Recurring slope lineae (RSL) are dark, narrow features that appear frequently on low-albedo steep slopes on Mars (McEwen et al., 2011; Hargitai and Kereszturi, 2015). Such features are around 40% darker than nearby areas, originating from warm steep slopes and incrementally lengthening (McEwen et al., 2011). These features have three main properties: They gradually lengthen during warm seasons, disappear in colder periods, and recur annually (McEwen et al., 2011; McEwen et al., 2014; Stillman et al., 2014; Hargitai and Kereszturi, 2015). Sites of RSLs are continuing to be found in high-resolution images taken by the High-Resolution Imaging Science Experiment (HiRISE) onboard the Mars Reconnaissance Orbiter (MRO) (McEwen et al., 2011). The confirmed RSL sites have been noticed to lengthen, disappear, and reoccur, whereas candidate or potential sites lack evidence of these three noted features. These RSL sites generally have low-albedo exposed rocks. Previous studies on RSL in the equatorial region are most significant with warm seasons in Martian solar longitude (Ls) = 50°–200° (end of northern spring and whole summer season) on slopes facing north and in Ls = 200°–290° (mid northern autumn and early northern winter season) on southern facing slopes with a length of 100 m in Valles Marineris (McEwen et al., 2014).

RSL on Mars are hypothesized to originate from snow melting, deliquescence, dry flow, or shallow groundwater (Abotalib and Heggy, 2019). The RSL activity on sun-facing slopes, which are generally warmer, has convinced many to support the presence of volatile probably fresh water or brines (McEwen et al., 2011; Levy, 2012; Grimm et al., 2014; Ojha et al., 2014; Stillman et al., 2016). Experimental studies found that perchlorate and other salts in RSL could potentially form liquid brine via deliquescence with favorable environmental conditions (i.e., temperature, water vapor) (Zorzano et al., 2009; Gough et al., 2011; Fischer et al., 2014; Nuding et al., 2014; Nikolakakos and Whiteway, 2015; Nuding et al., 2015; Nuding et al., 2019). A Raman microscope equipped with an environmental cell was used to understand the water uptake properties of calcium perchlorate at RSL locations (Nuding et al., 2019), where the Raman spectra indicate that water uptake by salts may occur as relative humidity increases, which has directly impacted the atmospheric water cycle, causing particle growth, and thus triggering a dry granular flow (Nuding et al., 2019).

Other studies also have been conducted to investigate the source of water for RSL as water vapor from the Martian atmosphere through the deliquescence mechanism (McEwen et al., 2011; Leung and Rafkin, 2015; Rafkin et al., 2016; Pál and Kereszturi, 2020). The water vapor column on Mars varies according to numerous factors, including latitude and time of year. Another theory suggested is porous substrate wetting and capillary flow (Grimm et al., 2014). This suggests that, if a minimal amount of solution from the subsurface to the surface is found at RSL as a result of capillary draw, the surface volume of the solution reduces considerably, whereas a large solution volume could still take place in the subsurface (Grimm et al., 2014). However, the exact processes and origin of the hypothesized water remains uncertain.

Over the last decade, the level of information accessible about our neighboring planet and its history has increased dramatically, which enables us to further understand the geologic record of Mars in detail and gain insights into the evolving climate and geologic history of the planet. Several minerals were found, mapped with specific geological units, and used to comprehend geological processes. Alteration materials, including clay minerals, sulfates, carbonate, hydrated silica, and zeolite, have also been found on Mars, retaining evidence of the history of contact of water and rock in the crust (Ehlmann et al., 2008). The early Noachian units on Mars are shown to contain clay minerals, especially Fe/Mg phyllosilicates and chlorite (Mustard et al., 2008), which are indicative of early weathering environments with liquid water. Fe/Mg clays in late Noachian–early Hesperian paleolakes were also found in fans and deltas (e.g., the Jezero, Eberswalde, and Holden craters), probably as both detrital sediments and in situ diagenesis (Grant et al., 2008; Goudge et al., 2015). Whereas hydrated mineralogy in Noachian or Hesperian areas are found to be widespread in the highlands, only a small amount is observed in the smaller northern lowlands (Ehlmann et al., 2008; Mustard et al., 2008; Munaretto et al., 2020). Recently, for many RSL sites, CRISM research has found spectral profiles that indicate an occurrence of hydrated oxychloride salts (chlorates and/or perchlorates) (Ming et al., 2014). A catalog of previous literature’s sites and newly added RSL sites, making a total of 940 RSL sites, have been shown globally on Mars (Stillman et al., 2014) as shown in Figure 1A, whereas Figure 1B shows the confirmed RSL site observed in this study.

In this paper, we perform a detailed geomorphological and compositional analysis of the Asimov and Hale craters, two well-established study areas having confirmed RSL sites, and investigate the atmospheric conditions and their role in RSL activity to understand their formation. This paper is arranged as follows. We first detail the study area in Section 2. Then we present the data and methodology in Section 3. We discuss the results in Section 4, which consists of weather cycling, geomorphological study, mineralogical mapping, and image interpretation. Section 5 is the discussion. We end with a conclusion on geomorphology and composition of these two well-established RSL hosting crater sites.

Asimov is an 84-km diameter crater (47°S, 5°E) located in the heavily cratered plains in the southern ancient highlands of Mars (Morgan et al., 2011). It is distinguished from other craters in this region because it contains both a ring depression just within the crater rim and a pit near the center. Figure 2A is a Context Camera (CTX) mosaic image showing details of the internal part of the crater, which has an off-center depression and an annulus of valley systems. CRISM data observation footprints at Asimov crater are shown with red polygons. Figure 2B shows Mars Orbiter Laser Altimeter (MOLA) Digital Elevation Model (DEM) of the study region, showing total elevation difference in the crater of 3724 m. The topographical analysis shows the crater’s degraded rim in the northwest. Figure 2C shows the slope map of the study area, showing the RSL sites to be occurring around 22°–36° in the crater. A previous study by Morgan et al. (2011) shows the Asimov crater to be Noachian-aged, having material fill of Noachian–Hesperian age and a superimposed annulus of valleys encircling the margins of the crater floor. It is also regarded as one of the oldest exposed areas of the planet (Scott and Carr, 1978). There have been two confirmed RSL sites possessing RSL lengthening, fading, and recurrence properties, observed in the central pit and the southwest of crater. One candidate RSL site possessing two properties (lengthening + recurrence) and four RSL candidate sites having one property are observed in the Asimov crater HiRISE image as highlighted in Figure 2.

Hale crater is located at 35.7°S, 36.4°W in the vicinity of the junction of Uzboi Vallis and the northern edge of Argyre Basin in the southern hemisphere with a north–south diameter of 125 km and east–west diameter of 150 km. It is basically a complex crater with several terraces and a central peak (Melosh, 1989), estimated to be around 1 Ga in age (Jones et al., 2011). It has a clearly defined rim and terrace structure in some parts north of the crater. One confirmed RSL site (lengthening + fading + recurrence) and two candidate RSL sites with two properties (lengthening + recurrence) of RSL have been observed in the central peak, whereas two further candidate RSL sites having one property are observed in crater’s HiRISE image as shown. Figure 3A shows the CTX mosaic of Hale crater with several terraces and a central peak, whereas Figure 3B shows the MOLA DEM of Hale crater, revealing a total elevation difference of 5354 m. Figure 3C shows the slope map of the Hale crater showing the RSL sites to be occurring around 14°–28° in the crater.

In this study, we used data from MRO HiRISE, CRISM, CTX, and Mars Global Surveyor (MGS) spacecraft MOLA data for comprehensive investigation of the RSL and their related atmospheric, geomorphological, and mineralogical parameters. The methodology used in this study is shown in Figure 4. With the help of its telescopic lens, HiRISE produces images with very fine detail. HiRISE data have a spatial resolution of 0.3 m/pixel (McEwen et al., 2011), which helped to investigate RSL activity in detail using HiVIEW: Image Viewer software specifically developed for handling HiRISE products. We also accessed the MOLA data set to investigate slope and elevation, using the slope tool in ArcGIS Software. MRO CTX has provided images that have been used to evaluate large-scale changes (tens of meters) to rapidly evaluate and record possible RSL activity (Malin et al., 2007). They have a spatial resolution of ∼5–6 m/pixel (Murchie, 2007). The Mars Climate Database (MCD) version 5.3 web interface has been used to study the surface temperature and water vapor column of the same Martian year and date when the HiRISE image of the Hale and Asimov craters was taken. The MCD is a set of simulations of Mars atmosphere meteorological patterns and has been validated using observational data (Francois et al., 1999; Millour et al., 2018). Martian atmosphere researchers utilize this database as a source for scientific investigation and designing future Mars missions (Howari et al., 2020). The MRO CRISM data has been gathering visible and shortwave infrared surface spectra (0.4–3.9 μm) since 2006 (Carter et al., 2013). Spectra from CRISM helps to understand the surface composition and identify the geological processes that have created iron and manganese-rich facies, evaporite, and hydrated deposits of minerals (Ehlmann et al., 2008; Mustard et al., 2008; Munaretto et al., 2020). A qualitative multiparametric understanding of the surface mineralogy is now possible with the help of CRISM browse products. FAL, CHL, IC2, FEM, and FM2 browse products are derived to study the composition in the Hale and Asimov craters. FAL is an improved infrared false color representation of the scene, whereas FEM and FM2 are used for mapping Fe minerals, particularly ferric and ferrous forms. IC2 shows information related to water or carbon dioxide frost or ice, whereas CHL shows information related to chloride deposits. The RGB band combination of these browse products and formulation defined by (Viviano et al., 2014) are as follows:

In this section, detailed geomorphological and compositional analysis is presented with consideration about the atmospheric parameters with the aim to investigate the origin of RSL present in these craters. Characterizing the details of temperatures in the study area by giving humidity values or daily cycling of temperature and humidity is difficult. Because the unique temporal behavior of RSL was not recognized in earlier research, the HiRISE images were used for purposes other than deducing the annual cycling of temperature and humidity. Literature review indicates that RSL could be dry or only transiently wet at very cold temperatures (McEwen, 2018; Rivera-Valentín et al., 2020). However, other work suggests that putative deliquescent RSL sites could be habitable (Maus et al., 2020). Other temperature or season-dependent factors may influence sand movement and contribute to RSL seasonality. It is possible that local heating of slopes plays an important role in outcrop scale wind patterns. For RSL, in particular, it may be necessary to resolve the effects of topography at the scale of outcrops and local slopes (Newman et al., 2017). Additionally, temperature changes could influence the cohesion of sand, for instance, due to variations in the amount of adsorbed water (McEwen et al., 2011; Dundas et al., 2017).

The reported data and observations in the current paper could strengthen the hypothesis that the Asimov crater’s RSL formation is because of wet surface or subsurface deliquescence processes. Other observations reported in the literature are also in line; for example, the surface temperatures corresponding to RSL activity are above the freezing points for salty solutions, which can be as low as ∼200 K (Möhlmann and Thomsen, 2011; Martínez and Renno, 2013). However, explaining the source of sufficient water for seepage is extremely difficult in the present-day Martian environment (Mellon and Phillips, 2001; Dundas et al., 2017).

A previous observation by Nuding et al. (2019) shows that, with the increase in relative humidity increases, visual growth and darkening are observed, which are caused by increases in water uptake, also are confirmed by Raman spectroscopy equipped with an environmental cell (Nuding et al., 2019), which can be considered as one of the trigger mechanisms for RSL. In the Asimov crater, the surface temperature (K) and water vapor column observed to be increased from 158 k to 261 K (inside crater) consistently increased in water-vapor five times in the crater, which indicates regolith-atmosphere interaction and high deliquescence potential in the surface and subsurface, and it also suggests an active water cycle.

The RSL activity in the Hale crater and associated atmospheric parameters show the temperature to be increased and water vapor column. Previously studied results from thermal analysis establish that local temperatures are observed to be high enough to allow either the melting of brines or deliquescence of salts during the observation period. However, the slope and aspect distributions of RSL activity predicted for these processes are not consistent with the observations in the Hale crater, which is also seen in the current study. There was no significant relative albedo difference, which suggests that RSL at Hale crater are not caused by seeping water that reaches the surface, but are best explained by other hypothesis and mainly by dry flows of granular material. Ojha et al. (2015) detect temporally and spatially initial change in surface hydration linked with RSL activity. Perchlorates (chlorate) in situ presence on the surface of Mars is also confirmed at the Phoenix landing site, where visible water ice was also found (Chevrier et al., 2007). Recent evidence of perchlorates at RSL sites strengthens the hypothesis that at least some RSL are produced by liquid water flow (Mitchell and Christensen, 2016). The variations in RSL as the formation of liquid brine (water uptake with the increase of relative humidity) and calcium perchlorate particles undergo visual transformations as the relative humidity increases in the presence of water uptake. Heinz et al. (2016) report that highly deliquescent salts are known to exist on Mars and may temporarily trap atmospheric water in extremely small quantities, perhaps sufficient to darken the surface but not sufficient for seepage down slopes (Gough et al., 2019; Nuding et al., 2019). Some workers have speculated that small quantities of water could trigger granular flows [ 44, 63].

MarsWRF experimental simulations found that perchlorate and other salts in RSL could potentially form liquid brine via deliquescence with suitable environmental conditions (temperature, water vapor) (Nuding et al., 2019). MarsWRF (Toigo et al., 2012) is a numerical Martian atmospheric model explicitly simulating dynamical processes and parameterizing unresolved processes (e.g., radiative transfer, atmospheric phase changes, etc.). During those analyses, calcium perchlorate particles demonstrate water uptake and loss (Nuding et al., 2019). To understand the water uptake properties of calcium perchlorate at RSL locations, Raman microscope equipped with an environmental cell is used (Nuding et al., 2019). In their experiment, the Raman spectra indicate that water uptake is occurring as relative humidity is increased. Additionally, the calcium perchlorate particles undergo visual transformations (growth and darkening) as the relative humidity is increased, and the presence of water uptake is confirmed by Raman spectroscopy (Nuding et al., 2019), which supports the theory of deliquescence as one of mechanism for the regolith-atmosphere interaction and RSL formation.

In the future, spatial and spectral high‐resolution analyses of RSL sites will help to ascertain whether the origin and formation of RSL is a singular process on a global scale or instead consists of a range of processes. As RSL are found at many latitudes and geologic environments on Mars, it is possible that a range of RSL formation mechanisms exists. From the observations in this study, alternative hypotheses for RSL formation that do not involve significant surface or subsurface deliquescence of liquid flow should be further explored. Likewise, without the existence of significant amounts of brine, capillary wicking of brines to the surface may account for the apparent darkening of the surface (Grimm et al., 2014). In addition, granular flow due to lubrication of grains by a thin film of water (Möhlmann and Kereszturi, 2010; McEwen et al., 2011) is considered more plausible due to the relatively low demand of resources needed to initiate it.

In this work, we analyzed HiRISE, CRISM, CTX, and MOLA data to perform detailed geo-morphological and compositional analysis of the Hale and Asimov craters along with a detailed atmospheric analysis with the aim to investigate the origin of RSL present in these craters. A total of 940 candidate and confirmed RSL sites were mapped using previous literature and newly identified sites globally on Mars. The study concludes that, in Asimov crater, local temperatures are high enough to allow either the melting of brines or the deliquescence of salts during the HiRISE observation period because water vapor column values are observed to be nearly five times those measured before RSL appearance in the image as well as the slope and aspect distributions are consistent with the observations. This marked increase in water vapor indicates the possibility of surficial or subsurface deliquescence. Furthermore, the possible local cyclical interaction between the atmosphere and regolith could be described as the Martian water cycle. The composition of Asimov crater is found to be coarse-grained Fe, particularly low-calcium pyroxene. Nanophase or crystalline ferric oxides are also present in traces. Along with hydrated mineralogy (phyllosilicates), chlorides are also observed. In the eastern Asimov crater, traces of CO₂ frost are also found, whereas water ice frost is observed in the southwest of Asimov crater where RSL sites exist. The surface is dust- and ice-free with more mafic content. In Hale crater, the slope and aspect distributions of RSL activity predicted are not consistent with observations. Based on the CRISM RGB band combination, the area was dust- and ice-free. However, the surround could have some ice. The investigated scene is observed to have a more mafic area, which is free of dust, hence dust not triggering the RSL appearance in the area.

The noted earlier finding suggests that RSL at Hale crater are not produced by seeping water, but are best explained by other hypothesis and mainly by dry flows of granular material. Hale crater is mostly composed of nanophase or crystalline ferric oxides along with chlorides and hydrated minerals. It also has water ice or frost and indicates the surface as dust- and ice-free and more mafic. Previous experimental study by researchers using the MarsWRF model also supports the theory of deliquescence as a major mechanism for the regolith-atmosphere interaction and RSL formation, which suggests that minerals absorb moisture from the environment until the minerals dissolve in the absorbed water and yield a solution, which is also supporting the current study.