Early Mars The Lake Planet

10.1.1 Water as a Prerequisite

Whether on Mars or on Earth, ancient lakes are Rosetta stones and key paleoenvironments to explore. Their waterlain sediments retain the record of seasonal and orbital changes, the variations in hydrological regimes of the channels, and the main hydrogeological events that punctuated the history of their catchment areas. Lacustrine environments are of the utmost importance from an astrobiological perspective as well [7, 8, 16]: Terrestrial lakes host abundant and diversified biota. They facilitate its inception and expansion through complex food webs. Lacustrine life follows the temporal fluctuations of the environment through climate and hydrological changes, which in time lead to pathways of evolution or

N.A. Cabrol and E.A. Grin, Ancient and Recent Lakes on Mars. In: Water on Mars and Life, Tetsuya Tokano (ed.), Adv. Astrobiol. Biogeophys., pp. 235-259 (2005)

springerlink.com © Springer-Verlag Berlin Heidelberg 2005

extinction. Subtle or dramatic modifications in the spatial distribution of life can occur at very small-scale (cm to m) within the margins of a lake triggered by the variations of water influx (e.g., clusters of hydrothermal springs of various temperatures, water salinity and chemistry). Lakes are equally unique in their ability to preserve the record of life: when organisms die, they accumulate on lakebeds and are covered under regularly deposited sediments. Rapid burial allows the generation of anoxic conditions favoring the formation of fossils by preventing oxidation and decay.

It is not known yet if life appeared on Mars and left a fossil record [32, 33] but past and current missions are providing supporting evidence that ponding over extended geological periods might have happened, generating both perennial and ephemeral lakes [9, 10, 12]. Commonly, the term lake refers to water as being the ponding fluid for terrestrial features, while "lava" lakes specifically designate a fluid from volcanic origin. However, no unambiguous evidence exists so far about the nature of the fluid involved in the formation of flow features on Mars. If we search for life as we know it, water is a prerequisite. Cold temperatures, thin atmosphere, abundance of CO2, and spatial distribution of recent gullies in an environment that does not allow water to flow at the surface anymore, have led to suggest that CO2 [50, 72], exotic flows (e.g., cryovolcanism and clathrates [36] -although contested [59]) and not H2O [66, 69] were responsible for carving recent flow features. Those would generate ponds, if any, of a very different nature than their terrestrial counterparts.

Some go farther. Recent studies proposed that CO2 is the more likely fluid not only for the gullies but for all Martian flow features (valley networks, outflows, gullies), whether recent or ancient [49]. Highly debated, this model encounters many difficulties. It lacks an explanation of how large CO2reservoirs could have accumulated and subsisted near the surface for thousands of years under unstable conditions (e.g., magmatic activity and impact-induced tectonic) especially in warmer sub-tropical to mid-latitudes (30°S-45°S) where gullies are also observed in abundance [66]. Moreover, this model is not consistent with the morphological observations and mineralogical measurements performed recently by the MOC and TES instruments. Although conditions at sub-polar latitudes would allow liquid CO2 to be theoretically stable at depths corresponding to the location of the gully sources, the gully morphology does not support this hypothesis. CO2 is liquid at -56.4°C and 5.1 bars. It is potentially stable at depth in its liquid phase but would become instantaneously unstable at contact with the surface where the pressure of 6 mbars is ~1000 times lower than required for its liquid phase. The outcome of surface exposure would be at best a sudden outburst followed by immediate sublimation. Gullies are characterized by terraced ravines, debris aprons, and channels narrowing from upstream to downstream suggesting a regular evaporation consistent with that of water. The presence of some percentage of briny material in the water would have allowed flow at freezing temperatures [62, 90].

The hematite deposits discovered by TES [23] may be additional evidence for past water activity. Large-scale hydrothermal systems (magma and water/ice interactions) or precipitation in a shallow lake were suggested for its formation. Both hypotheses involve water at some time in the Martian past and cannot be explained by CO2 processes. The observation of massflows [2, 13, 15, 55] with morphologies inconsistent with CO2-driven processes at latitudes where currently ground ice is stable near the surface [76], and in an environmental domain shared with the gullies, is a very strong support of the H2O-based hypothesis. Mars Odyssey (MO) has cast an even larger shadow on the CO2 model by delivering a global epithermal neutron map of Mars [34]. This map shows the distribution of hydrogen abundance as ice and hydrated minerals at global scale and is consistent with a large reservoir of H2O ice even today in the very near subsurface. From about 55° latitude to the poles, the Odyssey data show an average of 50% water by mass. Closer to the equator, this value is 2-10% beneath several centimeters of dry soil. See Chap. 5 by Mitrofanov for more details. Moreover, climatic models indicate that liquid water can form in localized transient events today, although it would not be in equilibrium with the environment [44]. The following sections, therefore, discuss lakes on Mars assuming that the most likely fluid associated with their formation is water.

10.1.2 Abundance and Diversity

The generation, abundance and distribution of lakes on early Mars would directly result from the initial physical and climatic conditions of the planet during the Noachian era (e.g., warm and wet vs. cold and dry) [86]. Far from being resolved, the debate about these early conditions is currently being revived in light of MGS and MO new data.

The abundance of valley networks and channels is consistent with Mars being a water-rich planet in its early times [20]. The distribution of hydrogen abundance as a current reservoir of ice and hydrated minerals shown by MO could indicate that Mars still has some significant water/ice resources and could have started with a H2O reservoir proportionally larger than the Earth's. The morphology of the majority of valley networks and channels [65] and the lack of large carbonate deposits [3] are also consistent with a planet which was never much warmer than at the present time. Although Mars was likely warmer during the Noachian because of a thicker CO2 atmosphere [43, 57], possible greenhouse effects, and an increased internal heat flux, current data may be showing that "warmer" might define at best a periglacial climate. Morphological, geological, and mineralogical evidence point to an early cold and wet Mars, with enough moisture in the cold atmosphere to generate snow precipitation [53]. Impacts have been presented as possible agents for local rainfalls forming hydrographic systems [84], although the drainage maturity of Martian valley networks provides little support to this hypothesis [11]. In a cold and wet scenario, lakes could still have been generated through various processes, including seasonal melt, subglacial flow, and groundwater circulation locally associated with hydrothermal activity. They formed in a variety of physical and geographical environments comparable to their counterparts on Earth. The combination of both climatic, geographical, and physical conditions would have led to a diversity of types, whose classification is likely as complex as that of terrestrial basins. Lakes on Mars were neither accidental occurrences nor rare [9, 10, 29, 31, 37, 74, 83]. Because Viking data lacked resolution and altimetric precision, the first attempt at identifying and classifying Martian lakes [9] used the very confined limits of structural impact crater basins as identification criteria, leading to define a physical classification of open, closed, and lake-chain basins. About 200 impact crater lakes were identified, mostly Hesperian, with some occurrences in the Amazonian [10]. Since, MOC images and the MOLA altimetry have finally made it possible to identify topographic basins as well. Much smaller and older lakes are now accessible and their number is, as expected, orders of magnitude more elevated than perceived with Viking. For instance, the Ma'adim Vallis watershed region alone counts at least as many lakes, ponds, and larger basins (topographic and impact-related) as the total number identified at global scale with Viking data [63]. High-resolution allows access to smaller surface areas of eroded geological units (e.g., Noachian) and ultimately will lead to the reassessment of the timing of the maximum peak of lacustrine activity. As unravelled by MGS and MO, early Mars could be defined as the "Lake Planet". Lacustrine basins of varied dimensions and types were numerous and widely spread over the entire globe, possibly contemporaneous with the largest of all basins: the northern ocean.

The diversity of early Mars lacustrine activity is best characterized by the many methods that can be used to define it. Except for the biological activity, which often defines lakes on Earth but still has to be demonstrated on Mars, terrestrial criteria could be valid classifiers, with possibly the exception of plate-tectonic induced basins which may or may not be applicable to early Mars [6]. MGS and MO data show the potential for a much more diverse classification than Viking allowed, with classifiers which could encompass: the origin of the water source, the hydrologic regime of the systems supplying the lakes, associated climatic conditions; the origin of the topography of the basin itself (e.g., tectonic, structural, volcanic); its dynamics (open, closed, lake-chain); and its relative lifetime (perennial, ephemeral). At a high level, lakes can be characterized using only one of these criteria (e.g., volcanic) or they could be more precisely defined using a suite of criteria, such as a perennial, closed, impact crater lake. Some criteria directly result from data interpretation: they are inductive classifiers (e.g., closed, open, impact crater, tectonic basins) and should be used as the foundation for the classification of Martian lakes as they rely on observation. Deductive classifiers result from analytical and logical processes that are not necessarily directly supported by observation. For instance, the presence of a large paleobasin showing evidence of shorelines, terraces, and a delta may lead to the assessment that the basin was occupied by a perennial lake. Although normally an inductive classifier on Earth, perennial is deductive on Mars. Because it comes in this particular case as a deduction from a series of direct observations, it can be defined as a strong hypothesis and be used as a classifier. The more the deductive process relies on pure logic and departs from observation, the more the end result will lean toward speculation which should never be used for classification. For instance, "brackish" cannot be used at this stage to classify Martian paleolakes, although intuitively one can assume that brackish lakes were common [24, 58]. The situation is complicated on Mars by the fact that ground-truth is rare (surface missions). Thus, even direct observations (especially image interpretation) rely most of the time on analogy with Earth. In all fairness, "direct" classifiers should be considered hypotheses. Therefore, the following classification proposed in Table 10.1, revised compared to the Viking-based one [9] uses classifiers resulting from strong analogies and strong hypotheses.

Table 10.1. Classification of Martian Paleolakes





Snowfall ' Glacial Melt1"""'

One major water/sediment input peak during melt season. Latitude/altitude dependent.

Lakes at the front of glaciers. Maximum peak at melt season.


Large-scale thawing of ice-rich permafrost


Rainfall (if any)

Limited to warmer latitudes. Seasonal peak(s) depends on geographical location. Role of asteroid impacts?


Volcanic/Hydrothermal. Input follows activation of volcanic centers.

Polar Basal Melting. Circulation varies in time with location and extent of polar caps, and depends on pore space availability.

Regional/Local subsurface circulation. Lakes and ponds supplied by local groundwater reservoir (e.g., residual perched aquifers).



Formed by dissolution of soluble rock by percolating water. See also permafrost for cryokarstic lakes.

Structural /Volcanic


Compression or extension of the Martian crust traps water circulating in surface and subsurface.


Lakes in craters, maars, cinder cones, calderas, and collapsed lava tubes

Impact Craters

Lakes in impact basins



Lakes in local to regional low topography (e.g., inside valleys, between high-relief).



Lakes can be open, closed, linked in chains



Perennial or ephemeral (relative lifetime)




1 Separation between glacial melt and snowfall: glacier morphologies are observed in MOC images. If glaciers, they resulted from snow accumulation but glacial melt can still occur while glaciers are receding at the end of wet climate episodes.

1 Separation between glacial melt and snowfall: glacier morphologies are observed in MOC images. If glaciers, they resulted from snow accumulation but glacial melt can still occur while glaciers are receding at the end of wet climate episodes.

To date, although suggested [35] no evidence for evaporite or large carbonate deposits has been detected. This may reflect paleoclimate conditions [3] or result from surface global dust cover or resolution and specification of the instruments [60]. However, if/when detected, mineralogical clues will be the very first true inductive classifiers. They will give deeper insights into Martian lacustrine evolution. They will broaden the classification to mineralogy, chemistry (e.g., fresh, brackish), physics (e.g., stratification, fractioning), and will give a better assessment of a lake's lifetime. The Mars Exploration Rover (MER) Spirit should provide for the first time such a mineralogical ground-truth while testing a paleolake hypothesis in Gusev crater [18].

0 0

Post a comment