Hundreds of thousands of dark streaks are crawling down Martian hillsides, vanishing, and reappearing with the seasons. Two studies published in Nature Communications have now used artificial intelligence to build the largest catalogs of these features ever assembled, drawing on nearly two decades of orbital imagery from NASA’s Mars Reconnaissance Orbiter. Their shared conclusion upends a long-held suspicion that liquid water might be involved: the streaks appear to be dry, wind-driven dust avalanches, and they are far more widespread than anyone previously documented.
The findings, which drew attention from both NASA and the European Space Agency earlier this year, carry practical stakes. If wind-lofted dust is actively reshaping slopes across broad swaths of Mars, engineers planning future landers and crewed missions will need to factor in surface instability and reduced visibility at candidate landing sites.
Two massive catalogs, one core answer
The first study applied deep learning to Context Camera (CTX) frames collected by MRO between 2006 and 2024, flagging more than 2.1 million individual dark slope streaks. Only a small fraction could be linked to meteorite impacts or marsquakes. The dominant trigger, repeated across multiple Martian years, was seasonal dust carried by wind.
The second team took a different methodological path, constructing a global catalog of roughly 500,000 slope streaks. Its geostatistical analysis directly challenged the hypothesis that subsurface water seeps to the surface to darken slopes, pointing instead to dry mechanisms: sand avalanches kicked off by wind gusts and, occasionally, meteoritic impacts.
The gap between 2.1 million and 500,000 detections reflects differences in detection thresholds, image coverage windows, and classification criteria rather than a disagreement. Both teams converge on the same central point: these streaks are shaped by dust and wind, not by hidden water.
NASA highlighted how AI accelerated what had been a grueling manual process. Researchers who once spent months hand-tagging streaks in orbital frames could now scan tens of thousands of images in a fraction of that time, according to NASA’s overview of the research. The raw MRO camera archives sit in NASA’s Planetary Data System, publicly accessible for anyone who wants to verify the work.
A separate darkening in the northern plains
While those catalogs focused on slope streaks scattered across Mars, the European Space Agency documented a distinct but visually related phenomenon in Utopia Planitia, the vast northern basin where NASA’s Viking 2 lander touched down in 1976. Using its Mars Express orbiter’s High Resolution Stereo Camera (HRSC), ESA identified a blanket of dark material that has been spreading across the region since the Viking era.
Comparing current HRSC frames with Viking orbiter photographs taken roughly 50 years ago showed a measurable increase in the dark material’s extent. ESA offered two leading explanations: either volcanic ash is being redistributed by prevailing winds, or bright surface dust is being scoured away to expose darker basaltic rock underneath. Both scenarios point to an active, wind-dominated process on a planet often assumed to be geologically quiet.
The connection between these two phenomena remains an open question. Slope streaks favor inclined terrain, while the Utopia Planitia darkening covers a broad, relatively flat basin. Whether the same wind and dust dynamics drive both, or whether they represent separate geological processes that happen to look alike from orbit, has not been settled in the published literature as of April 2026.
Where the uncertainties lie
Neither Nature Communications paper has published detailed error-rate metrics for its AI detection model, which makes it difficult for outside researchers to independently evaluate the precision and recall of each catalog. Until those numbers are available, the exact global count of active slope streaks should be treated as an informed approximation, not a final tally.
ESA’s Utopia Planitia analysis introduces its own complications. The Viking cameras had far lower resolution and different spectral sensitivity than modern instruments, so any precise measurement of how much the dark area has grown over five decades carries inherent uncertainty. The before-and-after comparison rests largely on ESA’s institutional interpretation; the raw Viking metadata for the specific frames used has not been independently reprocessed by a third party.
One hypothesis that neither study has tested directly involves Mars’s extensive shallow water-ice deposits. Seasonal temperature swings could, in theory, destabilize slopes in ways that pure dry-avalanche models do not capture. Both research teams explicitly favor dry mechanisms based on their statistical evidence, but neither has ruled out a secondary ice-related contribution with direct observational data.
There is also the question of timescale. The MRO-based analyses span fewer than 20 years, a thin slice of Martian history. If dust storm intensity or prevailing wind patterns shift on longer cycles, the current rate of streak formation and fading may not reflect what happened in earlier epochs, limiting how far scientists can project these findings into the geological past.
What this changes about Mars
For years, dark streaks on Martian slopes fueled speculation that liquid water might still flow, at least intermittently, on the planet’s surface. That narrative drove public excitement and influenced where mission planners considered sending future rovers. The convergence of two large-scale, peer-reviewed studies on a wind-and-dust explanation marks a meaningful pivot: Mars is actively reshaping itself, but through atmospheric processes, not hidden aquifers.
That distinction matters beyond academic debate. Landing-site selection for NASA’s planned crewed missions and sample-return campaigns depends on understanding how stable the ground actually is. A slope that darkens and sheds dust seasonally poses different engineering challenges than one moistened by briny seeps. The AI-generated catalogs give mission planners their first planet-wide inventory of where those changes are happening and how often.
As MRO continues to orbit and Mars Express monitors the northern plains, both catalogs will grow. Researchers expect to tighten detection accuracy, resolve whether the Utopia Planitia darkening shares a mechanism with slope streaks elsewhere, and test whether shallow ice plays any supporting role. For now, the picture that emerges is of a planet still very much in motion, its surface rewritten not by water but by the thin, restless wind that never quite stops blowing.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
