys. Res. 97, 7699 (1992).
6. M. C. Malin, K. S. Edgett, J. Geophys. Res. 106, 23429 (2001).
7. MOC consists of three cameras: one high-resolution, narrow-angle telescope, and two wide-angle cameras for low-resolution global and regional monitoring. A description of the experiment can be found in (5); results from the first Mars year of MOC operations are presented in (6). As of 21 October 2003, 155,000 images have been acquired by the MOC, of which nearly half are at scales of 1.4 to 12 m/pixel. Approximately 3% of the planet's surface area has been imaged at these scales.
8. P. R. Christensen et al., Science 300, 2056 (2003).
9. In general, images of surfaces taken at other than visible wavelengths should be used cautiously for geomorphic interpretation because of characteristics that may arise from the behavior of surface materials at those wavelengths. A "bright" feature in an infrared (IR) image is hot, not light-toned; indeed, because dark-toned features on planetary surfaces will be warmer than light-toned features during daytime hours (all other factors being the same), an IR image would be anticorrelated with albedo. Nighttime IR images show mostly differences in how areas retain heat after sunset, and are thus a measure of thermal inertia (which depends on the density, heat capacity, and thermal conductivity of the materials). However, the primary factor affecting temperature during the day is the declivity and azimuth of the surface slope. Daytime IR images look for the most part like shaded relief representations of the surface, with slopes that receive direct solar heating being warmer, and therefore brighter, than slopes that face away from the Sun. Such images are nearly ideal for geomorphic studies.
10. D. E. Smith et al., J. Geophys. Res. 106, 236B9 (2001).
11. For the purposes of discussion, we define a fan as any accumulation of debris transported by fluvial processes under laterally constrained conditions and deposited in the absence of that constraint. In this broad definition, a delta would be a subclass of fan in which rapid deposition of fluvially transported material occurs as a subaerial flow attempts to maintain and adjust its base level upon entering a body of water. We acknowledge that this and many other classification schemes [e.g., (25-27)] fail to capture the range and complexity of depositional environments found on Earth, and that inferences derived from any of the atternative schemes used to classify terrestrial fans and deltas [for example, that subaerially deposited materials tend to have steeper longitudinal slopes than those deposited in lacustrine or marine environments (27)] may be of limited applicability to martian studies.
12. MOLA measured the mean relief of spots about 150 m in diameter spaced about 300 m along the MGS orbit ground track from March 1999 through June 2001. Ground track separation was not systematically controlled, hence coverage is not uniform. Laser altimetry for the region north and east of Holden Crater was converted to digital elevation models (DEMs) with the GMT software package (28, 29). Gridded and contoured products were produced using programs blockmedian, surface, and grdcontour with ground sampling increments of 48 m. The DEMs were converted to grayscale images and coregistered with MOC and THEMIS mosaics. Topographic profiles were located on the images and the corresponding relief values determined from the DEMs.
13. Inversion of relief occurs by denudation in cases where a material that once was at the floor of a channel or depression - either intrinsically more resistant to erosion (e.g., coarser grained) or rendered more resistant by cementation - forms an armor that protects underlying materials from removal (in volcanic settings on Earth, lava may flow down a preexisting valley or into a depression, thus forming an erosion-resistant unit that can likewise protect underlying materials, although this is not what occurred in Holden NE Crater). The surrounding less resistant materials are eroded away, and the former depression or channel floor then stands higher than surrounding terrain. Burial, exhumation, and inversion of relief are common attributes of martian geology (6), and numerous examples of inverted channels have been identified in MOC and THEMIS images; examples are located in northeast Arabia Terra near 32°N, 314°W (MOC images M15-00823, M20-00098, R09-00568, R09-01403, R09-02724, R09-03590, and R10-1078), southeast Arabia Terra near 11°N, 313°W (THEMIS image 101646005), southwest of Juventae Chasma near 4°5, 63.5°W (MOC images R07-00999, R08-02192, R09-00590, R09-01445, R09-02851, R09-03652, R09-04056, and R10-01114; THEMIS images V05558002 and V06307001), and south-central Arabia Terra near Z°N, 343.5°W (MOC images M10-03450 and R05-03450; THEMIS image V07257022). In addition, a fan-like pattern of exhumed, inverted channels exhibiting some of the cross-cutting and meander characteristics of the Holden NE Crater complex is located in Aeolis near 6.5°5, 209°W (MOC images FHA-00772, M00-02962, M02-02414, M03-02792, M20-01792, M21-00360, E05-02402, E13-01135, E18-00307, and R10-03569; THEMIS image V05588002). Relief-inverted paleochannels described in this paper and at the other martian locations listed above are strikingly similar to those found in central Saudi Arabia (30) and eastern Oman (31).
14. We use the term "progressive" to mean that emplacement occurred both in a specific, sequential manner (e.g., in which lower, smaller distributaries were covered by later, larger distributaries) and by advancing distally.
15. As summarized more than two decades ago by the Mars Channel Working Group (32) and further amplified by Carr and Malin (20) using MGS data, characteristics of martian landforms are often ambiguous on the subject of how much fluid was involved in their formation, how it moved, and over what time scales. At large scale, valley networks appear to have the ramified pattern of integrated, contributing fluid and debris transportation systems, but higher resolution views show these networks to lack the smaller, lower order tributaries seen in most terrestrial river systems that occur as a natural consequence of overland flow. They instead more closely resemble terrestrial systems in which groundwater-both in the form of wall seepage and as base flow into small streams substantially underfit to the valleys in which they occur - controls valley development by undermining and collapse of valley headwalls and side-walls. Such groundwater-related processes still require surface flow and a level of persistence necessary to break down and remove material derived from the walls, but the discharge rates and volumes can be substantially smaller and occasional rather than perennial.
16. J. Weihaupt, J. Geophys. Res. 79, 2073 (1974).
17. D. J. Milton, J. Geophys. Res. 78, 4037 (1973).
18. D. J. Milton, V. R. Baker, Icarus 23, 27 (1974).
19. R. P. Sharp, M. C. Nalin, Bull. Geol. Soc. Am. 86, 593 (1975).
20. M. H. Carr, M. C. Matin, Icarus 146, 366 (2000).
21. 3, and filling times of decades). These calculations simply show that the relations are internally consistent with similar relations seen on Earth, not necessarily that the situations are identical.
22. As part of our study, we targeted 158 locations identified by previous investigations [e.g., appendix B in (34)] as potential "alluvial fans" and "deltas" (34, 35) and more than 100 additional locations exhibiting similar topographic relations (valleys entering depressions). As of October 2003, some 200 MOC images covering approximately 80 locations had been acquired and inspected. All of these images show features quite different from those discussed in this work, generally falling into two categories. The most prevalent category is one in which the floors of the valley and crater are concordant, showing no discernible expression of deposition (e.g., MOC images E04-01284, E23-01302, and R02-00995). in these cases, alluvial deposits may exist but have been buried by some process that filled the crater, or may have once existed but have since been completely stripped away. In a relatively small number of cases (the second category), a discernible apron of material is seen at the point where the valley enters the crater. Although the aprons have some attributes of alluvial fans (they are conical in three-dimensional form, have longitudinal slopes ≥2° and convex transverse sections, and occur adjacent to high-standing relief), they have three characteristics that distinguish them from the fan described in this work. They consist of a, single (rather than multiple) lobe of material, they lack a radial (or distributary) pattern of conduits, and they display concentric steps in their surface's descent to the crater floor (e.g., MOC images E02-00508 and R02-00093). The concentric steps are unique to the aprons, as the adjacent crater walls do not display such forms (that is, the steps are not wave-cut terraces). In some cases, the volume of the apron appears to be equal to the volume of the valley (e.g., MOC images E05-02330, E09-00340, and E11-00948). These aprons appear to be the result of mass movements rather than fluvial processes, with the concentric steps resulting from successive surges of the material as it moved out of the valley or, more likely, as the expression of compressive stress in the material as it came to rest within the crater.
23. We use the term "rhythmically layered" to denote a sequence of tens to hundreds of repeated layers (or packages of layers too fine to resolve in MOC images) of essentially identical thickness and outcrop expression.
24. M. C. Malin, K. S. Edgett, Science 290, 1927 (2000).
25. W. E. Galloway, in Deltas, Models for Exploration, M. L. Broussard, Ed. (Houston Geological Society, Houston, TX, 1975), pp. 87-98.
26. W. Nemec, in Coarse-Grained Deltas, A. Colella, D. B. Prior, Eds., Intl. Assn. Sedimentol. Spec. Pub. 10, 3 (1990).
27. T. C. Blair, J. G. McPherson, J. Sediment. Res. A 64, 450 (1994).
28. GMT - The Generic Mapping Tools (http://gmt.soest. hawaii.edu).
29. P. Wessel, W. H. F. Smith, Eos 72, 441 (1991).
30. R. P. Miller, J. Geol. 45, 432 (1937).
31. J. Maizels, Palaeogeogr. Palaeoclimatol. Palaeoecol. 76, 241 (1990).
32. Mars Channel Working Group, Geol. Soc. Am. Bull. 94, 1035 (1983).
33. P. Komar, Icarus 37, 156 (1979).
34. N. A. Cabrol, E. A. Grin, Icarus 149, 291 (2001).
35. G. G. Ori, L. Marinangeli, A. Baliva, J. Geophys. Res. 105, 17629 (2000).
36. We thank R. A. MacRae for stimulating discussions, and R. M. E. Williams and V. R. Baker for their perceptive and insightful comments and suggestions that were instrumental in refining and focusing this paper. We acknowledge the contribution to this work made by the MGS/MOC and Mars Odyssey/THEMIS operations teams at Malin Space Science Systems, Arizona State University, the Jet Propulsion Laboratory (JPL), and Lockheed Martin Astronautics. Supported by JPL contract 959060 and Arizona State University contract 01-O81 (under JPL contract 1228404 and NASA prime contract task 10079).