Abstract– Six large iron meteorites have been discovered in the Meridiani Planum region of Mars by the Mars Exploration Rover Opportunity in a nearly 25 km-long traverse. Herein, we review and synthesize the available data to propose that the discovery and characteristics of the six meteorites could be explained as the result of their impact into a soft and wet surface, sometime during the Noachian or the Hesperian, subsequently to be exposed at the Martian surface through differential erosion. As recorded by its sediments and chemical deposits, Meridiani has been interpreted to have undergone a watery past, including a shallow sea, a playa, an environment of fluctuating ground water, and/or an icy landscape. Meteorites could have been encased upon impact and/or subsequently buried, and kept underground for a long time, shielded from the atmosphere. The meteorites apparently underwent significant chemical weathering due to aqueous alteration, as indicated by cavernous features that suggest differential acidic corrosion removing less resistant material and softer inclusions. During the Amazonian, the almost complete disappearance of surface water and desiccation of the landscape, followed by induration of the sediments and subsequent differential erosion and degradation of Meridiani sediments, including at least 10–80 m of deflation in the last 3–3.5 Gy, would have exposed the buried meteorites. We conclude that the iron meteorites support the hypothesis that Mars once had a denser atmosphere and considerable amounts of water and/or water ice at and/or near the surface.
At the time of this writing, six iron meteorites have been discovered on the surface of Mars by the Mars Exploration Rover (MER) Opportunity along its approximately 25 km-long traverse through Meridiani Planum (see e.g., Arvidson et al. 2011). The meteorites have been informally named “Heat Shield Rock” (HSR), “Block Island” (BI), “Shelter Island” (SI), “Mackinac Island” (MI), “Oileán Ruaidh” (OR), and “Ireland” (Ir) (Fig. 1). HSR was discovered in January 2005 (Arvidson and Squyres 2005) and given the approved official name “Meridiani Planum” after the location of its find (Connolly et al. 2006). BI, SI, and MI were found about 10 km to the south of HSR between July and October 2009 (Fleischer et al. 2010a; Ashley et al. 2011). OR and Ir were found about 3 km to the south of BI between September and October 2010 (Fig. 2). The meteorites are relatively large, with the long axes of HSR, BI, SI, MI, OR, and Ir nearing 31, 60, 52, 30, 50, and 30 cm, respectively. These sizes probably approximate their original postfall contours (Schröder et al. 2008; Ashley et al. 2011). The meteorites sit on friable sulfate-enriched materials partly blanketed by saltating sand particles (Ashley et al. 2011). Polygonal networks of cracks are abundant around the meteorites (Fig. 1), and probably relate to wetting and drying of surface sediments.
The MER Opportunity Alpha Particle X-ray Spectrometer (APXS) shows that the elemental compositions of HSR, BI, and SI are nearly identical, all being type IAB complex iron meteorites (Wasson and Kallemeyn 2002) based on their Ni, Ge, and Ga contents (Connolly et al. 2006; Fleischer et al. 2010a, 2011; Ashley et al. 2011). Compositional information, however, was not obtained from MER Opportunity’s scientific instruments for MI, OR, and Ir, although they have similar VIS-NIR spectral patterns and morphologies, suggesting that they are also iron meteorites (Fleischer et al. 2010a; Ashley et al. 2011). This information suggests that at least BI and SI, probably MI, OR, and Ir, and possibly HSR and some others yet to be discovered, may be pieces of a single (or 2 or 3) parent body that fragmented in the atmosphere before landing, although this possibility faces some problems (Chappelow and Golombek 2010; and also see the Aqueous Weathering and Astrobiological Implications section below). As the hypothesis that the iron meteorites are paired is an important subject, yet not affecting the results of this work, we leave this alternative as an open question.
Across the same area, at least four other rocks similar to mesosiderite silicate clasts have been investigated by Opportunity (Fleischer et al. 2010b). On the basis of their different stages of weathering (Schröder et al. 2010), these four rocks are probably stony-iron meteorites, although their origin is not definitively resolved. The stony meteorites are estimated to have fallen on Mars in more recent times than the iron meteorites, probably during the Amazonian (Fleischer et al. 2010b; Schröder et al. 2010), and therefore do not carry a record of chemical weathering that might be linked to an early wet Mars. The stony-iron meteorites will not affect the results presented in this article, because they are not genetically linked to the iron meteorites analyzed here, and they are significantly smaller than the iron meteorites.
Herein, we propose that the occurrence on Meridiani of the six iron meteorites can readily be explained as the result of their impact into a soft, unconsolidated and wet surface with subsequent burial, acid aqueous weathering, and ultimate differential erosion of once entraining materials, exposing them at the Martian surface, similar to the meteorites found on Earth in zones of significant wind deflation (Zolensky et al. 1990; Cassidy et al. 1992; Welten et al. 1997, 1999). We also discuss this hypothesis in the context of some alternative scenarios already proposed by others, as well as the implications of our model for the climatic and hydrologic histories of Mars.
Impact into a Soft and Wet Surface, Burial, and Ultimate Exhumation by Differential Erosion
We estimate that the six iron meteorites fell on Meridiani more than 3 Ga ago, based on three arguments. First, all six meteorites lack an associated nearby fresh impact crater of the size expected by their impact on the surface (see the Discussion of Alternative Hypotheses section). Second, calculations show that iron meteorite decay due to chemical weathering in the equatorial regions of Mars should have occurred on a multibillion-year time scale, even considering the continuous availability of water close to the meteorite in a “wet” early Mars scenario (Bland and Smith 2000). In contrast, in terrestrial polar regions, complete chemical weathering of meteorites occurs over a 105–106 year time scale (Scherer et al. 1997; Welten et al. 1997, 1999). Third, a collection of several dozens of stony meteorites of various sizes should be expected for every large iron meteorite found, as this is the observed ratio of stony meteorites versus iron meteorites on Earth (e.g., Bischoff 2001). However, a large population of stony meteorites is so far missing on the surface of Meridiani analyzed by Opportunity. As stony meteorites are less resistant against weathering than iron meteorites, it is likely that ancient stony meteorites contemporaneous to the six iron meteorites have been destroyed during the last 3 Ga. The stony meteorites observed today probably correspond to very recent events (Fleischer et al. 2010b; Schröder et al. 2010), independent from the ancient iron meteorites. Therefore, the fact that the MER Opportunity is not finding the population of stony meteorites expected to complement the six iron meteorites, is a robust argument supporting the hypothesis that the irons fell on Mars in pre-Amazonian times. Together, all this information suggests that the iron meteorites that rest on the Martian surface today are not the product of recent events, but probably, the remains of Noachian-Hesperian impactors. Also, Ashley et al. (2010) highlight that “little argues against their landing >3 Ga.”
We propose here that the iron meteorites found in Meridiani fell sometime during the Noachian or the Hesperian on a soft, unconsolidated and wet surface characterized by the presence of significant amounts of water, such as a shallow oceanic margin, a playa environment, a mud ocean, fluctuating ground water, and/or an iced landscape. A combination of some or all of these environments in different times and places on Meridiani is certainly a possibility with the presence of alternating amounts of water, including the potential evolution from those involving a greater amount of water to the less watery. The existence of a much denser atmosphere during early Mars when compared with the present-day atmosphere, probably produced by outgassing associated with the formation of Tharsis (Dohm et al. 2001; Phillips et al. 2001), could have contributed to decelerate the meteorites at low enough speed to survive impact against the soft and wet surface. Upon impact and/or by subsequent burial by water- and/or ice-enriched materials, the meteorites were encased underground for a long time, up to billions of years.
Subsequently, during the Amazonian, surface water almost completely disappeared, the landscape desiccated, and the dry sediments underwent induration. Differential erosion of Meridiani Planum sediments, including wind-driven deflation (Golombek et al. 2006), ultimately exposed the buried meteorites on the Martian surface. Currently, the dominant active process in Meridiani is eolian reworking of sediments by erosion, transport, and deposition. The broad plains of Meridiani commonly erode into stair-stepped outcrop patterns with some localized occurrences of massive outcrops. Around the perimeter of the continuous sediments are numerous outliers of buttes, mesa, and yardang-like formations, all of which suggest that the sediments were more extensive at the time of emplacement (Hynek et al. 2002). Erosion rates at Meridiani during post-Noachian times are 2–5 orders of magnitude lower than the slowest continental denudation rates on Earth, indicating that long-term action by water was not an important erosional agent (Golombek et al. 2006). Such erosion rates account for at least 10–80 m of deflation in the last 3–3.5 Gyr, including 1–10 m of erosion and redistribution of sand in the last 0.4 Gyr (Golombek et al. 2006). Denudation of several tens of meters may very well account for the uncovering of pre-Amazonian buried meteorites.
A clear implication of the sequence of events proposed here is that the materials that buried the meteorites mainly involved fine-grained sediment such as silt, sand, and/or pebbles, but not large clasts, because, in this case, subsequent deflation would have resulted in the formation of a rocky subaerial landscape, including possible rock gardens. This implication constrains the type of environment existing in Meridiani prior to the impact events. Deflation of once fine-grained, water-enriched material, or an ice body acting on fine-grained materials are the best candidates. The lateral and vertical extents of Meridiani are not constrained, but there is no reason to believe that they are much different from the sulfate-rich, fine-grained sediments of the Burns Formation, an outcrop exposing sedimentary rocks and forming a set of genetically related strata at the Opportunity landing site (Grotzinger et al. 2005; Ruff et al. 2005). Our hypothesis is summarized in Fig. 3.
Similar sequences of events are common on Earth. Large populations of meteorites are found on Earth in areas where wind deflation has removed ancient sand covers, or has differentially eroded thick sequences of friable sulfate-enriched materials, in cases resulting in deflation basins or blowouts that have been excavated upon a mantle of eolian coversands (Zolensky et al. 1990). A good example is the Chinle Formation of Triassic age, near Flagstaff, Arizona, where the more resistant rock materials such as chert and Petrified Forest material are often concentrated. The Chinle Formation is composed of mudstone as a primary rock type and sandstone as a secondary rock type, and also contains calcarenite, conglomerate, limestone, and siltstone as tertiary rock types (Blakey 1989). The formation records aqueous conditions, including tropical streams, lakes, and swamps. The friable materials, which outcrop along the northern margin of the younger San Francisco volcanic field lavas, was examined remotely by the MER science team during the final field test prior to the launch of the MER rovers (Anderson et al. 2006).
Aqueous Weathering and Astrobiological Implications
The chemistry and appearance of the meteorites (Fig. 4) support our interpretation of their impact into weakly indurated and wet materials, which includes at least intermittent interaction with groundwater and/or ice. All specimens are well exposed on the Martian surface, and the weathering state of the six meteorites is notably variable: MI and Ir are the most weathered meteorites and are severely hollowed, SI is more weathered than BI, BI is more weathered than OR, and OR is more weathered than HSR. This weathering difference may represent a range of residence times, burial depths, or environments (Ashley et al. 2011), arguing against a common origin from a single parent impactor. We suggest that the different weathering states of the meteorites would probably reflect their different residence times on the surface of Mars, including during an epoch when environmental conditions were warmer and wetter than today. The most altered meteorites probably fell on Mars early during the Noachian, when water and/or water ice was more abundant on the surface and subsurface, and the resulting chemical weathering was faster and more significant; whereas those iron meteorites with a lower degree of alteration probably fell at the end of the Hesperian, when liquid water was scarce, ice was prevalent, and the chemical weathering was slower and/or more difficult. In any case, based on degrees of pitting and rind development, the interaction of the meteorites with water is evident. Subsequently, during the Amazonian, low temperatures and hyperaridity would slow chemical reactions that can destroy the meteorites (Cassidy et al. 1992), and therefore the conditions for their preservation over time would be more favorable in such a colder and dryer environment.
The six meteorites have portions of their surfaces that are smooth, and they show ubiquitous hollows and pits that appear similar to regmaglypts, probably formed by ablation during the descent through the atmosphere (Schröder et al. 2008; Chappelow and Golombek 2010) and/or by subsequent aqueous interaction once on the surface (Fleischer et al. 2010a; Johnson et al. 2010; Ashley et al. 2011) or while encased underground. Cavernous features seem to indicate differential acidic corrosion removing less resistant material, suggesting chemical weathering produced by aqueous alteration removing softer inclusions (common to iron meteorites), such as troilite or cohenite (Fig. 5) (Fleischer et al. 2011). Chemical corrosion is also indicated by the shape and angularity of the cavernous features (Ashley et al. 2011). This morphology points to the interaction with acidic surface water, groundwater, and/or ice, although the interaction with just thin films of acidic water shaping BI’s cavernous pit has also been proposed (Ashley et al. 2011). As water has been mostly absent from the surface and the upper subsurface of equatorial Mars such as in Meridiani Planum during the Amazonian, the described chemical alteration probably occurred shortly following impact, when wet conditions were pervasive in the region. Therefore, the aqueous chemical alteration shown by the meteorites and the extreme arid conditions of the Amazonian Period also confirm our estimations of a “Martian age” of the meteorites (the time in which the meteorites have been on Mars) of over 3 Ga.
The meteorites are partially coated by a discontinuous cover of variable thickness, analyzed for the cases of HSR, BI, and SI. The coating is composed of iron oxides and oxyhydroxides (mainly hematite, but also maghemite, goethite, and lepidocrocite, among other secondary minerals; see Fleischer et al. 2010a, 2011; Johnson et al. 2010), and is not a remnant fusion crust (Fleischer et al. 2011). The chemistry of this secondary weathering coating is different from that of the substrate, and the same coating is observed in rocks at Meridiani. These observations suggest an interaction among the primary rock materials, which would include the six meteorites, and water and/or water ice, as the presence of the hydroxyl radical denotes water exposure, and the formation of oxyhydroxide minerals is favored in the presence of water (Gooding et al. 1992; Ashley et al., 2010). The coating is thin, less than 1 μm thick, an observation that has been proposed to indicate a low degree of chemical weathering through the influence of small quantities of water (Fleischer et al. 2001). Alternatively, an original thicker coating, produced by a significant degree of chemical weathering through the influence of important amounts of water and/or ice, could have been mostly stripped away by physical erosion as the meteorites were exhumed, leaving only the thin vestige observed today.
In this sense, if wind erosion occurred over a long period of time, ventifacts would be expected on the meteorites. Minor smooth areas can be identified on the surfaces (see Fig. 4), but no real ventifacts, which suggests that the iron meteorites might have been exhumed relatively recently. Eolian weathering could also be the cause of some pits and curved features in the iron meteorites, as regmaglypts can assist in eolian weathering because the hollows may offer a preferred path of mass removal to erosive forces such as eolian scouring (Ashley et al. 2011). Similar processes of eolian weathering have been reported in terrestrial meteorite analogs (Al-Kathiri et al. 2005).
Together, all this information allows us to suggest here that the six iron meteorites found by the MER Opportunity may be indicating the presence of significant quantities of water and/or ice at Meridiani at the time of impact. The demonstration that the iron-nickel alloy present in iron meteorites provides (1) a suitable source of energy for the active growth of iron-oxidizing microorganisms (González-Toril et al. 2005), and (2) a possible protection against the deleterious effects of solar UV radiation if the microorganisms are in direct contact with thin layers of meteorite material (Rettberg et al. 2004) gives an interesting perspective to the presence of these meteorites on the surface of Mars, especially considering their landing on wet environments. Iron meteorites could be motivating targets to search for signs of life on Mars, as they could be one of the latest (and maybe remaining) niches where a possible ancient Martian biosphere could have retreated as conditions on the surface became increasingly hostile for life.
Discussion of Alternative Hypotheses
Schröder et al. (2008) list three possible explanations for the presence of iron meteorites in Meridiani (1) meteorites are spalled-off fragments of an impactor; (2) meteorites fell in the last 0.4 Ga. and excavated craters were later eroded away; and (3) meteorites were decelerated to near free-fall velocities prior to impact with no resulting impact craters. None of these alternatives is incompatible with our hypothesis, as follows.
Regarding hypothesis 1, it is now accepted that observations of rounded shapes and smooth surfaces confirm that the meteorites are not spall fragments that fractured off the impactor during impact, because such fragments would have surfaces dominated by fracture planes, broken edges, and planar or angular features (Chappelow and Sharpton 2006; Chappelow and Golombek 2010). In addition, the meteorites do not show any signs of impact damage or deformation. Therefore, the six meteorites do not seem to be fragments of a larger impactor excavating nearby impact craters. In addition, iron meteorites can survive velocities in excess of 2.0 km s−1 (Hörz and Cintala 1984; Bland and Smith 2000), so the meteorites would not have broken after the impact, and thus would not show any planar or angular features, but rather remain intact, especially if the target was a soft and wet surface. However, over 3 Ga., there is sufficient time for aqueous and eolian erosion to smooth any original broken edge and irregularity into the rounded shapes observed today, in the same way that weathering/erosion has formed the pits and hollows. If this is correct, then the meteorites may be fractured fragments of larger impactors. In any case, this possibility is not incompatible with the hypothesis presented here, particularly if the target material was ice.
Concerning hypothesis 2, impact craters attributable to HSR, BI, SI, or MI as primary impactors have certainly not been observed, but the possibility that they have since been destroyed is arguable. For example, HSR can potentially excavate a crater of approximately 10 m in diameter and up to 3 m in depth (Grant et al. 2006). Craters in this size range have been repeatedly observed during Opportunity’s traverse along Meridiani, with variable degradation states ranging from fresh appearance to ghost craters (Golombek et al. 2006; Grant et al. 2006). This variation indicates that the crater formed by HSR should be somewhat visible today around the meteorite, and it is not. Therefore, the combined probability that all six meteorites fell in the last 0.4 Ga, excavated an impact crater and all six craters have been completely erased, is negligible. Regardless, similar to hypothesis 1, hypothesis 2 concerning crater destruction is not incompatible with our model, especially because a crater excavated in mud or in a shallow sea floor would have been easily destroyed soon after the impact.
Finally, considering hypothesis 3, Chappelow and Sharpton (2006) and Chappelow and Golombek (2010) analyze the possibility that iron meteorites could be landed recently without forming a primary crater and being destroyed or severely deformed. Only a small impact pit or structure near or under the meteorites would be expected, and meteorites could have ricocheted away from the point of landing to their current location. Chappelow and Sharpton (2006) identified the conditions for impact survival in the case of HSR, and Chappelow and Golombek (2010) did the same for BI. They conclude that such landings could have occurred even during a low-density Martian atmosphere similar to the one at present, although only quite rarely and in very specific and limiting conditions concerning initial mass, speed, and entry angle. Such events would account only for 0.08–0.007% of incoming iron meteoroids (Chappelow and Golombek 2010). The combined probability that this event has occurred in the six cases of the iron meteorites discovered by the MER Opportunity in less than a 25 km traverse is negligible. In fact, landing meteorites have not been described on Earth. Chappelow and Golombek (2010) also proposed that the iron meteorites could be the result of a single aerial fragmentation event that dropped a strewn field over the area, and this would not require several very low probability events to occur and could result in several meteorites and no craters. But, if meteorites had been landed, Chappelow and Golombek (2010) expected about 12 HSR-sized iron meteorites for each BI-sized one found, and almost as many HSR-sized for each SI-sized. After the last discovery of OR, another 12 HRS-sized meteorites should be expected. However, such an expected large population (>30) of HSR-sized meteorites balancing BI, SI, and OR is missing. As any big rock that differs at the Martian surface from the local friable, sulfate-enriched materials, such as the relatively large iron meteorites, would have been easily identified through Opportunity, it is a fair assumption that they simply are not there to be found. In this sense, every new large meteorite that the MER Opportunity may discover along its future traverse at Meridiani will make the landing hypothesis less likely. Nonetheless, the landing possibility is not incompatible with our hypothesis, especially because landing on a soft and wet target sometime in the past would encompass a less restrictive set of conditions than landing in the current dry layered rock at Meridiani, and therefore would be a less improbable phenomenon. Also, our model considers the option of transportation away from the point of impact by means of ice flow, and therefore our hypothesis is compatible with the possibility that the meteorites first touched ground in a location to be relocated later to another.
We have proposed here that the characteristics of the six iron meteorites found at Meridiani by the MER Opportunity could be explained as the result of their impact into a soft, unconsolidated, and wet surface sometime during the Noachian or the Hesperian. The meteorites would have been encased in materials upon impact and/or subsequent burial, followed by weathering, and ultimately exhumed at their current locations on the Martian surface due to differential erosion of the surrounding friable materials. The possibility exists that the meteorites were transported from the initial resting place to another locality by ice flow. We cannot rule out the possibility that the iron meteorites were produced by landing; or that they impacted, broke into fragments, and excavated a crater that has been ultimately erased. However, we find that the hypothesis that the meteorites fell on a soft and wet surface early in the history of Mars is a much more parsimonious explanation. Our hypothesis predicts that (1) iron meteorites should be relatively common in Meridiani; (2) more of them will be found by Opportunity during its continued traverse, even larger than the six discovered to date; and (3) a substantial population of iron meteorites may be buried under the surface in Meridiani Planum.
Acknowledgments— Comments and suggestions by G. Osinski, S. Kissin, and an anonymous reviewer have substantially contributed to improve this work.