Dark streaks running down Martian slopes have stymied scientists for years, with researchers divided as to whether they are cause by flowing water.
A groundbreaking new study analyzing the structure of these features, aptly called slope streaks and recurring slope lineae (RSL), with the help of artificial intelligence sifted through over half a million of those and has made the call that it’s not wet, instead it’s likely bone dry and formed by dust and wind.
The study, performed by researchers from Brown University and the University of Bern, trained a machine learning algorithm on tens of thousands of high-resolution images taken by NASA’s Mars Reconnaissance Orbiter.
This enabled them to compile the first worldwide map of more than 500,000 separate slope streaks. Comparing this map with different environmental parameters such as temperature, wind speed, dust deposition and surface hydration, the AI found that the streaks were well correlated with areas of high wind activity and dust coverage, but not so much when it came to water or frost.
These features, some of them 300 or more meters long, materialize on steep, dusty slopes for no obvious reason. Benard: RSL are not stable as are the streaks, which are persistent over a substantial part of a year or even all year, but are transient phenomena, appearing at some time of the year in the warm seasons, and disappearing afterwards.
AI analysis has suggested that these flow features are probably caused by small dust particles being knocked loose by high-speed winds, little meteorite impacts, or even slight marsquakes, and then flowing fluidly downhill, moving like liquid across the landscape thanks in part to the super-fine nature of the dust and the low gravity of Mars.
This find has major implications for our picture of the Martian surface and its potential to support life as we know it; life as we know it needs liquid water.
The results indicate that the present Martian environment is even drier than previously considered, limiting the possibility of near-surface habitable niches at these streak locations.
Moreover, this study has some significance for Mars exploration. Altogether, these desiccation cracks were once labeled as “special regions” since, even if they contain presumptive water, few if any planetary protection protocols are in place to protect from off-world organisms.
The recent evidence supporting a dry origin of the Moon may help to put those fears to rest, as it could lead to more leniency in the types of missions we are able to pursue.
The fact that AI could be successfully applied to a dataset of this magnitude shows how powerful machine learning is becoming in planetary science, able to find subtle patterns and correlations that would be difficult, if not impossible, to detect using traditional methods, researchers said.
The finding represents a major breakthrough in solving a Martian mystery, and is an important contribution at a time when humanity has set its sights on new horizons in space exploration and discovery thanks in no small part to AI.