<P> An analysis of data from the Mars Odyssey Neutron Spectrometer revealed that the RSL sites do not contain any more water than found at anywhere else at similar latitudes . The authors concluded that RSL are not supplied by large, near - surface briny aquifers . It is still possible with this data that water vapor from deeply buried ice, from the atmosphere, or from small deeply buried aquifers . </P> <P> Dry granular flow was proposed since the first observations of RSLs but this interpretation was ruled out because of the seasonality of the process . The first proposition of a seasonal triggering in a dry context was published in March 2017 using a Knudsen pump effect . The authors demonstrated that the RSLs stopped at an angle of 28 ° in Garni crater, in agreement with dry granular avalanche . In addition, the authors pointed out several limitations of the wet hypothesis, such the fact that the detection of water was only indirect (salt detection but not water). This theory pushed back the dry flow theory . Research published in November 2017 concludes that the observations are best explained by dry flow processes, and remark that there is no actual spectrographic evidence for water . Their research shows RSL exist only on slopes steeper than 27 degrees, enough for dry grains to descend the way they do on faces of active dunes . The RSL do not flow onto shallower than 27 degree slopes, which is inconsistent with models for water . A 2016 report also cast doubt on possible sources of underground water at RSL sites, but the new research article acknowledged that hydrated salts could draw some humidity from the atmosphere and seasonal changes in hydration of salt - containing grains might result in some trigger mechanism for RSL grainflows, such as expansion, contraction, or release of some water, that would change the cohesion of grains and cause them to fall or "flow" downslope . Furthermore, neutron spectrometer data by the Mars Odyssey orbiter obtained over one decade, was published in December 2017, and shows no evidence of water (hydrogenated regolith) at the active sites, so its authors also support the hypotheses of either short - lived atmospheric water vapour deliquesence, or dry granular flows . Nevertheless, the footprint of this instrument (~ 100 km) is much larger that the RSls (~ 100m). </P> <P> These features form on Sun - facing slopes at times of the year when the local temperatures reach above the melting point for ice . The streaks grow in spring, widen in late summer and then fade away in autumn . This is hard to model in any other way except as involving liquid water in some form, though the streaks themselves are thought to be a secondary effect and not a direct indication of dampness of the regolith . Although these features involve water in some form, the water could be either too cold or too salty for life . At present they are treated as potentially habitable, as "Uncertain Regions, to be treated as Special Regions". </P> <P> The "Special Regions" assessment says of them: </P>

Recurring slope lineae in equatorial regions of mars