Atmospheric conditions on early Mars and the missing layered carbonates



[1] Widespread, massive layered sediments, studied by surface rovers and Mars-orbiting spectrometers, are found to be rich in sulfates. No similarly massive carbonates have been detected. We present the results of coupled atmospheric and geochemical calculations of the formation of the sulfate-rich Meridiani sediments, and offer an explanation for why extensive layered carbonates are not found there or elsewhere on Mars. Large scale volcanism from the build-up of Tharsis during the late Noachian would have injected large amounts of SO2 into the atmosphere. Efficient photochemical conversion of SO2 to H2SO4 would have caused widespread sulfuric acid/water clouds, similar to those seen on Venus today. Precipitation from these clouds and acidification of surface water would have sustained a thick, warm CO2 atmosphere via carbonate inhibition. Such an atmosphere could have been subjected to loss to space via impact erosion and sputtering. Once atmospheric SO2 gas production dropped and waters become more alkaline, the remaining CO2 in the atmosphere collapsed to form poorly consolidated carbonate patinas on rock surfaces and in open fractures.

1. Introduction

[2] Massive layered sediments such as those investigated by the Opportunity rover at Meridiani Planum leave little doubt that liquid water has chemically altered the surface of Mars. Stacks of mixed sulfate-siliciclastic layers, 7 m of which are exposed near the rim of Endurance crater, appear to be sands that were alternately saturated with water and cemented by its products. The upper beds were formed by the wave action of liquid surface water, most likely in an interdune, transient pond environment [Squyres et al., 2004]. MOC and THEMIS images show that the exposed outcrops at Meridiani are merely a portion of light-toned sediments that are up to 800 m thick and several hundred thousand km2 in extent [Hynek et al., 2002]. The OMEGA instrument on board the Mars Express spacecraft has identified at least some of these layers as evaporitic salts [Gendrin et al., 2005]. In particular, OMEGA has imaged spectacular layered sulfates in the Interior Layered Deposits (ILDs) of the Valles Marineris system [Bibring et al., 2005]. Near-IR spectral analysis shows a remarkable decrease in the hydration state of the sulfate as a function of altitude. The finely and regularly spaced layered deposits seen in the ILDs and throughout Meridiani were probably laid down in lacustrian environments, as evaporitic sediments [Squyres et al., 2004], although precipitation in a subsurface environment is also a possibility. While finding exposed sulfate-rich sediments on Mars is not a huge surprise, the attendant lack of evidence for carbonate deposits remains a mystery. Fundamentally, a body of water in contact with a carbon dioxide atmosphere should have laid down massive carbonates, as happened on Earth during the Archean. So where are the water-lain carbonates on Mars?

[3] We used a simple photochemistry model, coupled with a set of equations for SO2, CO2 and basalt solubility and reactivity, to establish a reasonable set of constraints on the nature and composition of the early Martian atmosphere, its role in maintaining the chemistry of Martian water, and why carbonates never formed massive layered deposits on Mars.

2. Sulfate Deposits on Mars

[4] The chemistry of Mars' atmosphere and water during the time when the Meridiani sediments were laid down can to some extent be deduced from the chemistry we see in the Meridiani rocks and elsewhere. The layered deposits at Meridiani are approximately 40% sulfates, mixed with siliciclastic particles and 10% hematite [Clark et al., 2005]. Sulfates are chiefly Mg and Ca, but in some places at Meridiani ferric sulfates are seen. It is likely that these formed by diagenesis; in this case the exchange of cations Ca and Mg for Fe. The OMEGA instrument aboard the Mars Express spacecraft has detected sulfates in numerous regions on Mars, including massive deposits in the center of Valles Marineris [Gendrin et al., 2005] and sulfate-flavored dunes encircling the north pole [Langevin et al., 2005].

[5] In none of these locations are layered sediments of carbonate found. On the Earth, carbonate sediments sit on the vast regions of the continental margins. Massive sulfates are seen, too, often in association with carbonates [Garrels and Christ, 1965]. The carbonates are the solidified remains of Earth's earliest atmosphere. They contain a massive 60 bar CO2 atmosphere [Holland, 1978]—roughly the atmosphere of Venus.

[6] The incontrovertible conclusion from MER is that surface water did indeed exist on early Mars, and chemical sediments attest to substantial interactions between bodies of water and the atmosphere. Geomorphologic evidence for fluvial deposits such alluvial fans left over from ancient rivers at the surface are a further link to a hydrologic cycle involving the surface and atmosphere [Moore et al., 2003; Malin and Edgett, 2003; Moore and Howard, 2005]. These features point to sustained rain or snowfall, runoff and ponding, sometime around the transition from the Noachian to the Hesperian. Under such conditions, the primary result on the Earth is the formation of carbonates. The oceans today are approximately saturated with carbonate, so sedimentary deposits can form easily in bodies of water that experience restricted inflow.

[7] The most likely situation for Mars, however, was that its waters were acidic [Moore, 2004; Fairen et al., 2004; Bibring et al., 2006; Greenwood and Blake, 2006] If large enough concentrations of sulfuric acid were present in Martian waters, sulfates would precipitate, but not carbonates. Sufficient sulfuric acid would be necessary to buffer the pH against alkalinity induced by dissolving metallic cations and the attendant OH out of basalt. If this were the case, CO2 from the atmosphere would remain dissolved in water but would not precipitate – the atmosphere would be protected from collapse.

3. Evolution of Atmospheric CO2

[8] It has long been speculated that Mars once had a large atmospheric reservoir of CO2 very early in its history, which provided greenhouse warming and potentially allowed liquid water to exist at least seasonally over a significant portion of the planet [Pollack, 1979; Squyres and Kasting, 1994]. It was also part of the Viking era canon that this CO2 would combine with altered basaltic minerals in the presence of water, drawing the CO2 reservoir down. The fact that the present Martian atmospheric surface pressure sits at water's triple point has been taken as evidence that this had indeed taken place [Kahn, 1985]. Hence missions carrying spectrometers and other mineral detectors were flown over the last decade in the anticipation that large water-lain carbonate beds would be found. The operative assumption was that carbonates derived from the atmospheric CO2 would be deposited in standing water. The subsequent failure to find carbonate beds on Mars led to doubt that liquid water had ever persisted for significant periods there [Bandfield et al., 2003]. Then the discovery of layered, water lain outcrops of sulfates in 2004 [Squyres et al., 2004; Gendrin et al., 2005] required a reconsideration of the role of CO2 in Martian climate evolution.

[9] The scenario that emerges from the Mars rover and spectrometer data is that Mars must have been warm enough for liquid water to be stable at the surface (even under a faint young Sun), and that sulfur photo- and geochemistry propped up a massive CO2/H2O atmosphere with an enhanced greenhouse. Some cloud scattering greenhouse would probably be necessary [Forget and Pierrehumbert, 1997], or the existence of methane in Mars' early atmosphere [Brown and Kasting, 1993]. This more massive atmosphere could have then been subjected to loss via impact erosion and solar wind pick-up-ion sputtering [Brain and Jakosky, 1998]. Stable isotope fractionation measurements [Jakosky and Jones, 1997] support the contention that 90% of Mars' CO2 could have been lost to space via the latter process. A primary objective of the upcoming 2011 Mars Scout mission is to greatly increase the understanding of atmospheric volatile loss processes, so questions of the ultimate fate of Mars' early CO2 inventory may finally be resolved. Atmospheric collapse of the remaining atmosphere ensued when volcanism waned. Indeed, data from the Thermal Emission Spectrometer (TES) on the Mars Global Surveyor spacecraft indicated that modern Martian dust, derived from weathered rock and other sources, could contain magnesite (MgCO3) at the level of between 2 and 5% by weight [Bandfield et al., 2003]. However, recent OMEGA and Mini-TES data show no evidence for carbonates in the Martian dust, so it is possible that the total inventory of carbonates in the Martian fines is small. Along with carbonates deposited in fractures and interstices, such as those seen in the Mars meteorite ALH84001 [McSween and Harvey, 1998], however, a significant amount of CO2 may be still sequestered in these materials [Bandfield et al., 2003].

4. Volcanism and the Noachian Atmosphere

[10] With the loss of atmospheric CO2 to space and/or surface reservoirs and an earlier, more volcanically active early Mars, it is likely that the early Martian atmosphere contained higher levels of CO2 and volcanic gases. The primary gaseous emissions from basaltic volcanism are H2O, CO2, and SO2, with lesser amounts of H2S, HCl, and HF [Kaula et al., 1981].

[11] High rates of volcanism, and in particular the formation of Tharsis, could have driven large quantities of sulfur gases into the atmosphere. As pointed out by Settle [Settle, 1979], the early Mars atmosphere would have been at least as wet as the Earth's stratosphere. Odd-hydrogen species with the help of solar photons would have efficiently converted S gases to H2SO4 aerosols. The most likely fate of SO2 in the early Mars atmosphere was:

equation image

with abundant OH supplied by the photolysis of H2O. This happens to a small extent in the Earth's stratosphere, and is also responsible for the massive sulfuric acid clouds on Venus. Clouds inhabit the region of Venus' atmosphere from 50 mbar, 235 K to 1 bar, 350 K. Extensive, sometimes global hazes extend to both lower and higher temperatures. On Venus, sulfuric acid rain evaporates and then thermally decomposes in the deep atmosphere. On Noachian Mars, whether warm and wet or cold and humid, photochemical sulfuric acid clouds would rain onto the surface. Acid rain during the Noachian, at least on some parts of the planet, was probably the agent that acidified Mars' waters. Settle [1979] calculated that the lifetime of SO2 in the early Martian atmosphere would have been about 10 years. The rate of H2SO4 input to Martian waters and the equilibrium abundance of SO2 in Mars' early atmosphere depend upon how much S was being injected into the atmosphere during the Noachian. This in turn depends upon the concentration of S in Noachian magmas and on the rate of volcanism.

[12] S is often found at abundances close to saturation in terrestrial MORBs [Jambon, 1994]. These basalts generally have between 1500 to 3000 ppm S. Fe increases the solubility of S in magmas, while O decreases it. The higher uncompressed density of Mars' mantle relative to Earth's has been attributed to a higher Fe content [McGetchin and Smyth, 1978]. Because the Mars crust appears to be higher in Fe and perhaps slightly more oxidized [Haggerty, 1978], Martian magmas may be expected to have at least as much S as terrestrial MORBs. However, if the early Martian mantle was more reducing, magmas could have had a good deal more S than 3000 ppm.

[13] From a careful analysis of the MGS MOLA and gravity data, Phillips et al. [2001] concluded that Tharsis appears to be a huge igneous construct, emplaced during the late Noachian. Lithospheric warping associated with its construction may have focused a good deal of the fluvial activity during that time. They estimated that it required about 300 million km3 of lava to build it over the approximately 300 My of the late Noachian. Although Tharsis must have been built by numerous events separated in time, some large and many small, the average volcanic flux from Tharsis during the Late Noachian must have been about 1 km3/yr.

[14] The above considerations permit an estimate of both the globally averaged H2SO4 precipitation rate during the Noachian, and the steady-state abundance of atmospheric SO2. If Rv is the volume rate of lava production, ρ the density of the magma, XmSO2 the mass mixing ratio of SO2 in the lava, and Rm the radius of Mars, then the column production rate of atmospheric SO2 during the Noachian would have been:

equation image

The column loss rate of SO2, LrSO2, can be calculated from its lifetime against photolysis in the atmosphere, τ:

equation image

where NSO2 is the atmospheric column abundance of SO2. Settle [1979] estimated the photolytic lifetime of SO2 on early Mars, based on careful extrapolations from terrestrial upper atmospheric S chemistry, would have been very short; about 10 years. The equation describing the evolution of SO2 atmospheric column abundance is therefore:

equation image


equation image

As t = > ∞, NSO2(t) goes to its steady-state value, N*SO2, so that:

equation image

[15] Because of the short lifetime of SO2 in the early Martian atmosphere, steady state SO2 levels would have been quite low; on the order of 0.03 μbar, or about 30 ppbv in a 1 bar CO2 atmosphere. Large excursions in atmospheric SO2 due to short intense volcanic activity would have been possible, but would not have sustained high levels for more than about a decade. This result is significant because it implies that arguments that invoke large quantities of atmospheric SO2 as a greenhouse gas [Postawko and Kuhn, 1986; Johnson et al., 2006] or as a middle-atmosphere warming agent to prevent CO2 cloud formation [Yung et al., 1997] would not apply to sustained conditions.

[16] The average rate of S injection into the atmosphere during the Late Noachian, from equation (2) is estimated to be about 6 million tons per year. This is approximately 1800 Gt of S over the 300 My of the Late Noachian. In an atmosphere a few 10's mbar or more sulfuric acid/water clouds are thermodynamically favored. The Venus atmosphere and Earth's stratosphere both host these kind of clouds, although convection induced by solar absorption at the ground would have been necessary for rainfall on Noachian Mars. Most of the sulfur would have cycled through massive sulfuric acid clouds and rained out into standing bodies of water, where it precipitated as Ca and Mg sulfates sufficient to cover the planet in a global layer 10 m thick. After mixing with basaltic aeolian sediments, burial and cementation, these became the layered sulfates of Meridiani and other deposits like them elsewhere on Mars.

[17] In order to understand how Martian waters could have formed the chemical sediments found at Meridiani, we modeled the precipitation of salts from water that is acidified by sulfuric acid rain and is in contact with mafic basalts and an early Mars atmosphere. Based on what was measured by APXS, mini-TES, and the Mossbauer spectrometer on the Opportunity rover at Endurance and Eagle craters, the two most prominent sulfate cations are Mg2+ and Ca2+. Since the water is in contact with CO2 of an indeterminate partial pressure, we know that H2CO3, H2CO3 and CO32+ are also in the water. Since we are investigating the precipitation of CO32+, and SO42− this will happen with the available cations, mostly Mg2+ and Ca2+. For the basalt, we consider the dissolution of fayalite and ferrosilite into ferrous iron and SiO2 (Figure 1).

Figure 1.

Geochemical reactions used to predict carbonate and sulfate stability at the surface of Mars. Equilibrium pH and ion abundances were determined by solving this system of equations using the LLNL thermodynamic data [Delany and Lundeen, 1990], and Debye-Huckel theory for activity coefficients.

[18] Precipitation must have occurred at the bottom of lakes or seas through three processes: Acid rainfall, evaporation, and freezing followed by sublimation. We can estimate what the average pH of these surface waters were by noting what conditions were necessary to begin the precipitation of carbonate. The stability of carbonates in the ancient aqueous environment of Noachian Mars is shown in Figure 2. For water in equilibrium with basalt and salts of Ca and Mg, the pH depends upon the partial pressure of CO2. Under present atmospheric conditions, such water would have a pH of about 7.7, owing to the alkaline nature of basalt. Carbonates are stable under present surface conditions. Under a one-bar CO2 atmosphere, water with these geochemical properties would have a pH of about 6.7, and carbonates would be stable. Sulfuric acid atmospheric precipitation adds sulfur, but no cations, and lowers the pH of surface waters. With the present 7 mbar of atmospheric CO2, water with an added molarity of 0.01 sulfuric acid and in contact with basalt would have a pH of 5. Carbonates are not stable under these conditions, and would rapidly decompose to CO2. Sulfate precipitates with the available cations, Ca2+ and Mg2+. Under one bar of CO2, water would have a pH of 4.5, and could persist on the ground for extended periods due to an enhanced greenhouse. Again, carbonates are not stable and sulfate precipitates out with Ca and Mg.

Figure 2.

pH of water in contact with basalt and salts of S and C, as functions of atmospheric CO2 pressure and H2SO4 molarity. Grey shaded contours show pH, while the stability field for carbonates under these conditions is depicted in the orange shaded region. With a one-bar CO2 atmosphere, carbonates are unstable with 1 μmolar of H2SO4 or more. This corresponds to solutions with a pH of 6.2 or less. At 10 mbar CO2, carbonates are unstable in water with pH < 7.7 (0.1 μmolar H2SO4 added).

5. Conclusions

[19] During the late Noachian (Figure 3), widespread volcanism, such as that associated with the formation of the Tharsis igneous province, most likely injected large amounts of SO2 into the atmosphere of Mars. Efficient photochemical conversion of SO2 to H2SO4 would have caused widespread sulfuric acid/water clouds, similar to those seen on Venus today. Precipitation from these clouds and acidification of standing bodies of water explains the chemistry and extent of the massive layered sulfate-rich sediments seen at Meridiani and elsewhere. CO2 would be inhibited from precipitating out of early Martian seas as long as liquid water remained at a pH of 7 or lower. Sulfuric acid rain dissolved in bodies of water would rapidly combine with altered basaltic minerals to form layered water lain deposits such as those seen at Meridiani. Bibring et al. [2006] has interpreted the stunning OMEGA discoveries of phyllosilicates as evidence that the earliest waters on Mars were not acidic, and that sulfates were deposited later. However, recent, more detailed geological interpretations of the Mawrth Vallis phyllosilicates suggest that they have been formed as drape deposits, post dating the formation of Mawrth Vallis [Howard and Moore, 2007]. As volcanism declined in the early Hesperian and the majority of the Tharsis build-up was completed, water would have become more alkaline, and it began drawing down the atmosphere at the same time as some phyllosilicate minerals were being formed. Colder conditions would ensue as the greenhouse effect declined. Any bodies of water on Mars would freeze and either sublimate or be buried by the large amount of sediments that had been produced during the Late Noachian. Since this all happened at the end of a time of hydrologic activity, carbonates would most likely have formed in an unconsolidated form on top of the km of sediments deposited during the acidic phase. Incorporated into the most mobile unit of the Martian geology, the fines, carbonates remain dispersed but plentiful, but never in the form of layered aqueous sediments like the earlier sulfates.

Figure 3.

Atmospheric and aqueous conditions on early Mars that probably led to the existence of massive sulfate rich layers, but no water-lain carbonates. During the volcanic formation of Tharsis during the Late Noachian, large amounts of CO2, H2O, and SO2 were outgassed. SO2 was rapidly oxidized to H2SO4 in the humid atmosphere via photochemical reactions very similar to what happens on Venus today. Sulfuric acid rain from the photochemically produced clouds acidified standing bodies on Mars, allowing sulfates, but not carbonates, to precipitate. Details of these reactions have been studied in the lab, in a Mars simulation facility [Bullock et al., 2004; Bullock and Moore, 2004]. As long as volcanism allowed the production of sulfates and kept surface water acidic, carbonates could not form and the atmosphere was propped up. As volcanism waned, small amounts of carbonates formed in fractures and as a globally extensive patina on the surface of rocks, where it was subjected to erosion and incorporation into dust.


[20] The present work has benefited greatly from conversations with K. Zahnle, D. Catling, and R. Haberle. It was supported by a NASA Planetary Geology and Geophysics grant to J.M.M., and by a NASA Planetary Atmospheres grant to M.A.B.