Implications for the origin and evolution of Martian Recurring Slope Lineae at Hale crater from CaSSIS observations

Highlights

• We report the first CaSSIS observations of RSL at Hale crater, Mars.

• We compare morning CaSSIS images with afternoon HiRISE observations.

• We do not find any relative albedo variations between morning and afternoon RSL.

• RSL lengthening is constrained to steep slopes.

• Our results indicate that RSL are consistent with dry flows.

Recurring Slope Lineae (RSL) are narrow, dark features that typically source from rocky outcrops, incrementally lengthen down Martian steep slopes in warm seasons, fade in cold seasons and recur annually. In this study we report the first observations of RSL at Hale crater, Mars, during late southern summer by the Color and Surface Science Imaging System (CaSSIS) on board ESA’s ExoMars Trace Gas Orbiter (TGO). For the first time, we analyze images of RSL acquired during morning solar local times and compare them with High Resolution Imaging Science Experiment (HiRISE) observations taken in the afternoon. We find that RSL activity is correlated with the presence of steep slopes. Our thermal analysis establishes that local temperatures are high enough to allow either the melting of brines or deliquescence of salts during the observation period, but the slope and aspect distributions of RSL activity predicted by these processes are not consistent with our observations. We do not find any significant relative albedo difference between morning and afternoon RSL. Differences above 11% would have been detected by our methodology, if present. This instead suggests that RSL at Hale crater are not caused by seeping water that reaches the surface, but are best explained as dry flows of granular material.

Introduction

The origin of Recurring Slope Lineae (RSL) is one of the most controversial science questions regarding present-day surface activity on Mars. They appear as narrow ($<5$ m) low-albedo streaks, often starting from bedrock outcrops and lengthening hundreds of meters down Martian steep slopes at equatorial and mid-latitudes (McEwen et al., 2011). RSL generally start lengthening during warm seasons, fade during cold ones and recur over multiple Martian years. However, RSL activity can be more complicated than a simple repeating cycle of lengthening and fading, as several well studied sites shows simultaneous lengthening, appearance and fading of RSL (Stillman and Grimm, 2018; Vincendon et al., 2019). Many models have been proposed to explain their origin, but a definitive explanation is still elusive. These can be broadly summarized in three classes: wet models, in which RSL are water-dominated features, dry models, that explain RSL as dry mass fluxes or aeolian features, and hybrid models, in which water plays an indirect role in RSL formation and lengthening.

The temperature dependence of RSL activity suggests that water or brines may be present (McEwen et al., 2011, 2014; Stillman et al., 2014; Huber et al., 2020). Spatial correlation with multi-scale fractures appears to support groundwater sources (Stillman et al., 2016; Abotalib and Heggy, 2019). However, topographic relationships cast doubts on these models, as they explain the seasonality of RSL activity, but require larger volumes of liquids than could be reasonably present on Mars at these latitudes (Grimm et al., 2014; Chojnacki et al., 2016). Moreover, many RSL source from local topographic highs, such as ridge crests or peaks, where a groundwater source is unlikely (Chojnacki et al., 2016). Another wet process that has been considered to explain RSL formation is the deliquescence of salts, which occurs when hygroscopic salts absorb water vapor to form a liquid solution (Gough et al., 2016). For this to happen, temperatures higher than the eutectic temperature of the solution and relative humidities (RH) greater than their deliquescence relative humidity (DRH) are required (Gough et al., 2016). Several laboratory studies have identified chlorides (Gough et al., 2016; Wang et al., 2019), perchlorates (Nuding et al., 2014) and chlorates (Toner and Catling, 2018) as potential candidate deliquescent salts. The required RH values for the deliquescence of CaCl₂ brines range from $12.9\text{\%}$ when $T=273$ K to $~20.9\text{\%}$ at $T=233$ K (Gough et al., 2016), while their eutectic temperature is $T=223K$. Numerical modeling by Gough et al., [2016] showed that atmospheric water vapor could sustain the deliquescence of hydrated CaCl₂ brines at 3-cm depth at the Phoenix landing site. However, it is not clear whether or not the same process can happen at RSL sites (Gough et al., 2016), even though Wang et al., [2019] recently proposed that the deliquescence of subsurface hydrated chlorides can be a thermodynamically viable (but not necessarily sufficient) process for triggering RSL activity. The deliquescence of Ca-perchlorates is also interesting because of their relatively low DRH (from $10\pm 4\text{\%}$ at 273 ​K to $55\pm 4\text{\%}$ at 223 ​K (Nuding et al., 2014)) and eutectic temperature of $T=200$ K. Finally, chlorates may be even more interesting candidates as they have a lower DRH than perchlorates (Toner and Catling, 2018).

In these models, RSL lengthening and formation are interpreted with dry mechanisms involving granular flows or aeolian processes (Dundas et al., 2017; Schmidt et al., 2017; Vincendon et al., 2019; Schaefer et al., 2019). These have been initially invoked to explain the correlation between the angle of repose of granular material and the slope angle at which RSL stop (Dundas et al., 2017), although more recent measurements show that $~25\text{\%}$ of RSL reach slopes below the angle of repose (Tebolt et al., 2020; Stillman et al., 2020). In addition, Schmidt et al., [2017] discussed morphological inconsistencies between RSL and wet flows, such as the lack of terminal or lateral levees; Vincendon et al., [2019] proved that RSL formation occurs outside the time-frame compatible with the existence of liquid water and does not show a preference for the warmest slopes. Finally, the spectral detection of hydrated salts at some RSL sites (Ojha et al., 2015), which would support the presence of liquids, was recently proven to not be robust (Leask et al., 2018; Vincendon et al., 2019). Although dry flows do not require a source of water, a trigger mechanism has not been fully established yet. Schmidt et al., [2017] proposed a Knudsen pump as a trigger mechanism, but it doesn’t predict RSL at the times and locations where they are observed (Stillman and Grimm, 2018; Vincendon et al., 2019). Vincendon et al., [2019] and Schaefer et al., [2019] proposed that RSL are aeolian features resulting from the removal of bright dust by winds. A fading mechanism has been recently proposed by Vincendon et al., [2019], which showed that the progressive brightening of RSL can be attributed to dust deposition. On the other hand Schaefer et al., [2019] discovered that the fading of RSL is similar to that of other dust-related albedo features (dust devil tracks, rockfalls) on the basis of relative albedo analyses at Tivat crater (${45}^{\circ }$ S, $9^{\circ }$ E). They propose that RSL fade due to the widespread removal of dust from the neighboring slopes, which progressively darken until matching the RSL reflectance. While this mechanism applies well to Tivat crater, it cannot explain the concurrent fading and lengthening of RSL observed in several other RSL sites (Vincendon et al., 2019; Stillman et al., 2020).

Other hypotheses envision an indirect role of water in triggering RSL activity. Massé et al., [2016] showed that boiling fresh water can trigger granular flows by violently displacing grains. While this model is interesting because it requires less water than wet-based models, boiling water requires high temperatures that are hardly met at the onset of RSL activity in early spring (Stillman and Grimm, 2018; Vincendon et al., 2019). More recently, Shoji et al., [2019] proposed that moisture serves to stabilize steep slopes in cold seasons, which flow when drying during the warm seasons. In addition, Bishop et al., [2019] proposed that subsurface brine activity can create surface collapse features, perhaps initiating RSL when on steep enough slopes. The High Resolution Imaging Science Experiment (HiRISE, McEwen et al., [2007]) on board the NASA’s Mars Reconnaissance Orbiter (MRO) mission is the optimal instrument to image RSL due to its high pixel scale ( $~0.3$ m/pixel), which allows for digital terrain models (DTM) at $~$ 1 ​m spatial resolution. However, the Sun-synchronous orbit of MRO limits observation times between $~$ 2–4 PM, during the local afternoon.The Color and Surface Imaging System (CaSSIS, Thomas et al., 2017), on board the ESA ExoMars Trace Gas Orbiter (TGO) mission, images the Martian surface at 4.6 ​m/pixel in four color bands. It carries a rotation mechanism that allows it to acquire stereo observations with identical illumination angles for the production of DTMs. The ${74}^{\circ }$ inclined orbit of TGO allows CaSSIS to observe the surface at different times of day, such as the local morning. This capability is pivotal for RSL studies, since morning observations of these features are not possible otherwise. The only exceptions are extremely early morning views by Mars Odyssey, with poor SNR and much lower spatial resolution. Morning activity can provide great insights on the nature and formation mechanism of RSL. Laboratory studies (Gough et al., 2016) and analyses of data from both the Rover Environmental Monitoring Station (REMS) on board Curiosity and the Thermal and Electrical Conductivity Probe (TECP) on the Phoenix lander shows that daily maximum RH values occur during the early morning of local summer (Steele et al., 2017; Fischer et al., 2019), suggesting that these local hours are the most favourable for deliquescence. If RSL are liquid-based flows, their activity should increase in the morning, when both temperature and air humidity are high enough to favor the deliquescence of salts and increase the stability of brines. Under this scenario, dehydration would occur in the afternoon, so we would also expect RSL to be darker in the morning than in the afternoon. Instead, if RSL are caused by the melting of shallow subsurface ice (such as from a deep groundwater source), then they should be darker in the afternoon. This assumes that the quantity of water or brines at play is relatively small, so that evaporative losses exceed the supply of water (Hillel, 2004). If instead RSL have a high water content, pore water concentration would not be affected by evaporation and there wouldn’t be any albedo change (Pommerol et al., 2013; Levy et al., 2014). The latter case, however, may not be in agreement with the very low water content of RSL estimated at Garni crater (Edwards and Piqueux, 2016). Finally, if RSL are dry flows, then we do not expect any albedo change between the observations. In this study we present the first observations of RSL performed by CaSSIS during the local morning and compare them with HiRISE observations acquired one month earlier in the afternoon. In particular, we search for any differences in their overall morphology (i.e. length, slope, aspect) and report relative albedo measurements performed on morning and afternoon images. We analyze the thermal conditions of the surface and shallow sub-surface at the time of the CaSSIS observation to assess whether temperatures would either allow melting of brines or deliquescence of salts.