NASA's 2015 Evidence for Flowing Water on Mars, and Why It Faded

In September 2015 NASA announced what it called the strongest evidence yet that liquid water flows on Mars today. The evidence was real: an orbiting spectrometer found water-bearing salts on dark streaks that grow down Martian slopes every warm season. The interpretation, however, did not last. By 2017 a closer study of the slopes suggested the streaks behave like dry sand, and a 2018 fix to the instrument's data processing weakened the salt signal itself. Today most researchers think little or no liquid water is involved, although the exact mechanism is still debated.

Here is what was found, why it caused so much excitement, and why the story turned out to be more complicated.

NASA’s 2015 Evidence for Flowing Water on Mars, and Later Doubts

Dark streaks running down slopes in Hale crater, shown in false colour on a 3D terrain model with the height exaggerated 1.5 times. The bluish tones upslope come from the mineral pyroxene, not from the streaks. Credit: NASA/JPL/University of Arizona.

What NASA announced in 2015

The streaks are called recurring slope lineae, or RSL. They are narrow, dark lines, usually a few metres wide and up to a few hundred metres long, that appear on steep, often sun-facing slopes. They lengthen during the warmest part of the Martian year, when ground temperatures rise above about minus 23°C, and fade when it turns cold again. The seasonal rhythm is what first made scientists think of water.

On 28 September 2015 a team led by Lujendra Ojha, then a graduate student at Georgia Tech, published a paper in Nature Geoscience. Using the CRISM spectrometer on NASA's Mars Reconnaissance Orbiter, they reported chemical signatures of hydrated salts at several RSL sites. The salts best matched a mix of magnesium perchlorate, magnesium chlorate and sodium perchlorate.

Crucially, the signal appeared only when the streaks were at their widest, and vanished at the same spots when the streaks were small. That timing suggested the salts were tied to whatever formed the streaks. NASA's science chief at the time, John Grunsfeld, said the results appear to confirm that water — albeit briny — is flowing today on the surface of Mars.

Why salty water seemed to make sense

Pure water cannot stay liquid on the Martian surface for long. The air is so thin and cold that it would freeze or boil away quickly. Salt changes that. Just as road salt melts ice on Earth, perchlorate salts can keep a brine liquid at temperatures far below zero, in some cases down to about minus 70°C.

Perchlorates were already known on Mars. NASA's Phoenix lander found them in polar soil in 2008, and the Curiosity rover detected them in Gale crater. The 2015 study was the first time they had been identified from orbit, and it placed them exactly where the mysterious streaks were growing. The proposed picture was a thin flow of brine moving just under the surface, wetting the soil enough to darken it.

Dark streaks up to a few hundred metres long running down the walls of Garni crater, on a 3D terrain model with 1.5 times vertical exaggeration. Credit: NASA/JPL/University of Arizona.

Ojha had been the first to notice these features back in 2010, as an undergraduate at the University of Arizona sifting through images from HiRISE, the high-resolution camera on the same orbiter. The 2015 paper combined years of HiRISE photos with CRISM's mineral mapping.

An animated fly-around of Hale crater, where dark streaks lengthen down the slopes in warm seasons. Each streak is roughly as long as a football field.

The 2017 study: flowing sand, not water

Two years later, a team led by Colin Dundas of the US Geological Survey took another approach. Using stereo pairs of HiRISE images, they built 3D models of 151 streaks at 10 sites and measured the exact slope where each one started and stopped.

The result was striking. Almost all the streaks occurred on slopes steeper than about 27 degrees, and each stopped at roughly the angle where loose, dry sand comes to rest on a dune. Liquid water would be expected to keep trickling onto gentler ground; these flows did not. As Dundas put it, the slopes are more like what we expect for dry sand.

The team did not rule out water completely. Hydrated salts can pull small amounts of moisture out of the atmosphere, and that might help trigger the slides. But the findings argued strongly against enough liquid water to support microbial life at these sites.

More doubts about the salt signal

Further work chipped away at the wet interpretation. Neutron measurements from the Mars Odyssey orbiter found no extra hydrogen at active RSL sites, although that instrument sees areas far larger than a single streak. In 2018 the CRISM team reported a processing artefact that could create false mineral signals in some pixels, including signals that looked like perchlorate. That did not erase the 2015 result entirely, but it made it much less certain.

Other explanations have been proposed since: dry sand flows set off by gas escaping from warming ground, the expansion and collapse of salts as they absorb and lose moisture, or seasonal frost. None of them explains every observation, which is why RSL remain an open question rather than a solved one.

Is there water on Mars at all?

Yes, and plenty of it, just not flowing across the surface. Mars has large amounts of water ice in its polar caps and buried under the ground at middle latitudes, where it has been exposed in fresh craters and cliffs. Rovers have also found clear signs that rivers, lakes and possibly seas existed billions of years ago, including the ancient lake bed of Jezero crater now explored by the Perseverance rover.

What remains unproven is liquid water at or near the surface today. That question matters for two reasons. Liquid water is the main thing scientists look for when asking whether life could survive on Mars now, and any future crew would benefit from water they could extract locally.

Why the 2015 headline still matters

It would be easy to call the announcement a mistake, but that misses how science normally works. The 2015 team reported a genuine measurement and a reasonable explanation for it. Other scientists then tested the idea with new data and found problems. The conclusion changed because the evidence did.

There is also a practical consequence. RSL sites are treated under international planetary protection rules as areas that might support Earth microbes, so spacecraft that could carry contamination are kept away. Some scientists argue that if the streaks are dry, those restrictions could be relaxed, allowing a rover to visit one and settle the question on the ground.

For readers, the lesson is simple. When a discovery is described as the strongest evidence yet, it is worth checking a few years later to see whether follow-up studies agreed. In the case of flowing water on Mars, they mostly did not.

The Mars Reconnaissance Orbiter, which made both the 2015 and 2017 studies possible, has been studying the planet since 2006. It is managed by NASA's Jet Propulsion Laboratory and was built by Lockheed Martin. More background is available on NASA's Mars Reconnaissance Orbiter page and its Mars exploration pages.

Sources: NASA press release, 28 September 2015; NASA/JPL: Dark streaks on walls of Garni crater; NASA JPL, 20 November 2017: Recurring Martian streaks, flowing sand not water?; Wikipedia: Recurring slope lineae.

Related: