Planning for Mars Returned Sample Science: Final Report of the MSR End-to-End International Science Analysis Group (E2E-iSAG)

1.1. A proposed MSR implementation architecture

The currently proposed architecture for the MSR Campaign entails four primary elements: a rover mission to collect and cache the samples; a lander mission (MSR-L) to transfer that cache to orbit; an orbiter mission (MSR-O) to retrieve the orbiting cache and return it to Earth; and a Mars Returned Sample Handling (MRSH) project that would encompass all the functionalities necessary to receive the samples (in a special facility referred to as the Sample Receiving Facility) and provide access to the laboratory infrastructure necessary to carry out the scientific investigation of the samples. Note that the first mission in this sequence (the mid-range rover) would have compelling in situ objectives sufficient to justify it as a stand-alone mission, and the decision to fly the 2018 rover would not automatically mean commitment to fly the remaining elements of MSR. Also, the length of time the cache would remain at the surface may be indefinite (see NRC, 2011).

The campaign would entail three launches, as depicted in Fig. 1 (Mattingly and May, 2011; NRC, 2011). The current baseline for the first mission is the proposed NASA/ESA Mars 2018 joint rover mission, which would utilize a heritage Mars Science Laboratory (MSL)–style entry, descent, and landing (EDL) system (Steltzner et al., 2006) to land on Mars a jointly developed rover with the capacity to meet both in situ and MSR scientific objectives. This rover would collect and cache samples that would be ready for future pickup. The proposed second mission is the MSR Orbiter (MSR-O), which could be two Mars launch opportunities (i.e., 4 years) later. MSR-O is projected to launch before the MSR Lander (MSR-L) so that it could provide telecommunications infrastructure for the lander mission operations during arrival and after landing. The third mission is the proposed MSR-L, which would also use an MSL-style EDL system to deliver the lander platform to the surface of Mars. The lander platform would dispatch a fetch rover to retrieve the sample cache. The cache may be augmented by a lander-collected sample of gas and possibly also regolith. The landing platform would also serve as a launch pad for a Mars Ascent Vehicle (MAV), which would be used to insert an orbiting sample container (OS) into a 500 km near-circular orbit. After monitoring the MAV launch through release of the OS from the MAV, MSR-O would rendezvous with and capture the OS, which would then be packaged into an Earth entry vehicle (EEV) carried with MSR-O. The orbiter would then return to Earth, release the EEV a few hours before entry, and divert into a non-Earth-return trajectory. The EEV would hard land on the surface and then be transferred to a secure sample receiving facility (SRF; see Beaty et al., 2009, and references therein) for quarantine before samples would be made available for scientific research.

2.1. Introduction

Several groups, most recently MEPAG (2010) and the National Research Council (NRC, 2011), have outlined the science objectives for Mars exploration. The objectives are diverse enough that no single mission could address them all. It has been recognized for several years that sample return is the single activity that would make the most progress toward the entire list. Most recently, the NRC (2011) concluded that the state of our knowledge of Mars is such that MSR is the next logical major step in Mars exploration and that there are no other strategic missions worth flying before MSR. However, one of the realities of sample-based science is that not all samples are equally useful for all kinds of scientific questions. A first step, therefore, in planning the MSR Campaign is to identify and prioritize the science objectives. This is then input into choosing where to go to get the samples and to determine the many sample-related attributes of the flight hardware such as mission lifetime, EDL capabilities, mobility parameters, sampling capabilities, on-Mars measurements, and sample preservation capabilities.

The ’03/’05 MSR mission was planned in 1999–2000 with seven scientific objectives (listed in Table 1 of MEPAG ND-SAG, 2008), all of which were highly generalized. These objectives provided very little guidance on where to send the mission or the kinds and quantities of samples to collect. Based on the results of the NASA and ESA exploration missions during the period 1996–2007, the MEPAG ND-SAG (2008) proposed 11 much more specific statements of “possible scientific objectives,” with the caveat that “the achievable degree of progress towards these scientific questions would depend on the choice of landing site” and several other factors, thereby deferring several essential questions of scientific priority to successor planning teams and to a later landing site selection process.

Several additional recent studies have played a major role in setting the scene for the present update of MSR scientific planning. The NRC's Astrobiology Strategy report (NRC, 2007), MEPAG ND-SAG (2008), iMars Working Group (2008), MEPAG MRR-SAG (2010), and the Decadal Survey report (NRC, 2011) all emphasized that we are now ready to be much more specific about MSR scientific planning, strategies, and priorities, and that astrobiology-related objectives have become important enough to become a driving force behind the mission.

The scientific value of a potential MSR has been discussed in the literature for at least 30 years (see, e.g., NRC, 1978, 1990a, 1990b, 1994, 1996, 2003, 2006, 2007), and the scientific rationale for returning samples has evolved over time. Early studies (e.g., NRC, 1978) emphasized the need for samples to better understand the evolution of Mars as a planet. More recently, however, emphasis has shifted to the search for life (e.g., NASA, 1995; NRC, 2007).

Finally, for the purposes of this study, the vision of the MSR Campaign was constrained by what was considered to be practical (see charter assumptions in Appendix A2). Thus, objectives that might require high latitudes, high elevations, deep (>2 m) drilling, and large sample masses, for example, were given lower priority. The intent was not to provide a prioritized list in the abstract, but one that could be used to guide a sample return campaign in the context of our current knowledge and expectations of future engineering and fiscal resources.

2.2. Specific proposed science aims and objectives for Mars returned sample science (organized by topic)

In the present study, the E2E-iSAG considered all previous statements of scientific objectives for Mars returned sample science and more recent perspectives from a broad segment of the Mars science community, along with its own perspective and judgment. The E2E-iSAG also developed eight implementable statements of scientific objectives, which are organized under four higher-level scientific aims (Table 1). The eight objectives have been prioritized, and it is our recommendation that these priorities guide landing site selection and a broad range of implementation considerations. Whereas the ND-SAG and MRR-SAG objectives encompassed a range of possible MSR missions (for example, a mission to return an ice sample), the present study is intended to be limited to missions that could be implemented with currently plausible resource and political constraints, as reflected in the assumptions presented in our charter (Appendix A2).

2.2.3. Aim C. Planetary evolution

2.2.3.1. Objective C1. Quantitatively constrain the age, context, and processes of accretion, early differentiation, and magmatic and magnetic history of Mars

Understanding the main phases of planetary-scale geological evolution of Mars is essential to providing context for other scientific objectives. Significant insights into planetary-scale evolution of Mars have been derived from studies of martian meteorites—the only samples available for research until MSR occurs. Such studies provide a complementary approach to Mars exploration (McCoy et al., 2011), but the meteorite studies have certain limits. The martian meteorites are all igneous rocks, formed from the eruption or emplacement of mafic magmas through variable degrees of accumulation of olivine or pyroxene (e.g., McSween, 1994) that were ejected from Mars through impact within the past few million years (Nyquist et al., 2001; Fritz et al., 2005). Their martian origin was established through the discovery of high ratios of ⁴⁰Ar/³⁶Ar and ¹²⁹Xe/¹³²Xe in trapped gases in shock-altered phases of the Antarctic meteorite Elephant Moraine A79001 that closely resembled the martian atmosphere measured by the Viking lander (Bogard and Johnson, 1983). The majority of the 55 known martian meteorites date from the younger Hesperian to the Amazonian (1.3 Ga and 200–600 Ma; Nyquist et al., 2001); ALH 84001 is the only sample from the Noachian (4.09 Ga; Lapen et al., 2010). Several attempts have been made to identify the source craters of the martian meteorites based on spectroscopy of laboratory samples compared to spectra from Mars Global Surveyor Thermal Emission Spectrometer (MGS-TES), with only limited results (Hamilton et al., 2003; Lang et al., 2009). Thus, the martian meteorites lack “field” context and cannot be used to provide ground truth for orbital missions.

Detailed studies of the mineralogy, petrology, and geochemistry of the martian meteorites have yielded insights into the geological evolution of Mars, including the timing of differentiation, mantle mineralogy and water content, and the thermal state of the interior (e.g., Stolper and McSween 1979; Bertka and Fei, 1997; Borg et al., 2003; Leshin and Vicenzi, 2006; Papike et al., 2009). Initial ⁸⁷Rb/⁸⁶Sr and ⁸⁷Sr/⁸⁶Sr values for meteorite whole rock samples plot on a line with a slope corresponding to 4.5 Ga (Borg et al., 1997), and a similar 4.5 Ga isochron emerges from Sm-Nd isotopic compositions (Borg et al., 2003). The 4.5 Ga “age” is interpreted to represent the signature of ancient mantle sources established early in the differentiation of the martian interior via crystallization of a magma ocean (Borg et al., 2003). In this scenario, isotopically distinct mantle reservoirs remain unmixed and undisturbed until the melting event that produces a magma, which crystallizes in a near-surface intrusion or lava flow (at 200–600 Ma, in the case of the basaltic martian meteorites). Alternatively, some workers have suggested that the same meteorites have ancient crystallization ages (4.0–4.5 Ga) and that younger ages reflect the timing of resetting by impact or metasomatism (e.g., Bouvier et al., 2008). Much of the debate about crystallization ages of martian meteorites stems from the lack of geological context, as well as the overprinting of original geochemical signatures by the effects of impact shock (e.g., Fritz et al., 2005). If the 200–600 Ma ages are indeed crystallization ages, as most workers contend, the overrepresentation of Amazonian rocks is attributable to a sampling bias toward younger, more competent igneous rocks; older rocks representing primitive martian crust and weakened by impacts and aqueous alteration are discriminated against in the launch process (Walton et al., 2008, and references therein).

Recent results of Mars surface exploration demonstrate that the martian meteorites are not representative of the bulk of the rocks found at the martian surface (McSween et al., 2009), which raises doubt as to whether the geological evolution of Mars inferred from the meteorites can be applied to Mars as a whole. As an example, basaltic martian meteorites have CaO/Al₂O₃ ratios that are greater than chondritic, an observation that is best explained by majoritic garnet fractionation in a deep magma ocean (Bertka and Holloway, 1994; Bertka and Fei, 1997). However, the ∼3.7 Ga Gusev basalts have chondritic CaO/Al₂O₃ ratios; along with other elemental ratios and concentrations, it is apparent that Gusev basalts and basaltic martian meteorites are derived from different mantle sources (McSween et al., 2004; Schmidt and McCoy, 2010). This discrepancy suggests that different regions of the martian mantle may have undergone different processes and calls into question a global magma ocean. If the mantle inferred from martian meteorites is not applicable to all of Mars, then estimates of mantle water contents derived from martian meteorites (e.g., Mysen et al., 1998; McCubbin et al., 2010) may also have limited applicability.

The main shortcomings of the over 55 unpaired martian meteorites are listed in Table 2. Any version of MSR must enable science in the area of this objective over and above what is possible using the martian meteorites.

OPEN IN VIEWER

Table 2.

Shortcomings of Martian Meteorites as Representative/High-Priority Samples of Mars

No.	Shortcomings of Martian Meteorites
1	Sampled out of context
	- Launched from unknown sites, with unknown geological context on Mars by impact
2	Inadequate sampling of the Noachian and Hesperian
	- With one exception, all are < 1.4 Ga, i.e., Amazonian
3	They are all shocked to varying degrees
	- Experienced peak pressures of up to 60 GPa; affects mineralogy and radiogenic isotopes
4	No evolved igneous compositions
	- All are mafic to ultramafic
5	Most have been affected by liquid-crystal fractionation
	- Limitation for determining compositions of their parental melts
6	Orientations relative to planetary surface are not known
	- Not usable for magnetic studies
7	All have been subjected to deep space radiation (affects some isotopes)
8	Do not provide ground truth for orbital or surface missions
9	All (but especially the finds) have been affected by some form of terrestrial alteration and/or contamination

2.2.3.1.1. Sample types of interest—C1

Igneous rocks would provide the best possible samples for age determination by radiogenic isotopes and for constraining the martian interior with trace elements, isotopic composition, and other analyses. Locations should be identified from composition and geomorphological context, for example, igneous intrusions or lava flows. Samples should be as unaltered and unweathered as possible so as to preserve the igneous texture and distribution of elements as set during crystallization; preservation of these aspects would be more important than other attributes such as compositional or age diversity. In-place sampling would be strongly preferred over float. Samples from float that could be tied to a local outcrop (“subcrop”) or volcanic vent (e.g., pyroclastic materials) are strongly preferred over other rocks that may have been carried some distance by impact (and therefore are also more likely to have witnessed high shock effects). Samples should be located so that their stratigraphic age could be narrowly constrained, in order for their absolute ages, once determined on Earth, to provide constraints on the stratigraphic sequence from which they were obtained (and contribute to Objective B2). Age diversity would be less important than obtaining unaltered or unweathered samples with known stratigraphic context.

The following sample types are of priority interest for achieving this objective:

1. Ancient igneous rocks. There is no sample of the presumed primitive crust of Mars among the meteorites, although studies of martian meteorites indicate that silicate differentiation on Mars occurred by 4.51 Ga (Borg et al., 2003) and crystallization of a magma ocean, and subsequent overturn, may have been responsible for establishing initial reservoirs, including a primitive crust (e.g., Elkins-Tanton et al., 2005; Blinova and Herd, 2009). Thus, early Noachian igneous rocks would be preferable for this purpose. Similarly, there are no samples of late Noachian to early Hesperian among the martian meteorites. Within these constraints, the following rock types are preferred:

a. Noncumulus basalt (e.g., chilled flow margin). Two examples of noncumulus basalts are present among the martian meteorites. Of these, Yamato 980459, a glassy, olivine-bearing basalt whose bulk composition matches that of its presumed parent melt (Musselwhite et al., 2006), is thought to represent a primitive melt from the martian mantle. Samples similar to Yamato 980459 would be preferable in order to remove some of the difficulties in reconstructing parental melts, for example, from melt inclusions. A suite of samples from chilled flow margin to flow interior would be highly desirable.

b. Ultramafic (including xenoliths of possible mantle or lower crustal origin). The majority of the martian meteorites are mafic to ultramafic in composition. None of the meteorites has compositions and mineralogy consistent with an ultramafic mantle xenolith or a sample of the lower crust. Although the sampling of such a rock type would be fortuitous (see sampling strategies below), such samples would provide insights into the mineralogy of the martian interior, allow tests of mantle mineralogy inferred from meteorites (Bertka and Fei, 1997), and place further constraints on the thermal state of the martian interior, in conjunction with geophysical data (Khan and Connolly, 2008). A sample suite that included mantle xenoliths(s) with basaltic melt counterpart(s) would provide links between mantle source composition and derived melt(s) and would be desired but not required. A cumulate ultramafic rock would provide insights into crystallization processes in a martian magma chamber and processes of magma evolution on Mars in general.

c. Evolved igneous compositions. The range of igneous compositions on Mars is limited; martian meteorites are limited to basalts+ultramafic rocks, whereas igneous rocks at Gusev are primarily basalts, with some compositions falling in the picrobasalt and tephrite fields (McSween et al., 2009). More felsic compositions have been identified with thermal emission spectroscopy that potentially include evolved, granitic rocks (Bandfield et al., 2004; Christensen et al., 2005). A sample of an evolved igneous rock would provide significant insights into the potential range of igneous compositions that may be produced by melting of the martian interior. A sample suite of igneous rocks, for example, related by liquid line of descent would be desired but not required.

2. Young volcanic rocks. The youngest martian meteorite is ∼170 Ma (Nyquist et al., 2001); nonetheless, orbital observations show clues for volcanic activity in the last hundred million years and possibly even in the last million years (Neukum et al., 2004). Although examples are generally older, IR spectroscopic observations suggest that volcanic eruptives of diverse compositions may exist on Mars (Christensen et al., 2005). If a young volcanic rock could be sampled, it would provide significant insight into the range of volcanic processes and the diversity of mantle sources planet-wide through time. Furthermore, a young volcanic sample would provide information on the martian core if paleomagnetic measurements were made. Finally, such a sample would be an excellent target for evaluating the longevity and history of water on Mars because our current models suggest that water has only existed episodically on its surface in the recent to modern times.

Significance of sampling for magnetic studies. The magnetic history of Mars is a fundamental question that relates to the geological evolution of the planet as a whole (e.g., Lillis et al., 2008; Weiss et al., 2008). The ability to quantify the duration and magnitude of Mars' magnetic dynamo would provide novel insights into the thermal evolution of the martian interior. Paleomagnetic measurements could also be used to test the hypothesis that Mars experienced plate tectonics and true polar wander, and as a tool for field geology (e.g., using magnetostratigraphy as a chronometer to correlate rock sequences). The ideal suite that would likely address this question would include Noachian to Hesperian samples, because Mars' core is generally thought to have been active during this time frame (although it is not entirely clear when the dynamo died). If only one sample could be selected, it should be as old as possible in order to get a sample that records paleomagnetism and allows for measurement of the intensity of the magnetic field at that time point (as determined from radiometric dating of the same sample). Any early Noachian rock would likely preserve a record of the dynamo. It would not be necessary to target high-magnetization areas (as identified from MGS results; Acuña et al., 2001), although those areas would be of interest to explain the strong fields observed.

The most useful rocks for this purpose would be igneous because they tend to contain primary igneous magnetic minerals (e.g., titanomagnetite and pyrrhotite). However, sedimentary rocks, especially those containing iron oxide minerals (e.g., hematite, maghemite, and goethite) as cement, may also be useful and would provide the advantage of paleo-orientation (using bedding) and various field tests that could demonstrate primary magnetization (fold test and conglomerate test).

Although the intensity of the magnetic field would provide significant information, the fraction of science value achieved with intensity data alone would be less than that which could be achieved with oriented samples. To determine the absolute direction of the ancient martian magnetic field with only a few samples, it would be extremely important that the orientation of each sample with respect to the martian surface be recorded. The direction could be used to test the dynamo hypothesis, to test hypotheses that Mars experienced plate tectonics and true polar wander, and to conduct magnetostratigraphy.

Implications for the sampling system. The importance of in-place sampling would be very high for achieving Objective C1. Samples collected within a known geological context, including those of a local (i.e., outcrop), regional, and planetary scale, would provide the primary advantage over martian meteorites and allow for the information obtained from their study on Earth to be directly related to the specific phase of Mars' planetary evolution they represent. Sampling of rocks from probable Noachian terrane would enable the presumed age to be tested by radiometric dating on Earth. Sampling of rocks from outcrop, within the context provided by outcrop-scale characterization, would allow for selection of igneous rocks that have undergone lower levels of shock metamorphism and selection for compositional diversity. Therefore, the implications on the sampling system might include

(1) Need to visit Noachian to early Hesperian igneous outcrop, targeting those least affected by alteration, weathering, or impact shock metamorphism. Implication: capacity to get beneath weathering rinds to fresh samples; capacity to assess level of shock metamorphism.

(2) Target rocks that exhibit mineralogical or textural characteristics that suggest rapid cooling from a melt (e.g., quenched flow margin). Implication: capacity to characterize texture at sufficient spatial resolution.

(3) Samples of opportunity: During the traverses of the sampling rover, we should constantly check for exotic blocks that would allow us to further expand the range of rock types sampled. Ultramafic xenoliths in outcrop (e.g., within a basaltic unit) would also represent a sample of opportunity. Implication: capacity to assess compositional diversity and macroscale textural variations at a distance.

(4) The implication for paleomagnetic samples: capacity to sample such that the orientation with respect to Mars' surface would be preserved. Samples would need to be isolated from magnetic fields (ideally less than 0.2 mT; Weiss, written communication, 2011).

2.2.3.2. Objective C2. Constrain the origin and evolution of the martian atmosphere, accounting for its elemental and isotopic composition with all inert species

Understanding the evolution of the martian atmosphere is essential to explaining the occurrence of liquid water on the surface in Mars' early history, and that in turn influences conclusions on the habitability of the planet. Existence of liquid water on the martian surface, other than short-lived concentrated brines, requires higher atmospheric temperatures and pressures than present-day conditions. Although precisely how the warmer conditions were achieved remains unknown, a thicker CO₂/H₂O atmosphere (released from the martian interior by extensive volcanism in the early Noachian) is implied. Evidence from surface features on Mars suggests, however, that this atmosphere, and thus persistent surface liquid water, had largely disappeared by the end of the Noachian, which led to the dry and low-pressure conditions observed on Mars today.

Models of formation, evolution, and loss of the martian atmosphere are based for the most part on the isotopic compositions of noble gases (Pepin, 2000, 2006). Currently available data on some atmospheric components are from in situ analyses by the Viking landers, but the isotopic compositions of neon, krypton, and xenon are exclusively from analyses of the martian meteorites, in particular EETA79001 (see review by Swindle, 2002). The latter contains unfractionated atmosphere in its impact shock glass (Table 3).

OPEN IN VIEWER

Table 3.

Current and Potential Pre-2018 Knowledge of Gas Chemistry of the Martian Atmosphere (EMTGO is Currently a Proposed Mission)

		Knowledge as of 2011		Expected results from future missions before MSR, 2011 estimate			Analysis of returned sample
GAS	Parameter of interest	Precision	Source	MSL rover (SAM instrument) on ground	2013 MAVEN orbiter above homopause (>120 km)	2016 EMTGO orbiter (MATMOS instrument) orbiter	Precision *4	Magnitude of MSR benefit over MSL	Gas volume, cm 3 STP *6
Ar	40Ar/36Ar	>10%	meteorites	0.2–1.5%	• measures atmosphere above homopause	none	<<0.01%	MODEST	8E-03
	38Ar/36Ar	5%	meteorites	0.5–2.5%	• atmosphere inhomogeneous, dominated by diffusion, particles start to ionize due to solar wind, irradiation, and escape	none	<<0.01%	MODEST
N2	15N/14N	<5%	meteorites	2%			<<1%	MODEST
CO2	delta 13C	0.4%	Phoenix *2	0.2%		0.1%	?	MODEST
	delta 18O	0.5%	Phoenix *2	0.3%		0.1%	?	MODEST
	delta 17O	none		0.5%		not specified	?	MODEST
Kr	isotopes	<1% for major isotopes5% for 78Kr	meteorites	major isotopes: 0.2–1.5%		none	<<0.1%0.2% for 78Kr	MAJOR	5E-08
Xe	isotopes	<2% for major isotopes10% for 124,126Xe	meteorites	major isotopes: 0.2–1.5%	MAVEN data are not comparable with bulk atmosphere	none	≤0.1%0.7% for 124,126Xe	MAJOR	2E-08
Ne	isotopes	7% no data on 21Ne	meteorites	issues *1		none	0.01%	MAJOR	1E-06
He	isotopes	none		none		none	0.01% 4He0.3% 3He *5	MAJOR	1E-06
methane	abund.	10±10 ppb	several	<1 ppb		<1 ppb	not stable and reactive
	13C/12C	none		0.2%		15 ppt
	D/H	none		TBD		4%
ethane	abund.	none		TBD		5 ppt	stable?
S gases	H2S	none		10s ppb		0.5 ppb	molecules not stable and reactive
	OCS	none				1 ppt
	SO2	none				2 ppt
	isotopes	none		issues *3		?
H2O	D/H			0.2%		0.6%	<1‰	MINIMAL
	delta 18 O			0.3%		0.10%	<0.1‰	MINIMAL
	delta 17 O			0.5%		not specified	<0.1‰	MINIMAL

*1

requires complex gas separation system.

*2

Niles et al., 2010.

*3

could be close to detection limit, 1–2% if close to fumaroles.

*4

Numbers are based on one aliquot analysis, see Table 2.

*5

based on the terrestrial ⁴He/²⁰Ne, and ³He/⁴He ratio, no Mars data available.

*6

Concentration as expected in one aliquot of the gas sample that results in the given precision.

Important, but so far unsolved questions require high precision composition data of all noble gases and their isotopes, which would be only achievable with a returned atmospheric gas sample analyzed in terrestrial laboratories. Those questions involve (a) the formation of the atmosphere: primordial versus volcanic outgassing versus late veneer (volatiles delivered from outside, for example, by comets); (b) duration of the thicker warmer atmosphere; (c) age of the atmosphere: young (continuous outgassing) versus old (primordial); and finally (d) the cause of the atmospheric loss: impact erosion during the late heavy bombardment versus a steady atmospheric escape after the disappearance of a global magnetic field (Fig. 3).

OPEN IN VIEWER

FIG. 3.

Overall scientific objectives of atmospheric gas samples. Color images available online at www.liebertonline.com/ast

Atmospheric noble gas science is mainly based on EETA79001 impact glass, but interpretation of meteorite data is complicated due to possible mixing of different noble gas components. Spallogenic products mask the light noble gases (He, Ne, Ar) and the minor isotopes of Kr and Xe, and a significant correction is required for the more abundant isotopes. Krypton and xenon in meteorite finds are also altered due to adsorption of terrestrial atmospheric gases. Other martian meteorites revealed the presence of multiple components: martian interior, a soil component, and a paleoatmospheric signature (see review by Swindle, 2002).

Future in situ analyses by the MSL rover Curiosity, the 2013 MAVEN, and the proposed 2016 ExoMars Trace Gas Orbiter (EMTGO) orbiters would analyze the trace gases (CO₂, H₂O, O₂, CH₄, N, sulfur molecules, C₂H₆, etc.) and their H, C, N, O, S isotopic compositions with high precision (written communications: MSL, Paul R. Mahaffy, 2010; MAVEN, Bruce Jakosky, 2010; EMTGO, Mark Allen, 2010). Table 4 summarizes the strengths and limitations of both in situ measurements and returned samples. The former provide high spatial and temporal resolution data of the atmospheric composition and are also capable of analyzing nonstable gas molecules (e.g., sulfur molecules, CH₄, H₂). A returned gas sample, in contrast, would allow high-precision analyses of all stable components, including the minor isotopes of the noble gases and other stable species (as CO₂ and N₂). As a consequence of the limitations of the meteorite data and orbital measurements, the analysis of noble gases would be a major objective for a sample return mission (Table 3).

OPEN IN VIEWER

Table 4.

Strengths and Limitations of In Situ and Return Sampling of Mars Atmosphere (EMTGO is Currently a Proposed Mission)

	Mission/samples	Strengths	Limitations
in situ	Viking	Noble gases (abundances, isotopic compositions)	Uncertainties (10–20%)
	MSL	• Capture of seasonal and spatial variability	No He, Ne, minor Kr and Xe isotopes with large uncertainties, no spatial resolution
	MAVEN	• Abundances+isotopic composition of nonstable molecules (as sulfur molecules, CH4 (if it exists)	Composition of the uppermost atmosphere (ionized particles, escape)
	EMTGO/MATMOS	• High precision on abundant species	No noble gases
returned sample	SNC meteorites	Heavy noble gases, = basis for many atmospheric models	Significant cosmogenic contribution to all noble gases, He too low, no control on when gas was introduced nor on different gas components
		High precision
	MSR	All stable atmospheric species (noble gases, CO2, N)	No spatial or temporal variability
		High precision also for minor isotopes	Nonstable atmospheric species do not survive

FINDING #5: We will have at least a partial understanding of the geochemistry of most major components of the martian atmosphere prior to the proposed receipt of samples on Earth. The area in which we anticipate still having the most significant unresolved questions is in the noble gas isotopes.

FINDING #6: The noble gases in martian meteorites are often mixtures of different components that introduce ambiguities in interpretation of measured data. A direct sample of the atmosphere would remove this ambiguity.

Detailed open questions related to martian noble gas geochemistry:

A better ³⁶Ar/³⁸ Ar value would help to constrain models of atmospheric loss. The martian atmosphere is the only one in the Solar System that shows a significant deviation from solar ³⁶Ar/³⁸Ar. Presumably, this is due to atmospheric loss of the lighter isotope, which is somewhat mitigated by mantle outgassing. Resulting loss rates should be in conjunction with the Ne isotopic composition, the ¹⁴N/¹⁵N, and the ¹²C/¹³C in CO₂ composition.

Ne and He isotopic composition (and abundance) to constrain atmospheric evolution and mantle degassing. Neon abundances and its isotopic composition in the martian atmosphere are still poorly known. Both ²⁰Ne and ²²Ne are difficult to measure at Mars because of serious ⁴⁰Ar and ⁴⁴CO₂ interferences. In martian meteorites, isotopes are masked by cosmogenic Ne. MSL will not be able to produce high-precision data due to the lack of a gas separation mechanism. The Mars Atmospheric Trace Molecule Occultation Spectrometer (MATMOS) instrument proposed for EMTGO would not measure noble gases. Thus, all Ne science would be deferred to MSR.

The He abundance and isotopic composition in the martian atmosphere is unknown as it is entirely masked by cosmogenic contribution in meteorite samples. The He isotopic composition in fluid inclusions in igneous rocks that are not exposed to galactic cosmic rays (GCR) would deliver valuable information on the degassing history of Mars.

Combined atmospheric Kr and Xe signatures. Krypton and Xe and their isotopic compositions are key in determining the starting composition of the volatiles on Mars. However, modeling the evolution of the martian atmosphere is seriously affected by the absence or imprecision of critical data. Xenon in the martian atmosphere is either of solar or chondritic origin. However, whereas the ¹²⁹Xe suggests significant martian interior outgassing, missing heavy Xe isotopes related to ²⁴⁴Pu fission (^131–136Xe) indicate the opposite. Precise and accurate measurements of the rare ^124–126Xe and ¹²⁸Xe isotopes would not only reveal the extent of mass fractionation of the starting composition but also the degassing history of Mars (Pepin, 2006).

The Kr isotopic composition of Mars' atmosphere may be either solar (Pepin, 1991) or fractionated in favor of light isotopes (Garrison and Bogard, 1998). Presumably, atmospheric Kr has been replenished by degassing after early atmospheric loss. Additionally, elevated amounts of ^80,82Kr indicate neutron capture by Br. However, it is unclear whether this is a feature of the rock (i.e., a soil component) in which the Kr was measured or an atmospheric signature. It is clear that a model of the origin and evolution of the martian atmosphere has to include and coherently explain the isotopic compositions and abundances of all noble gases and other volatiles.

FINDING #7: A key to understanding the origin and evolution of the martian atmosphere would be to measure the isotope systematics of each of the noble gases and construct a model that is consistent for all of them.

Seasonal variability of the present martian atmosphere. The present martian atmospheric pressure seasonally varies by about 20%, due to the polar CO₂ ice cap formation and sublimation. Swindle et al. (2009) proposed that a substantial fraction of Xe and less of the Kr and Ar are incorporated into polar clathrates (or perhaps also in ice). Accordingly, the seasonal formation or dissociation of a small amount of clathrates could change the atmospheric Kr/Xe ratio and would probably have some effect on the isotopic composition of Xe. Detectable seasonal changes of C and O isotopic composition due to sublimation/condensation of CO₂ ice are also considered possible (Mark Allen, personal communication, 2011). Respective fractionation of the Xe and Kr isotopes, either due to CO₂ ice or clathrate formation, thus could not be ruled out. No further studies have been carried out (Tim Swindle, Bob Pepin, Oliver Mousis, personal communications, 2011). However, MSL may be able to address this question by monitoring the Kr/Xe and a Xe isotopic ratio (best ¹²⁹Xe/¹³⁶Xe) over 1 martian year. If significant variations are found, two atmospheric gas samples should be taken, one at pressure minimum and the other at pressure maximum. The high-low pressure pair could assess the extent of seasonal clathrate formation (and the incorporation of Xe). Secondly, it could reveal differences driven by other mechanism in the atmosphere (Bob Pepin, personal communication, 2011). Finally, the second sample would also mitigate risk by serving as a backup if one sample is lost (e.g., on Mars due to sealing failure).

FINDING #8: There would be scientific value in collecting atmospheric samples at two different seasons; however, although this is considered desired, it is not required. If it were possible to return more than one sample, we do not currently recognize a need to collect such samples from more than one location.

2.3. Prioritization of scientific objectives

A key part of the E2E task was to “prioritize a reference set of ‘campaign-level’ science objectives.” Given the science objectives described above, the E2E team developed a top-level criterion and a set of subcriteria to guide the prioritization (Table 5). Since the chief aim of the proposed MSR Campaign would be to return martian samples to Earth for detailed study, the key criteria for prioritizing objectives is the value of the incremental knowledge to be gained by returned sample analysis. This value is judged by the degree to which sample analysis could address the highest-priority questions in Mars science, as most recently summarized by MEPAG (2010) and the NRC (2011). The value of returned samples is highest if they address questions for which little progress could otherwise be made due to instrument or sample preparation complexity, or to the scale of the investigation. Laboratory analyses generally have notable advantages over in situ analyses in terms of precision and accuracy, and laboratory results could be confirmed by alternate methods. In addition, laboratory instruments and techniques could be adapted to discoveries much more easily than instruments that are locked into a spacecraft payload.

Using these criteria, we have ranked the science objectives in priority order (Table 6). It is worth noting that, as a multidisciplinary team, the members of the E2E committee have diverse interests and perspective, and there was a certain amount of variation in how scientists from different disciplines viewed these priorities. However, despite that, the placement of Objectives A1 (life) and C1 (planetary evolution) as the top two priorities (in that order) was strongly felt by the scientists from almost all disciplines represented on the team.

As per the arguments discussed above in Section 2.2.1.1, the search for evidence of extant life cannot be effectively prioritized with the same criteria. The importance of this objective is superimposed on all the others.

3.1. Relationship between field and sample science

For some laboratory studies of samples an associated field context is available, and for some studies it is not (e.g., meteorite studies). The former is always better constrained and therefore more useful than the latter, and for the kinds of complex astrobiology/geology questions posed in this report, field context would be required. As applied to MSR, this is what distinguishes a “grab-and-go” version of MSR (MEPAG, 2002) from a selected-sample version of MSR (NRC, 2011). Robust interpretation of even the most sophisticated laboratory analyses of geological specimens ultimately relies on the quality and extent of observations of the field relationships used to select samples. This fundamental insight was aptly summarized by the legendary sedimentary geologist, Francis Pettijohn: “For only by a firsthand acquaintance with the primary phenomena of geology, obtainable only by field study, can significant research be distinguished from the trivial” (Pettijohn, 1984, p 248). Accordingly, for the MSR Campaign to obtain the greatest information, it is considered crucial that samples be collected such that their geological context is well understood. Some of this context could be characterized from orbital data obtained prior to landing, but many details would only be obtained by in situ study at, and around, the sites of collection.

3.2. Information hierarchy

Figure 4 illustrates the time-honored approach to answer scientific questions that involves the use of both a selected field site and laboratory analysis data—such studies are extremely common in the natural sciences. It begins with well-posed scientific questions and selection of a field site for sampling that best addresses those questions. Geological relationships of the site are characterized through a variety of approaches (e.g., lithological mapping, stratigraphic traverses, textural observations). Appropriate sample suites are then collected on the basis of those field relationships. Ideally, this is done in a sequential manner, but on Mars (and most investigations on Earth) logistics would impose a more iterative process between characterizing geological context and selecting samples. Samples returned from the field are examined and analyzed in laboratories, and finally results from both the field and laboratory are then integrated to obtain “answers” that respond to the original questions.

OPEN IN VIEWER

FIG. 4.

Schematic flow chart outlining the roles of samples and geological field characterization in the evaluation of scientific questions. Color images available online at www.liebertonline.com/ast

It is virtually axiomatic to science that any investigation of a scientific question will lead to new, unanticipated questions and to new ways to pose the original question that provide better insights. Accordingly, another advantage of characterizing geological context as thoroughly as practicable during sample selection is that it would allow for unanticipated and revised questions to be addressed without necessarily returning to the sampling site, a critical asset for planetary missions.

3.3. Collect early, exchange later

In addressing geological questions that involve coupled field work and sample-based laboratory studies, there is commonly a tension between collecting potentially valuable samples when they are encountered and waiting to understand the geology better before starting to fill up the sample collection. In principle, it would be ideal to complete all geological mapping first and then use that information to decide which samples would be most important. However, field studies are rarely carried out this way because of the extreme inefficiency of returning to localities a second time to sample. Therefore, samples are typically collected as the field survey is being carried out. Some of the earlier samples will inevitably turn out to be less valuable than those encountered later (especially because the geology is better understood), but the time-honored way of solving this on Earth is to toss the earlier, less valuable samples out of the backpack and replace them with later-collected more valuable and appropriate samples. This capability greatly reduces the pressure on sample collection decision-making early in the traverse and allows the geologist to benefit from his or her increasing knowledge as fieldwork progresses.

On Mars, the inefficiency of returning to outcrops a second time to collect samples would be unacceptable given the limited lifetime and traverse capability of a rover. Thus, the importance of building the collection “as you go” is even higher on Mars than on Earth. Therefore, as discussed above, to support effective decision-making, the capability to swap out samples is judged to be extremely important. The collective judgment of the E2E-iSAG team is that an excess sampling capability of 25% should be sufficient to address this need. That is, 25% of the core tubes could be replaced with cores of higher value for the final suite that would come back to Earth.

3.4. The importance of rock and soil sample suites

It is possible for individual, unrelated samples to provide useful data. However, most rocks record the effects of multiple geological processes, and these effects can only be unraveled uniquely by studying multiple samples in which the effects differ. Single samples could be likened to individual snapshots that do not provide information about gradients of change in time and space. It is common with such single datum samples not to know how they are positioned in the evolution of time and within the variability of environments. They provide limited help in identifying the geological and environmental context. The evaluation of complex systems like those anticipated for the MSR Campaign would be particularly context-dependent. It is critical to decipher the geological changes that remain impressed in the stratigraphic record of outcrops and other terranes by analysis of a series of samples. Therefore, it would be necessary to collect suites that comprise a series of samples that are genetically connected in order to build a vision of planetary environmental evolution through time. This finding reinforces a key conclusion previously made by the MEPAG ND-SAG (2008).

4.1. Integrated priorities for rock samples

Rock sample types are listed below in priority order with a single prioritization criterion: potential value to achieving the objectives in Table 6. Category 1 samples are given the highest priority because of their importance for addressing the possibility of past life and issues related to climate change and the history of water. Category 2 samples (fresh igneous rocks), the next highest in priority, are important for absolute dating and understanding the evolution of the interior of the planet.

Category 1A. Subaqueous sediments or hydrothermal sediments . These primary sample target types were rated to be of highest priority in the search for life due to the fact that these have the greatest potential for preservation of biological signatures, as well as prebiotic chemical signatures and observations that help establish an abiotic baseline. In addition, subaqueous and hydrothermal sedimentary deposits have the maximum potential to provide contextual information on the depositional environment, habitability, and preservation potential, based on the preservation of information in order and sequence (clearly recording past spatial-temporal relationships) in the outcrop. They were rated of equal priority primarily to ensure that a sampling of multiple environments that might have suitably hosted a broad spectrum of possible modes of life was captured. This includes not only a focus on sedimentary formations formed subaqueously that may preserve surface-based life but also includes sample types that encompass the possibility of life based on chemoautotrophic principles (similar to those discovered at hydrothermal vents, hot springs, and emerging subsurface groundwaters on Earth) (Chapelle et al., 2002; Kelley et al., 2005; Lin et al., 2006; Chivian et al., 2008; Proskurowski et al., 2008).

Category 1B. Hydrothermally and low-temperature fluid-altered rocks . The distinction between these is described above in Section 2.2.1.1. As with the samples in Category 1A samples, these fluid-altered rocks are also relevant to Objectives A1, C1, C2, and in a minor way to B2 (for fluid inclusions). Category 1A samples are of higher priority than those of Category 1B due to the inherently higher potential for preservation of both biotic and chemical signatures as well as for the higher potential for preserving contextual information regarding the timing and paleoenvironmental sequence and context for life. In addition, a sample suite derived from a sedimentary record would be more useful to Objectives C1 and C2 because of the added value of good temporal context.

Category 2. Unaltered igneous rocks . This sample type is required to achieve Objective C1, the second-highest-priority objective in Table 6. In contrast to the Category 1 samples, the less altered the sample by water-related processes, the better. It is strongly preferred that the sample would be free of shock effects and so should be sampled in place rather than from float, but the sampling rover could encounter lithological exotics on the martian surface that have been blasted in from somewhere else that would be well worth sampling. Although this sample type is crucial to Objective C1 and is of minor importance to C2 (because of fluid inclusions), it is of little importance to the other objectives.

Category 3. Regolith samples . These are vital for understanding the hazards that surface material may present to human exploration (Objective D1) and the opportunities that may exist for human resources (Objective D2). Regolith samples are also of scientific interest for insights into surface/atmosphere interactions and interactions with the space environment (Objective B3). Finally, if the regolith samples contain exotic fragments of rock that improve the diversity of the returned sample collection, they could be of great value to Objectives A1, B1, B2, C1, and C2.

Category 4. Atmosphere . This sample type is definitively required to achieve Objective C2, which is important, but in a relative sense is ranked priority #7 of 8 (Table 6). Fortunately, it would be an easy sample to acquire—no sample selectivity would be required. There is no value of this sample type to the other objectives.

4.1.4. Number of rock samples desired/required

In any geological study, the number of samples required to address a well-defined problem is always a difficult issue; depending on the question, the number may lie anywhere between a single sample (e.g., an internal isochron of an igneous rock) and many hundreds, if not thousands, of samples (e.g., high-resolution climate change records). What are appropriate numbers of rock samples to address the science objectives (Table 6) and sampling priorities (Section 2.3 of this report) of the MSR Campaign?

Considerable insight has been gained from the experience of the MER mission (see references below). The Spirit rover operated for more than 6 years in a target-rich environment in Gusev Crater and examined more than 75 rocks in situ. Within this extended period of time, the rover carried out a number of reasonably well-defined “campaigns” that lasted on the order of weeks to months and were designed to investigate, in detail, major discoveries. These campaigns, in turn, provide some guidance to the number of samples that might be required for MSR.

Four relevant Spirit campaigns bear some broad similarity to the science objectives defined for the MSR Campaign:

(1) A hydrothermal system defined by the exposures of opaline silica-bearing rocks and soils in the Eastern Valley near Home Plate (Squyres et al., 2008; Ruff et al., 2011);

(2) An alkaline igneous province defined by alkaline volcanic rocks in the Columbia Hills of Gusev Crater that had sources similar to the Adirondack-class volcanic rocks found in the Gusev plains (McSween et al., 2006);

(3) The Home Plate deposit comprising a 2 m layered section of poorly sorted pyroclastics debris overlain by well-sorted sandstones thought to represent aeolian reworking of the pyroclastics (Squyres et al., 2007);

(4) The ascent of Husband Hill, during which the Spirit rover traversed a target-rich geological setting of layered rocks and encountered numerous lithologies that could be considered as “targets of opportunity” (Arvidson et al., 2006; Squyres et al., 2006).

Table 7 summarizes the details of these campaigns and the number of samples each examined in some detail in order to address the hypothesis that initiated the detailed study. Also shown is our best estimate of campaign durations. Durations are difficult to estimate because, in many cases, campaigns were initiated some time after the initial observations, multiple issues were being addressed simultaneously, unrelated traverses were interspersed, and so forth. Nevertheless, an overall conclusion from this analysis is that on the order of ∼30–35 mixed rock and soil samples would be sufficient to achieve the major science objectives of the MSR Campaign and that the time required to obtain such samples would likely take on the order of a martian year or less (Fig. 5).

OPEN IN VIEWER

FIG. 5.

Schematic diagram, based on lessons learned from the MER Spirit, illustrating the hierarchical nature of obtaining sufficient geological context information in order to select the best samples for return to Earth. Estimates of targets within reach (†) is based on number of rocks readily visible to Spirit while traversing during the first martian year. Color images available online at www.liebertonline.com/ast

OPEN IN VIEWER

Table 7.

Approximate Duration and Number of Rocks/Soils Studied in Detail during Various Science “Campaigns” of the MER Spirit Rover

Campaign	Description	Duration	Samples
Hydrothermal System	Exposures of opaline silica-bearing rocks and soils in the Eastern Valley	∼100 sols	∼10
Alkaline Igneous Province	Igneous rocks preserved on the Gusev plains and Columbia Hills	∼60 sols	7
Pyroclastic Deposit	2 m Home Plate section of altered pyroclastics and reworked aeolian sediments	∼30 sols	7
Targets of Opportunity	Examples of distinctive rock classes encountered during ascent of Husband Hill	∼200 sols	∼8–10
TOTALS		∼390 sols	∼32–34

FINDING #15: For the kinds of landing sites of interest to the MSR Campaign, the number of high-priority rock samples is estimated to be ∼30–35. This reaffirms a key finding of the ND-SAG, who recommended 28 rock samples as a minimum number.

4.2. Considerations related to the number and type of regolith samples

4.3. Considerations related to the number and type of gas samples

5.1. Establishing a reference landing site set

Meeting the scientific objectives of the MSR Campaign (Table 6) is dependent on whether there are places on Mars that host the desired materials for sampling and whether a rover could access and sample them. To establish the potential for at least one site on Mars to both satisfy the highest-priority MSR science objectives and have the capacity to land, we conducted a process to identify potentially viable candidates. Our goal was to produce a reference landing set, consisting of several reference landing sites, each of which could potentially meet the objectives of Table 6 and illustrate a range of scientific and engineering attributes that span the trade space of interest and could be quantitatively evaluated.

The search for reference landing sites began with a review of the ∼60 landing sites proposed for the MSL mission (Grant et al., 2011) and ∼25 additional community-proposed landing sites identified for possible future missions (originating through a 2010 Future Landing Sites call). Although the overall objectives for the MSR Campaign would differ from those of these other missions, they were viewed as a good starting point because of some overlap in science objectives and because many of the sites considered for other missions are partially to nearly completely covered by high-resolution spatial and spectral resolution data (e.g., from Mars Reconnaissance Orbiter, Mars Express, and Odyssey). We chose sites with an eye toward providing a range of characteristics for both science and engineering that could be used to help define landing and roving requirements. Sites with substantial existing image coverage were favored because such data enable meaningful engineering studies of the MSR Campaign EDL system requirements for accessing the eventual landing site.

To screen candidate sites for the MSR Campaign, we used the four threshold criteria listed in Table 8. These threshold criteria relate primarily to the inferred depositional setting and age of the rocks considered to be of highest priority for sample return and include the strong desire for the presence of igneous rocks. Although additional preliminary qualifying criteria were identified, ranging from morphological evidence of setting to the age of volcanic units to be accessed (Table 9), these were not used in the E2E analysis process. (The job of prioritizing candidate landing sites needs to be done through a community-based process, separate from the E2E process.) Because the reference sites are not intended to serve as a short list for the actual mission, the more relaxed constraints allowed us to define reasonable science and engineering criteria for the proposed mission. A more rigorous and open landing site selection process is expected to follow. We anticipate that, once formal criteria are defined, a call for candidate sites would be made to the science community, which would initiate a comprehensive site selection process based on those employed for MER and MSL (e.g., Grant et al., 2004, 2011).

For reference landing set, the E2E-iSAG chose the following (Table 9): five candidates studied extensively during the MSL landing site selection process; the MER landing site in Gusev Crater; and an additional site at a relatively high northern latitude. We assumed a landing ellipse comparable to that of MSL (∼20×25 km). Each of the sites appears to encompass all the threshold science criteria and define a latitude range of approximately 35°N to 15°S, elevations ranging from ∼ −0.5 km (Mars Orbiter Laser Altimeter—MOLA) and lower (Fig. 6), and a variety of relief (much of which was viewed as unacceptable for MSL touchdown criteria relating to hazard probability). A description of each of the reference sites follows.

OPEN IN VIEWER

FIG. 6.

Map showing landing sites considered by the E2E-iSAG analysis. Those proposed as candidate reference sites for the MSR Campaign are highlighted.

5.2. Brief descriptions of the reference landing sites

The Nili Fossae region includes a diversity of hydrated minerals rarely obtained on Mars (e.g., Ehlmann et al., 2009). Three sites are proposed in this broad region, each with a different context and the presence of igneous rocks. Nili Fossae trough is located on Hesperian volcanic flows on the floor of the Nili graben. Locations favorable to astrobiology contain phyllosilicates in both layered and massive units (Mustard et al., 2009). More diversity is accessible east of the ellipse within the ejecta breccia of the crater Hargraves, which would enable a sampling of crustal rocks, both altered and unaltered, although not in place. About 200 km east of Nili trough, Jezero Crater is a paleolake identified by two delta fans (Fassett and Head, 2005). The fans contain hydrated minerals and possibly carbonates, and are likely to have collected material from the altered highlands (Ehlmann et al., 2008a, 2008b). The crater floor contains mafic minerals that show rough textures consistent with Hesperian-age volcanic flows. South of Jezero, the contact between altered highlands and unaltered Syrtis Major plains is very straight with a small scarp of lava flows that dominate altered highlands (Mangold et al., 2007). Landing on the plains at this NE Syrtis Major site would enable sampling of a well-defined volcanic unit with a go-to astrobiological objective inside the layered material beneath the lava flows and in the altered highlands. Sulfates may exist inside the layered material, and carbonates are present on this olivine-bearing unit of the Nili Fossae region (Mustard et al., 2007, 2009; Ehlmann et al., 2008b, 2009).

The Mawrth Vallis region is another area that displays evidence for substantial alteration, mainly in the form of phyllosilicates (e.g., Poulet et al., 2005; Loizeau et al., 2007, 2010; Bishop et al., 2008), which is considered of high interest for understanding the early martian environment and its astrobiological potential (Michalski et al., 2010). The site proposed is located northwest of the one that was under consideration for the MSL site selection in the final four list. In addition to layered materials containing Al clays and Fe/Mg clays that crop out inside the ellipse, it also contains igneous materials, some that are likely present inside the ellipse, and some Hesperian-age volcanic flows in the western side of Oyama Crater that are reachable after a 20–30 km long traverse.

Eastern Margaritifer Terra is located in the channeled highlands of Noachian age, south of Meridiani Planum. The small basin where the ellipse is located contains phyllosilicates and possibly chlorides (Osterloo et al., 2008). The sequence of units exposed by erosion in this basin has an unaltered, basaltic unit at the top of the sequence that overlies a phyllosilicate-bearing unit, which in turn overlies a chloride-bearing unit at the base (Christensen et al., 2008). These units would be reachable inside the ellipse after a short traverse out of the ellipse to the east.

Gusev Crater has proved to be mineralogically diverse despite an apparent lack of evidence for the paleolake that motivated its selection as a MER landing site. Within the Columbia Hills in the center of Gusev, Spirit encountered soil and outcrops of nearly pure opaline silica, a clear manifestation of hydrothermal processes (e.g., Squyres et al., 2008) and entirely consistent with a hot spring or geyser origin (Ruff et al., 2011). Outcrops containing as much as 34% by weight Mg-Fe carbonate were identified and perhaps represent another manifestation of hydrothermal processes (Morris et al., 2010). Sulfate-rich soils are yet another indication of the role of water. In addition to a range of igneous rocks within the Columbia Hills, they are surrounded by early Hesperian, olivine-rich flood basalts similar to lunar mare (Greeley et al., 2005). Gusev Crater provides both well-defined igneous units and a diversity of mineralogy and rocks formed in environments of astrobiological interest.

Ismenius Cavus is a 60 by 90 km elliptical trough located in the Ismenius Lacus region. It is the most challenging site proposed here, because it is at a high latitude (34°N) and because mafic minerals are currently only observed as sands, even if volcanic plateaus surround the circular trough. The interest in this site is that it shares the presence of a paleolake attested to by three delta fans at the same elevation, phyllosilicates on layered deposits on the floor of the trough, and midlatitude glaciers on its side (Dehouck et al., 2010). The latter may open a unique opportunity to collect material from the Early Mars period and from Amazonian ice deposits in the same mission.

The seven sites identified here, in addition to containing reachable igneous rocks, have been chosen to be complementary and representative of different types of astrobiological interest: two paleolakes (Jezero and Ismenius Cavus), one ancient basin (E. Margaritifer), three hydrothermal sites (Nili Fossae trough, NE Syrtis Major, and Gusev), and one altered crust (Mawrth) with a variety of alteration or secondary minerals, or both (phyllosilicates, sulfates, chlorides, carbonates).

5.3. Implications of the reference landing set for the major EDL and mobility parameters

The reference landing set (Table 9) has some important lessons for thinking about landing site-related planning.

• It is not easy to find sites where relatively unaltered igneous rocks are found close to sites selected for astrobiology where the rocks have been strongly altered by water. Retaining the igneous rock objective in the threshold criteria, therefore, likely would require rover mobility and lifetime to be large enough to access targets outside the landing ellipse. For example, among the seven reference sites proposed, most are likely “go-to” sites either for igneous rocks or astrobiology and would therefore require significant driving distance (i.e., ∼20 km) and rover lifetime on the surface (assumed to be at least 1 martian year, although not evaluated in this study).

An important caveat is that the seven reference sites are not known to be safe for the engineering criteria of the MSR Campaign (which would involve two landings at different times). The total number of potential sites would depend on how many of the sites in the reference set are ultimately judged to be acceptable to both science and engineering, and the reasons that any of them are rejected. However, the number of potential sites may never exceed 20–25. Consequently, all capabilities that could be added to the landing system to access additional sites should be considered.

Lastly, latitude limitations on the landing site, such as the −5°/+25° constraint on the original ESA ExoMars rover (and the E2E-iSAG Charter), would also impact the number of potential sites that are responsive to all proposed objectives. Many southern latitudes below −5° display considerable evidence for aqueous alteration and putative lacustrine landforms, including sites such as Gusev Crater (−14°) and two of the final four MSL sites (Eberswalde and Holden at about −25°). Northern latitudes in excess of 25° (25–40°) would also be of interest but may be less important than an opening of southern latitudes, especially because of the limited area of highland terrains above the +25° latitude. Thus, an enlargement of the southern latitude range, at least to −15° and perhaps to −25°, would help to introduce a more significant number of candidate landing sites into consideration.

6.1. Analysis flow for rock and regolith samples

As per experience gained by the global curatorial community through processing of extraterrestrial samples returned by Apollo, Luna, Stardust, Genesis, and Hayabusa, the analysis of rock and regolith samples returned from Mars would be expected to follow a four-phase process:

I. Preliminary examination

A significant part of the scientific interest in martian samples is because of their potential to contain evidence of past or present martian biology. For this reason, the second category above is far more important than it was for Stardust, Genesis, and Hayabusa. Steps I and II would therefore need to be done in a suitable containment facility, which for planning purposes has been referred to as the Sample Receiving Facility (Rummel et al., 2002; Beaty et al., 2009; and references therein). Depending on the outcome of the planetary protection assessments, it may be desired or required that Steps III and IV would also need to be done in containment. Another possible way to set up this planning, if the sample demands of the planetary protection assessments are too large, is to bypass Step II and carry out Steps III and IV in containment, treating the samples as if they were hazardous (although this approach would likely restrict the scope of experiments that could be performed, which would have significant negative consequences for the scientific investigations).

6.2. Implications for sample sizing: rock samples

The size of individual samples is a key parameter of any sample return mission. Requirements for sample mass, volume, shape, and dimension feed directly into the design of the sampling system(s). Sample dimensions and quantity constrain the design of the cache and the attributes of the return flight system. Total sample mass, along with the mass of containing and supporting hardware, imposes stringent constraints on the MAV and the flight elements that would return the samples to Earth.

Numerous studies of MSR over the years have addressed the issue of sample mass, with varying degrees of scientific rigor. The most recent analysis, by the MEPAG ND-SAG (2008), concluded: “A full program of science investigations would likely require samples of >8 g for bedrock, loose rocks and finer-grained regolith. To support required biohazard testing, each sample requires an additional 2 g, leading to an optimal size of 10 g.” Although Table 4 of MEPAG ND-SAG (2008) did include the concept of a reserve sample, it did not explicitly account for replicate analyses, and the amount of sample mass needed to carry out certain high-value investigations was underestimated.

Rock sample sizing was estimated with the following assumptions, and summary results are shown in Table 10. The interested reader is referred to MEPAG E2E-iSAG (2011) for more details. Because returned sedimentary and igneous rocks would be used differently to achieve the proposed scientific objectives, the investigations and sample mass/investigation are presented separately in Table 10 for these two rock types. This analysis shows that sample sizing of 15–16 g/sample is optimal.

• All required analyses may be done on any sample.

6.3. Implications for sample sizing: regolith samples

For each regolith sample, investigations in each of the following areas are thought to be necessary to achieve the scientific objectives proposed (key input acknowledged from Mike Hecht and the Granular Materials Focus Group; see also MEPAG E2E-iSAG, 2011, Section 3).

• Physical properties (Shape, texture; Size distribution; Diffusivity, tortuosity, permeability; Surface area and porosity)

While there has been great progress in analysis techniques for minute samples, grain diversity and handling considerations suggest that a minimum of 5–25 mg would be necessary for each investigation described above except the search for viable microbes, which would require at least 300 mg per sample. As discussed above in Section 6.1.4, it is a high priority that the samples be large enough to support three redundant measurements for each investigation. The amount of regolith sample that would be required to support planetary protection testing is hard to estimate but should not be assumed to be less than 10%, as per arguments discussed above in Section 6.1.2. Maintaining soil samples in a pristine state during transport and analysis by different laboratories would be more challenging than for rock and gas samples, and greater loss or compromise of samples is expected. Accordingly, for regolith samples, it is recommended that 67% of the sample be retained for future analysis.

Adding up all the above factors leads to a minimum recommended sample size of about 7.5 g. Some reduction of this total may be possible by using a carefully structured program of sample reuse (Finding #26). Granular materials have lower density than rocks, particularly after they have been disturbed. For example, the average density of the lunar regolith is 1.5 g/cm³ in the top 15 cm (Heiken et al., 1991). The bulk fines delivered to the X-ray fluorescence spectroscopy instrument on Viking had a density of 1.10±0.15 g/cm³ (Clark et al., 1977). Dust is deposited on the surface with an estimated bulk density of 0.95 g/cm³ (Moore, 1991). Assuming a density of 1.15 g/cm³ for disturbed soil, a 7.5 g sample would need a sample tube with a volume of at least 6.5 cm³. If collecting samples of this size were overly difficult, in the judgment of the E2E-iSAG, useful science could still be achieved with samples as small as 1 cm³ (by reducing the archived fraction to 40%, reducing redundancy to 2, using the minimum quantity for each measurement, etc.), but this is not recommended.

6.4. Implications for sample sizing: atmospheric gas samples

A detailed analysis of how the gas research community would process a returned martian gas sample is presented in Section 4 of MEPAG E2E-iSAG (2011). The minimum quantity of gas needed to achieve all high-priority objectives would be 1.9×10⁻⁵ mol, which is equivalent to ∼50 cm³ at Mars ambient temperature and pressure (see Section 4 of MEPAG E2E-iSAG, 2011). This calculation is based on the following assumptions:

• Fifty percent of the returned gas would be saved for the future.

If it were not possible to return enough gas to measure ¹²⁴Xe and ¹²⁶Xe, the next least-abundant species of high scientific interest would be ¹²⁸Xe and ⁷⁸Kr. The minimum sample size needed for those determinations (as well as all species more abundant than this), with the same assumptions as above, would be about 1.0 cm³ at Mars ambient temperature and pressure, which is a factor of 50 reduction from the above. However, ¹²⁴Xe and ¹²⁶Xe are important for the following reasons. Xenon is one of the most important elements for interpreting martian atmospheric formation and evolution. Its nine isotopes would allow us to unravel the starting composition of the atmosphere and processes that later altered its composition. The importance of ¹²⁴Xe and ¹²⁶Xe is as follows: (i) They are the rarest isotopes and, therefore, have been measured in meteorites with the largest uncertainties. (ii) In meteorites, there are large cosmogenic effects (up to 90%) because of their time spent in space, further increasing the uncertainties. (iii) Many other Xe isotopes have other sources: for example, ¹²⁹I→¹²⁹Xe, ²⁴⁴Pu fission→^131–136Xe (see review by Swindle, 2002). ^124,126Xe have no other contribution (apart from cosmogenic, which however should be negligible in the atmosphere) and would therefore be valuable to reveal the extent of mass fractionation of the starting composition and thus the degassing history (Pepin, 2006).

The development of multicollection mass spectrometers for noble gas analysis should reduce the required gas amount for a precise Xe analysis in the future. Note also that making gas sample allocations to only three laboratories would be a minimum credible plan—returning a substantially larger sample than this would be greatly welcomed by the science community. If a simple compressor were included, this quantity of gas could be packaged in a volume of 5 cm³ at a pressure of 10× Mars ambient or a volume of 0.5 cm³ at a pressure of 100× Mars ambient.

7.1. Observations required to understand geological context

As described in Section 3.2, to obtain the martian sample suites most suitable for further analysis on Earth, it would be necessary to make a wide range of field observations. These field observations would be essential to guide the sampling process, and they are critically important when it comes to interpreting the results of sample analyses on Earth. For example, in a sedimentary sequence on Earth, a field geologist typically acquires an overview of the area by examining rock types, textural features (e.g., grain size and sorting), bedding characteristics (contacts, lateral and vertical variations in thickness, extent geometry), stratigraphic relationships, and so forth in order to interpret the local geology. On this basis, a sampling strategy could be devised to answer particular problems. Such observations are essential to understanding what a sample represents. Field geology on Mars would be no different, as demonstrated by our experience with the MER mission (e.g., Squyres et al., 2004a; Grotzinger et al., 2005; McLennan et al., 2005). Accordingly, it would be necessary to analyze the geological setting at the landing site in order to identify where to sample and to establish the geological context of the samples to maximize their science value.

Accordingly, to successfully acquire the most important available samples within a reasonably well-defined geological context, and to acquire the necessary contextual information to enable the highest-priority returned sample science objectives to be met through future sample analyses, the following field capabilities would be required:

(1) Ability to detect and correlate variations in mineralogy, chemical composition, textures/structures (at micro-, meso-, and macroscale) in outcrops;

Organic geochemistry . In addition to the measurements listed above, we conclude that the capability to detect organic material is highly desired. This reaffirms previous findings. The ND-SAG (2008; Sections IV-E, IV-F) indicated that the data would be of high scientific priority for sample selection but also said that color imagery, remote spectroscopic observations, and contact geochemical/mineralogical analyses constitute the minimum set of techniques that would be needed to optimize sample selection, which thus indicates that organic detection would not be mandatory (i.e., “required”). The MEPAG MRR-SAG (2010) presented a mission vision based on the premise that the information to make effective sample selection decisions could be made only by using fast, relatively light, arm-mounted instruments. They recommended a suite of measurement capabilities (see their Section 6), including organic geochemistry, but without distinguishing “required” from “desired.” The MEPAG MRR-SAG (2010) study flowed into the NRC (2011) concept for MAX-C, which included a potential deep-UV instrument (for detecting organics), but this was a single-point design not based on requirements.

Samples that may contain organic material would be of high interest to Objective A1. Thus, information about the presence or absence of organic matter would be valuable to the sample selection process. However, since there are many other reasons to justify the return of samples, and we do not know whether martian samples accessible to our sampling systems would contain organics in a concentration or chemical form that is detectable by in situ instruments, establishing the detection of organics as a prerequisite to caching a sample is both unjustified and unwise. By analogy with hydrocarbon “source rock” studies on Earth, collecting materials that have a high likelihood of preserving organic compounds or any kinds of potential biosignatures is a viable strategy. Examples of such rocks include carbonate, phyllosilicate, or silica-rich materials. Color and textural features, such as laminae or stromatolite-like layers, would also inform recognition.

If organic detection equipment was included in the mission, then it should represent an analytical step intermediate between in situ visual inspection and ex situ comprehensive analysis. Such a step would inevitably reduce risk in successful target rock selection for astrobiological purposes. Yet, as an intermediate stage of analysis, the lack of conclusive organic matter detection by an in situ instrument on an apparently otherwise suitable rock would not preclude sample selection, given that more exacting tests would await back on Earth.

7.2. The Pasteur payload

As of this writing, ESA and NASA are pursuing a formal collaboration for the exploration of Mars. This collaboration is focusing initially on the 2016 and 2018 mission opportunities. However, it is anticipated that the collaboration would continue across several launch opportunities through to the return of samples from Mars and even beyond. The charter of the E2E-iSAG indeed initially was predicated on the assumption that two rovers (MAX-C and ExoMars) would be delivered together to the same landing site in 2018.

In response to budgetary constraints and other ESA/NASA considerations, the E2E-iSAG was asked in May 2011 to consider additional assumptions about the proposed 2018 rover in making our recommendations, which bear most directly on the nature of instrumentation that such a rover would need to support the MSR Campaign. These assumptions are summarized as follows:

(1) The mission would consist of a single joint rover to be delivered by the MSL skycrane system. (For the purpose of this report, we refer to this as the 2018 joint rover.)

Accordingly, the discussion below incorporates these added assumptions/requests to the degree possible that is also consistent with the original charter of the E2E-iSAG.

For reference, the currently planned instrumentation of the current ExoMars Pasteur payload is summarized in Table 11. The science payload would consist of a mast-mounted panoramic camera system (PanCam), a body-mounted ground-penetrating radar system, a (2 m deep) drill-mounted microscopic imager and IR spectrometer, and an Analytical Laboratory Drawer (ALD) consisting of five instruments capable of mineralogical, geochemical, and organic geochemistry measurements. In its configuration as of May 2011, the Pasteur payload did not include any robotic arm-mounted instrumentation.

OPEN IN VIEWER

Table 11.

The Pasteur Payload

7.3. Measurement needs of the proposed 2018 joint rover in addition to those of Pasteur

7.4. Sample collection and preservation system

Integrated concepts for core sample acquisition and caching consistent with a potential application to the mission objectives described in this report have been developed and either published (e.g., Collins, 2009; Backes et al., 2010, 2011) or attributable to private corporations in the public domain (e.g., ASI, Honeybee, Swales). The concept would utilize a 5-degree-of-freedom arm to deploy and manipulate a rotary percussive coring tool. The coring tool itself would provide coring, core break-off, core retention, as well as bit capture and release for bit change-out. A sample would be acquired directly into a single-use sample tube within the coring bit, and bit change-out would be used to transfer the sample tube to the caching subsystem where it would be sealed and stored. The sample storage canister, containing the individual sample tubes, could be left on the rover or deposited on the surface for later pickup by a subsequent mission.

As discussed above in Section 2 of this report, it is important that the sample acquisition system have the ability to drill into loose rocks, in addition to outcrop. This raises follow-up questions that will need to be considered by a successor panel, such as how small the rocks could be to avoid their moving when force is applied. In addition, the ability to deploy the corer to a wide range of targets away from the rover body, and against rover hazards, would be a key capability.

This particular caching subsystem concept is referred to as the Sample Handling, Encapsulation, and Containerization (SHEC) subsystem. Bit change-out and sample caching are combined in the design. There is one opening in the SHEC subsystem design; the coring tool interface port for transferring a coring tool bit. The tool bits are stored in bit holders on the bit carousel. A 2-degree-of-freedom transfer arm internally transfers sample tubes between bits on the bit carousel, plugs on the sample carousel, plugging station, and tube chambers in the sample canister. This system has the ability to switch 25% of the samples if higher-priority samples are found later in the mission (Section 3.3). The sample canister is in the center of the sample carousel and could be removed from the top of the SHEC by the rover arm; this would enable either deposition on the surface for later pickup by a rover from a possible future mission or direct removal by that future rover.

7.4.1 Sizing the sample cache

Previous studies of MSR missions have mostly assumed a returned sample mass of about 500 g based on trades of sample mass against the MAV and MSR-O projected capabilities and costs (e.g., Price et al., 2000). In this study, we estimated the minimum sample mass to be returned by examining how large each sample should be to do the desired analyses and how many different samples would be needed to sufficiently exploit the scientific potential of a particular landing site. The relationship between mass/sample and number of samples, and how these two factors contribute to the overall size of the returned sample collection, is shown in Fig. 7.

OPEN IN VIEWER

FIG. 7.

Relationship between sample size and sample numbers, including several possible packing scenarios. Color images available online at www.liebertonline.com/ast

For rock samples, the optimal mass/sample (derived in Section 6 of this report) is 15–16 g/sample, with minor differences between sedimentary and igneous rocks. A comfortable number of rock samples, as discussed in Section 6 of this report, is about 30–35. The product of these two figures is close to 500 g. If smaller numbers of larger samples were collected, there would be a significant risk that too few samples would be collected to characterize the landing site. On the other hand, if larger numbers of smaller samples were collected, there may be insufficient mass to carry out the analyses on Earth, and the amount of time needed to collect the samples may add inordinate amounts of time to required mission life.

The density of the shergottite and nahklite meteorites (martian igneous rocks for which we have data) averages 3.01 and 3.11 g/cm³ (± ∼7%), respectively (Britt and Consolmagno, 2003; Coulson et al., 2007; Macke et al., 2011). The bulk density of the sedimentary rocks on Mars has not been measured, but a starting place for discussion is 2.25–2.5 g/cm³ [estimated by using the mineralogies reported in McLennan and Grotzinger (2008) and Ruff et al. (2006) and the porosities reported in Perl et al. (2007)—Budney, written communication, 2011]. For reasons of both porosity and grain density, however, samples with density higher or lower than this could be encountered. While noting these uncertainties, the sample masses described above, if these densities are used, would be equivalent to volumes of ∼5.3 and ∼6.7 cm³ for igneous and sedimentary rocks, respectively. Note that in a practical sense it is impossible to pack a rock sample into a sample tube with 100% efficiency. Thus, the sample tubes would have to be larger than the above volumes in order to achieve the amount of sample mass required. The decision on sample tube sizing needs to be guided by scientific considerations, experimentation in a sampling test bed, and margin planning considerations (see MEPAG E2E-iSAG, 2011).

For regolith samples, a minimum volume of approximately 6 cm³ for each regolith sample would be required (Section 6 of this report). Depending on porosity, this is approximately equivalent to ∼8 g, although with high uncertainties.

An important conclusion is that the sample size for regolith and rock samples would be approximately equal in a volume sense. This means that it may be possible to use the same sample tubes for both types of samples. Using a standard sample tube size would have important benefits to engineering.

FINDING #32: The collection should be sized to a total sample mass of about 500 g.

8.1. Summary

8.2. Recommendation for future work

During the deliberations of the E2E-iSAG, it became apparent that there is a considerable amount of additional research and other effort required in preparation for the MSR Campaign. These can be conveniently divided into tasks best addressed by MEPAG and by NASA/ESA at both the programmatic level and in the area of research/development.

A2.1. Charter assumptions

A2.2. Methodology and approach

Introduction

Planetary protection encompasses both protecting Mars from terrestrial contamination (forward protection) and protecting the Earth from potentially harmful materials returned from Mars (backward contamination). A major concern of forward contamination is to limit introduction of any terrestrial materials onto the martian surface that might compromise detections of any extinct or extant life-forms or traces of any prebiotic chemistry by future missions. This is accomplished mainly through internationally agreed-upon controls on sterilization and bioburden that depend among other things on the mission type and target body (COSPAR, 2011). The main concern of back protection is to contain safely and usefully any returned samples until they are demonstrated to be harmless.

Information

When the returned samples are evaluated in Earth laboratories, the needs of both planetary protection and scientific research can only be met by making measurements on the samples. Since the quantity of sample material available would be highly limited, it would be crucial to develop an understanding of the measurements that would be in common to planetary protection and science. Although planetary protection and science may have different needs for accuracy and precision in returned sample measurements, it should be possible to determine which one has more stringent requirements for certain classes of information and to produce at least some of the data in a coordinated way. This would be an important strategy to conserve sample mass. Examples of data of relevance in both arenas might include mineralogy, texture, and organic carbon concentration and speciation. There would also be data required by both planetary protection and science that are not in common.

The justification for the mission is to conduct exciting and cutting-edge science. If this cannot be achieved, the mission should not be flown. The goals of planetary protection are also critical, and if they cannot be met, the mission also should not be flown. Because of engineering considerations, the amount of sample returned to Earth would be quite limited. Thus, there is very high pressure for both science and planetary protection to learn how to achieve their purposes with as small a quantity of sample as possible. Finding ways to minimize destructive sample testing and to maximize sample reuse (so that sample material from an initial use could all, or in part, also be used for a subsequent use) would be essential. Exactly how the sample-related demands between science and planetary protection would be balanced is deferred to future panels, but we encourage those panels to work cooperatively together.

Issues of sample size and diversity

As discussed at length in this report, to have the best chance of finding the signs of past or present life, sample variety must be maximized. Multiple small samples (perhaps in the 2–5 g range) would almost certainly be more useful than a few large (50 g) ones. However, the smaller the samples, the more difficult it may be to render judgments on the safety of individual samples and the collection as a whole. This trade-off between the desirability of heterogeneity from the point of view of science and the possible difficulty of dealing with heterogeneity by planetary protection must be studied and a proper balance found.

It is possible that traces of life on Mars may be very rare and heterogeneously distributed, and the concentration of organics in the soil may be very low. Thus, it is essential to avoid biological contamination as well as organic contamination of Mars and of the collected samples. Some science-driven limits for organic and nonorganic contaminants are given by Mahaffy et al. (2004) and MacPherson et al. (2005). The tables in these reports need to be updated to reflect more modern understanding of the needs and opportunities in the area. The proposed missions of the MSR Campaign would ultimately be flown with a single set of contamination control requirements, so it is mandatory that science and planetary protection converge on acceptable values.

The most recent and comprehensive summary of how returned martian samples might be treated on their return to Earth from a planetary protection perspective is that by Rummel et al. (2002). Analytical techniques have, however, evolved significantly since this study, and an update of this document is appropriate. Although the return of martian samples to Earth may be decades away, it is vital that sample requirements be understood in the early design stages of the sample acquisition system. We need to know what requirements planetary protection protocols would impose for such things as sample mass, bioburden limits, nonbiological contamination limits, whether and what type of bioassays are needed, and so forth. We therefore recommend that, as the science requirements for sample return are being formulated, a planetary protection panel be established to update or replace the Rummel et al. (2002) Draft Test Protocol in order that the sample acquisition system could be designed to satisfy both the science and the planetary protection requirements.