Color and Albedo Characteristics

Martian dust was first observed in the planet’s atmosphere via telescopic observations dating back to the 19th century, although an interpretation as dust was not universally accepted among astronomers. Many astronomers reported seeing “yellow” features that are now assumed to be related to dust storms (cf. Martin & Zurek, 1993). Dust storms that start in a localized region and then spread around the globe to become planet-encircling (Figure 1) were first confirmed by telescopic observations in 1956 and have been encountered by spacecraft at Mars in the decades since (Martin & Zurek, 1993). The Mariner 9 orbiter arrived at Mars on November 14, 1971, and encountered the planet fully veiled by atmospheric dust. The imaging campaign was delayed until January 2, 1972, at which point the dust had cleared sufficiently to observe surface features.

Open in new tab

Figure 1. An example of a planet-encircling Martian dust storm as viewed from Earth with the Hubble Space Telescope.

Source: NASA Science Mars Exploration Program.

Changing albedo patterns on the surface also were observed during the telescopic era but not confirmed as due to redistribution of surface dust until the spacecraft era, which began with the flyby of the Mariner 4 spacecraft in 1964. The origin of a phenomenon known as the wave of darkening was the subject of investigation even into the spacecraft era. In telescopic views, there appeared to be a darkening of the Martian hemisphere experiencing local spring and summer. The possibility that this was due to the emergence of vegetation or biology was long considered a viable hypothesis and was tested against the alternative hypothesis of dust transport as recently as 1967 (Pollack et al., 1967). Sagan and Pollack (1969) developed a quantitative model for the movement of dust and sand particles on Mars that provided strong support for the dust transport hypothesis as well as predicting the role of dust devils in the lifting of dust. With the arrival of the Mariner 9 orbiter in 1971, the phrase “wave of darkening” was recognized as a misnomer for the observable albedo variations it documented, and a role for biology was strongly disfavored although not excluded (Sagan et al., 1972). Combined with results from the two Viking orbiters, albedo variations outside the polar regions on Mars were attributed solely to eolian effects, and the concept of a wave of darkening was rejected (Veverka et al., 1977).

This combination of observations from two different orbiting missions separated by 4 years allowed for further investigation of variable albedo features, as documented by Veverka et al., (1977). Among these were streaks extending downwind from impact craters, both lighter and darker than adjacent surfaces. The lighter crater streaks were mostly unchanged over this period and were recognized as indicators of wind direction and attributed to an accumulation of dust that was protected from removal in the lee of the crater. Darker crater streaks were short-lived in comparison, attributed to fallout of dust from dust storms. Other albedo variations were noted during this period and generally attributed to the lifting or deposition of dust.

Thermal Characteristics

A relationship between albedo and particle size of surface materials at the global scale was first quantified by Kieffer et al., (1977) using data from the Viking orbiters’ infrared thermal mapper instruments. Albedo in this case is bolometric normal albedo defined as the fraction of total incident solar radiation not absorbed by the surface, determined for normal incidence and emission. Particle size was obtained from thermal inertia values, which were derived from predawn brightness temperature measurements. Thermal inertia is defined as I = (kρ‎c)^1/2, where k is thermal conductivity (W m^–1 K^–1), ρ‎ is density (kg m^–3), and c is specific heat (J kg^–1 K^–1), yielding values with units of J m^–2 s^–1/2 K^–1. Among these parameters, the one most responsible for variations in thermal inertia is conductivity, which is closely related to particle size. A strong anticorrelation between thermal inertia and albedo values was identified by Kieffer et al., (1977). Because the lowest thermal inertia values represent particles ≤100 μ‎m in diameter, high albedo regions are correlated with the smallest particles on Mars. Kieffer et al., (1977) suggested that the strong anticorrelation allows albedo alone to serve as a proxy for thermal inertia, especially in the case of high albedo values (>0.27), which serve to identify the smallest particle sizes.

Three distinct provinces were identified in northern equatorial regions that have high albedo and low thermal inertia, referred to as Tharsis, Arabia, and Elysium after their general geographic vicinity (Figure 2). The thermal inertia values for these regions were refined by Palluconi and Kieffer (1981), resulting in the determination that they are covered by material with an effective particle size of <40 μ‎m. The boundaries of these provinces are relatively sharp but lack a corresponding morphologic or geologic expression (Zimbelman & Kieffer, 1979). Christensen (1986) described them as regional dust deposits and constrained their properties using thermal, radar, and visual remote sensing observations. These observations and related considerations suggested a thickness between 0.1 and ~2 m formed over 10⁵ to 10⁶ years. Given the low likelihood that these deposits were formed for the first time within the past 10⁶ years, their young age suggests a cyclic process of deposition and removal linked to variations in orbital parameters and consequent changes in wind patterns and velocities. In this model, the present regional dust deposits could be eroded and deposited elsewhere as obliquity and eccentricity evolve over time (Christensen, 1986).

Open in new tab

Figure 2. Mars global thermal inertia map in units of J m^–2 s^–1/2 K^–1 produced with data from the Viking Infrared Thermal Mapper instrument, after Christensen and Moore (1992). Named purple and blue areas are covered in thick dust; coarsest material (including bedrock) is in red and white.

Source: Generated using JMARS.

The movement of surface dust particles by boundary-layer wind shear alone was shown to be extremely difficult based on wind tunnel experiments designed to simulate Martian conditions (Greeley et al., 1980). However, aerodynamic surface roughness was shown to be a contributing factor that could lower the wind speeds needed to entrain small particles (White et al., 1997). Furthermore, atmospheric vortices known as dust devils were long considered a likely candidate for efficiently lifting dust and larger particles, and they were first confirmed in orbital images from Viking spacecraft (Thomas & Gierasch, 1985). Although they were recognized from the surface by meteorological sensors on the Viking landers (Ryan & Lucich, 1983), dust devils were first imaged by the Pathfinder lander (Ferri et al., 2003) and first observed in motion with the Spirit rover (Greeley et al., 2006). They are now recognized as common and widespread features of the atmospheric boundary layer that contributes to the persistently dusty atmosphere.

The relationship between surface dust and albedo suggested from its strong anticorrelation with thermal inertia (Kieffer et al., 1977) was further supported with mapping of a particle-size-dependent spectral feature. Silicate particles in the size range <~40 μ‎m produce strong features in thermal infrared spectra that are not present in coarser particles. Known as transparency features, they serve as indicators of fine particulate materials such as dust. Ruff and Christensen (2002) developed a dust cover index (DCI) based on the average emissivity of a transparency feature from 1350 to 1400 cm^–1 in spectra from the Thermal Emission Spectrometer (TES) on the Mars Global Surveyor (MGS) orbiter. Global DCI values are strongly correlated with TES albedo values and more sensitive to dust cover than thermal inertia, so a few 10s of microns of dust cover can be identified compared to ~10 cm for thermal inertia data.

Composition

The distinctive composition of Martian dust allowed it to be identified and mapped globally. Visible to infrared (~0.3–4.1 μ‎m) spectral observations from Earth-based telescopes were the first means of obtaining compositional information. An early summary of these observations concluded that the bright regions of Mars are spectrally similar to each other and to dust in the atmosphere (Singer et al., 1979). A ferric oxide component was suggested at 6–8 wt% to account for the strong Fe³⁺ absorptions below 0.75 μ‎m and interpreted as evidence that the material in bright regions was more altered than that of dark regions. A weathering product of basaltic glass on Earth known as palagonite was recognized as a good candidate for bright material based on similar visible to near infrared (VNIR, ~0.3–2.6 μ‎m) spectral characteristics (Singer, 1982). The amorphous character of palagonite provides a better spectral match than crystalline phyllosilicate and oxide candidates and led to the generic description nanophase ferric oxide (npOx) as the component(s) that provides the reddish hue of Martian dust (Morris et al., 1990).

Moroz (1964) was the first to identify a feature in the 3-μ‎m region attributable to water bound in a mineral crystal structure and associated with bright regions. He suggested limonite as a candidate, which is a term for the mixture of fine-grained iron oxides and oxyhydroxides produced from the alteration of iron-rich minerals or iron ores. Although limonite is not a distinct mineral phase, its spectral similarity to Martian bright regions is notable and broadly consistent with the subsequent terminology of nanophase ferric oxides.

With the advent of visible to infrared spectrometers on spacecraft at Mars, additional compositional information on Martian dust was obtained. The 3-μ‎m feature attributable to bound water was clearly confirmed in spectra from the Infrared Spectrometer for Mars on the Soviet Phobos 2 spacecraft (Murchie et al., 1993), which operated for 2 months in 1989. It also was observed over dark regions, although weaker, and attributed to dust mixed with dark gray materials. No single mineral phase was identified to account for this feature, but possibilities included hydrated carbonates, ferric oxyhydroxides, and hydrated sulfates (Murchie et al., 2000). Spectra of dusty regions from the Observatoire pour la Mineralogie, l’Eau, les Glaces et l’Activité (OMEGA) instrument on the European Space Agency’s Mars Express orbiter lacked features near 1.4 and 1.9 μ‎m typical of hydrated minerals, leading Bibring et al., (2006) to conclude that Martian dust is composed of anhydrous ferric oxides.

Thermal infrared (TIR) spectrometers, collectively spanning the spectral range from 2000 to 200 cm^–1 (5–50 μ‎m), provided a notably different view of Martian dust. The Mariner 6 and 7 spacecraft included the Infrared Spectrometer (IRS) instruments that flew by Mars in 1969, followed by the Mariner 9 orbiter in 1971 with its Infrared Imaging Spectrometer and the MGS TES instrument in 1997. These instruments revealed a strong broad feature at ~1,050 cm^–1 attributable to one or more silicate phases present in atmospheric dust (Christensen et al., 1998; Hanel et al., 1972; Herr & Pimentel, 1969). Due to the different sensitivities of TIR spectroscopy compared to VNIR, TIR spectra demonstrated that Martian dust is dominated by silicates, likely framework silicates found in plagioclase feldspar and perhaps zeolite (Hamilton et al., 2005; Ruff, 2004). The ferric oxide phases that so dominate the VNIR spectra are not evident in the TIR because of their substantially lower abundance. However, indication of a bound water phase(s) based on the 3-μ‎m feature was evident in IRS spectra (Herr & Pimentel, 1969) and confirmed using the 6-μ‎m fundamental vibration of H₂O evident in TIR spectra from the TES instrument (Ruff & Christensen, 2002). Evidence for a small amount (2–5 wt%) of Mg-rich carbonate also was identified in TES spectra of dusty regions (Bandfield et al., 2003). Tentative identification of a comparable amount of carbonate in the dust was made by Pollack et al., (1990) using TIR spectra obtained from an airborne observatory on Earth. They also suggested the possibility of a sulfate component at 10%–15%.

Spacecraft missions to the surface offered additional means for investigating the composition and other properties of Martian dust. The Viking landers arrived in 1976 equipped with a retractable boom (arm) and collector (scoop) that could deliver regolith to onboard instruments. Neither lander was located in one of the classical bright regions, so the available soils (fines) were mixtures of darker, coarser grains and finer, brighter grains that presumably included dust. A key result from the X-ray florescence spectrometers was the remarkable similarity of the fines between the widely separated (~6,500 km) landing sites, indicative of globally mixed material, as documented by Clark et al., (1982) at the end of the mission. The elemental composition is indicative of mafic or ultramafic source materials to which S and Cl have been added, which were interpreted to result from evaporative salts. In addition, both landers carried magnet arrays, one of which was located on the lander deck to allow for the collection of any airborne particles. Both landers clearly had accumulated particles on these magnets, which was evidence for one or more magnetic phases in the dust (Figure 3). The iron oxide mineral maghemite was favored, probably as a component of composite particles (Hargraves et al., 1979).

Open in new tab

Figure 3. Viking Lander 2 calibration targets, one of which includes a magnet that has collected dust (arrow). Image cropped from larger mosaic.

Source: Wikipedia Commons.

The Sojourner rover deployed by the Pathfinder lander in 1997 carried an alpha proton X-ray spectrometer (APXS). Although the rover provided mobility, none of the soils measured by the APXS represented dust in isolation. However, these soils had notable similarities to those measured with Viking, but probably included inputs from the local rocks (Rieder et al., 1997). The Pathfinder lander also carried magnet arrays, which had a range of strength across five targets in order to help identify candidate magnetic phases. Unlike Viking with its sample delivery activity near its deck magnets, the Pathfinder magnets could only capture atmospheric dust. The results were similar to those obtained with Viking and helped further support the likelihood of composite particles containing one or more magnetic phases plus an assumed silicate phase, perhaps a phyllosilicate (Hviid et al., 1997). The favored interpretation was composite particles with ~6% maghemite as the dominant magnetic phase combined with claylike particles.

The Mars Exploration Rovers (MER) Spirit and Opportunity arrived in 2004 carrying an APXS as well as other instruments that could characterize Martian dust. The rovers’ mobility and extended duration made it possible to investigate dust across kilometer-scale distances and multiyear timescales. The Miniature Thermal Emission Spectrometer (Mini-TES) on each rover clearly documented the uniform TIR spectral character of dust both within and across both landing sites (Christensen et al., 2004). Notably, the Mini-TES spectra of dust matched those of the TES instrument in orbit, providing ground truth and a robust demonstration of the spectral uniformity across the globe (Yen et al., 2005). Measurements by the APXS and Mössbauer spectrometer (MB) included “bright dust” in small patches at the Meridiani Planum site and in a commonly occurring ~1-mm-thick layer covering dark soil deposits in Gusev crater, which also showed the uniform composition of dust across both sites (Yen et al., 2005). The measured SiO₂ abundance of ~45 wt% and other major elements indicated a generally basaltic composition but with substantial alteration as shown by S and Cl and oxidized Fe. Combined results from MB and APXS showed a strong correlation between a nanophase ferric oxide component and S (Yen et al., 2005), an indication of the relationship between alteration and S enrichment.

The MERs were equipped with three different magnets intended to capture atmospheric dust particles. The sweep magnet was designed to deflect all magnetic particles away from a circular central area, leaving only non-magnetic particles. Color Panoramic Camera (Pancam) images and spectra showed that particles accumulated only in an annulus around the central circle, with no particles detectable in the center, at the resolution of the camera. This provided evidence that atmospheric dust particles must be composites of a magnetic phase and other non-magnetic phases (Bertelsen et al., 2004). The other two magnets, located within reach of the arm-mounted instruments, were designed to capture atmospheric dust across a circular area large enough to be measured with the APXS and MB instruments (25-mm diameter). Key results were obtained from the MB, which identified Fe-bearing mineral phases. It showed evidence for olivine and pyroxene in the dust (non-Fe-bearing silicates are undetectable) and that the dominant magnetic phase is magnetite, not maghemite as previously suggested (Goetz et al., 2005). Given that magnetite is not visibly reddish, this showed that the characteristic hue of Martian dust does not arise from its dominant magnetic phase but instead probably comes from nanocrystalline hematite or possibly oxyhydroxides, including nanophase ferric oxides (Goetz et al., 2005).

The Phoenix lander mission, which operated in 2008, also included a magnet experiment with improvements over previous versions. In this case, there was evidence for a small amount of dust accumulation in the central area of the sweep magnet, an indication that there are non-magnetic or very weakly magnetic dust particles in addition to the strongly magnetic composite particles (Drube et al., 2010). Soil that was scooped and delivered to onboard instruments included “red fines,” which were interpreted as belonging to a global unit comprising airborne dust and bright surface soil (Goetz et al., 2010). The soil also showed evidence for ~3–5 wt% calcium carbonate (Boynton et al., 2009), but without distinguishing between dust particles deposited from the atmosphere and coarser soil particles derived from local sources. An atomic force microscope provided the highest resolution observations yet obtained for Martian soil, showing a particle size distribution in which the clay-sized particle proportion represents an exposure to liquid water for much less than 5,000 years (Pike et al., 2011). It is important to note again that it was not possible to isolate Martian dust in these observations.

Landscape Features

Relatively dark streaks were observed within lighter material on steep slopes in some of the highest resolution images from the Viking Orbiters, but their origin remained uncertain until the advent of higher resolution images from the Mars Orbiter Camera (MOC) on the MGS orbiter, as originally presented by Sullivan et al., (2001). They documented elongate dark streaks with ~10% contrast, sharp margins and no indication of relief, commonly with lengths in the 300–1,500 m range and length to width ratios of 5:30, typically with digitate downslope ends (Figure 4). Bright slope streaks also have been observed, but they are rare. Key observations are that dark slope streaks brighten over time and new ones formed during the MGS and subsequent missions. Although the possibility that they are stains from wet debris flows from breached aquifers was considered, Sullivan et al., (2001) provided a set of observations demonstrating that they most likely are scars from dust avalanches following oversteepening of air fall deposits, akin to shallow avalanche scars in dry loose snow on Earth. The streak contrast results from the exposure of darker slope materials in the avalanche scar, and the brightening over time occurs when this material is covered with new air fall dust. The typical occurrence of slope streaks in low thermal inertia, high albedo terranes on Mars demonstrates a connection to locations with substantial dust accumulation.

Open in new tab

Figure 4. Slope streaks observed in a contrast-enhanced grayscale image from the Mars Orbiter Camera (NASA/JPL/Malin Space Science Systems).

Source: NASA Jet Propulsion Laboratory.

The Tharsis region is a low thermal inertia, high albedo terrane with abundant slope streaks that also contains landforms thought to include a type of rock referred to as duststone. Bridges et al., (2010) proposed that distinctive eolian bedforms, yardangs, and indurated surfaces among the Tharsis Montes are composed of dust aggregates held together by electrostatic charging, which ultimately become lithified through cementation by salts. Dust aggregation can explain how eolian bedforms like those that typically form from sand-sized particles could form from dust-sized particles that separately are incapable of saltation (they would be suspended). Dust aggregates referred to as agglomerates were first recognized in images from the Microscopic Imager on the Spirit rover (Herkenhoff et al., 2004), lending support to the idea of the potential for saltation of dust and consequent landforms. However, the survivability of electrostatically bound or weakly cemented dust agglomerates during saltation remains an open question. The idea of duststones outside of the Tharsis region has been suggested, including the upper portion of the mound in Gale crater called Aeolis Mons (aka Mount Sharp) (Grotzinger et al., 2012), which could possibly be explored by the Curiosity rover.

Engineering and Operational Issues

The ubiquity and abundance of Martian dust resulted in a range of impacts to the engineering and operation of surface spacecraft, the most significant of which was the impact of dust on solar-powered spacecraft. Although the first such spacecraft were the Pathfinder lander and its Sojourner rover, their relatively short operational duration of nearly 3 months, which well exceeded their planned durations, meant that dust was not a significant factor in their operation. It was the Spirit and Opportunity rovers for which dust issues were fully manifested, starting with the design of the solar arrays. Based on the drop in power experienced by the Pathfinder lander from dust accumulation on its solar arrays, those of the MERs were designed with a sufficiently large area to accommodate dust accumulation and still achieve a minimum mission duration of 90 Martian sols (Stella et al., 2005). Unanticipated was the removal of accumulated dust from the solar arrays of each rover over the course of their missions. The most remarkable example occurred on sol 420 of the Spirit mission, manifested as a substantial increase in solar array output and thought to have resulted from the clearing of dust by one or more wind gusts or dust devils (Greeley, Whelley, et al., 2006). “Dust cleaning” events, evident as increases in power output from the solar arrays, were a repeated yet unpredictable feature of both MER missions. The darker, less dusty character of Meridiani Planum compared to Gusev crater demonstrates that dust lifting must occur more frequently, which also was evident in solar array dust cleaning events (Stella et al., 2005).

Spirit experienced more severe power loss due to dust accumulation on its solar arrays as well as its higher southern latitude location and consequent decreased solar insolation during the winter compared to Opportunity, which forced more extreme operational accommodations. During its first fall and winter season, Spirit was operated in a manner that maximized the tilt of its solar arrays toward the sun by driving and parking on north-facing (sunward) slopes wherever possible (Arvidson et al., 2006). In the subsequent two late fall, early winter seasons, diminished solar array output due to dust accumulation forced the need to park Spirit on north-facing slopes for the duration of the season in order to produce sufficient power to survive (Arvidson et al., 2008). During its fourth and final winter, Spirit had become embedded in fine sand in an orientation tilted away from the sun (Arvidson et al., 2010). The consequent reduction in power output from the dusty solar arrays forced the termination of all nonessential power usage, including radio communications, which were never re-established. A similar sequence led to the end of the Opportunity mission, although it was a global dust storm and accompanying increased atmospheric opacity that led to a dramatic decrease in solar insolation and consequent loss of power.

Dust coatings on rocks were a hindrance to science investigations of surface missions, most notable in the case of the Spirit rover. Many of the basaltic rocks it encountered on the crater floor had a light-toned appearance due to dust, which led to their misidentification in images, especially in grayscale images (Greeley et al., 2004). Furthermore, even relatively dark-toned rocks had sufficient dust cover to impact Mini-TES spectral measurements. Referred to as optically thin surface dust (Hamilton & Ruff, 2012), its infrared spectral character changes with time of day as the dust heats and cools at a different rate than the rock it covers. The temporal effect was recognized early in the mission (Ruff et al., 2006) but not properly attributed to thin dust until the end of the mission. This delay was due in part to optically thin dust that had accumulated on one or more mirrors within the Mini-TES telescope, which occurred on both rovers (Smith et al., 2006). Although both were equipped with an aperture closing mechanism, it was not designed with a tight dust seal.

Both MERs were equipped with a Rock Abrasion Tool (RAT) designed to grind several millimeters into rocks across an ~4-cm-diameter circular area in order to obtain a relatively fresh surface for science investigations. Two brushes on the RAT were designed to clear cuttings after a grinding operation and also to clear dust off rocks even without a grind (Gorevan et al., 2003). The brushing capability proved to be remarkably useful for removing dust coatings on rocks, producing essentially dust-free exposures for science investigations of rock surfaces. It is noteworthy that even rocks that appeared relatively dark in camera images from both rovers were made darker following the use of the RAT brushes, attesting to the ubiquitous presence of optically thin dust on Martian rocks (Figure 5).

Open in new tab

Figure 5. Dusty rocks in Gusev crater that were brushed with the Rock Abrasion Tool. Partial circles result from uneven surfaces. The diameter of each spot is ~4 cm. Images are from the Panoramic Camera (Pancam; NASA/JPL/Arizona State University).

Source: PANCAM images.

Composition

At the time of writing, the currently operating surface missions are the Curiosity rover in Gale crater, the Insight lander in Elysium Planitia, the Perseverance rover in Jezero crater, and the Zhurong rover in Utopia Planitia. The latter two have not yet published results, and Insight is not equipped for compositional analyses. ChemCam on Curiosity has a stand-off laser-induced breakdown spectrometer (LIBS) that provides quantitative elemental composition and has been used to investigate Martian dust. Because the measurement involves pulsed nanosecond laser shots, chemistry obtained from the first few shots typically represents that of dust covering the surface of rock and soil targets, which is effectively removed in the process to expose the substrate (Meslin et al., 2013). Lasue et al., (2018) refined the preliminary ChemCam results on dust chemistry using observations from the first 1,500 sols of the mission. They determined that every first shot obtained over this period presented a strong hydrogen signal, consistent with one or more hydrated components in the dust. This result supports previous indications that Martian dust is hydrated, as described in the previous “Composition” section. The bulk chemistry also is notably similar to results from previous missions in which dust or fines were measured. Curiosity includes an observation tray on which dust accumulates for measurements by its APXS instrument. There is close agreement between its results and those of the ChemCam LIBS, demonstrating that dust has a generally basaltic composition with the addition of S and Cl that likely represent one or more salt components (Berger et al., 2016; Lasue et al., 2018).

Curiosity carries a powder X-ray diffraction instrument known as CheMin that measured the fine fraction (<150 μ‎m) of an eolian “sand shadow” dubbed Rocknest, which likely includes a substantial Martian dust component (Bish et al., 2013). A key result was the recognition of an amorphous component representing ~30 wt% of the bulk material. Berger et al., (2016) made the case that much of the amorphous component is present in dust and probably is attributable to npOx and associated S and Cl. The nature and origin of the amorphous component remain unresolved, and its presence in dust has not been definitively shown.

Landscape Features

New orbital observations related to Martian dust have resulted from the highest resolution images yet obtained using the High Resolution Imaging Science Experiment (HiRISE) camera on the Mars Reconnaissance Orbiter starting in 2006. A class of features discovered with HiRISE images known as recurring slope lineae (RSL) that originally were considered possible evidence for flowing liquid brines (McEwen et al., 2011) are now viewed as features possibly related to surface dust cover. Similar to the slope streaks described in the previous “Landscape Features” section, these relatively dark, narrow (0.5–5 m), linear markings are observed to form incrementally on steep slopes during warm seasons and fade in cold seasons. A key observation that raised doubts about a water-related origin is the lack of a thermal signature consistent with either a liquid or a solid form of water (Edwards & Piqueux, 2016). Multiple papers followed, as summarized by McEwen et al., (2021), favoring a dry origin for RSL based on some form of granular flow. They suggest that the characteristics of RSL can be explained by dry flows of dust and/or sand on steep slopes and that dust lifting triggers and enhances RSL activity. Some details of this scenario remain unresolved, but a dry flow origin for RSL negates the most significant challenges of a wet origin: explaining the source and recharge of water.

The source of Martian dust is uncertain, but recent work by Ojha et al., (2018) makes a case for the Medusae Fossae Formation (MFF) as a candidate. The MFF has long been considered to be the possible eroded remnants of a regionally extensive ignimbrite deposit, as first suggested by Scott and Tanaka (1982). Ojha et al., (2018) demonstrated a correlation between the elemental chemistry in and around the MFF with that of dust measured in situ by different surface missions. Orbital data from the Gamma Ray Spectrometer onboard the Mars Odyssey spacecraft were used to make the connection. Most significant is the high abundance of S and Cl found in the vicinity of the MFF and a S:Cl ratio comparable to that measured for dust in situ. It is the S and Cl content of dust that distinguishes it from the otherwise basaltic composition that is comparable to typical Martian soil. Based on long-recognized evidence for the friable nature and extensive erosion of the MFF, Ojha et al., (2018) suggested that it could have supplied a volume of fines sufficient to account for the known dust reservoirs on Mars.