Nasa's Mars Reconnaissance Orbiter (MRO) is putting a new spin on the story of water on the Red Planet.
Some gullies look almost certain to have been cut by liquid flow...
The unprecedented detail seen by its instruments is challenging assumptions about just how wet our near neighbour was in the past - and especially today.
The orbiter has re-visited classic images purporting to show water-cut gullies, and found the evidence to be less compelling than first thought.
MRO scientist Alfred McEwen said a more complex picture was emerging.
"These are early results, and Mars is indeed a diverse planet; it's had a complicated geologic history," the principal investigator on the spacecraft's High Resolution Imaging Science Experiment (HiRISE) camera told the BBC.
The MRO team has published a number of findings in the journal Science based on the probe's first 100 days in orbit around Mars.
It is a tiny snapshot. The orbiter has already acquired more than eight terabytes of data and researchers have barely had a chance to look at all the images, let alone begin to analyse them in any detail.
The early foray into Science magazine, however, may just act as a brake to the some of the more runaway media reporting that has given the impression that the Red Planet has at times been sloshing with water.
Boulders on the northern plains complicate the story of an ocean
A key example is images reported last year of what appeared to be evidence of current water activity on the planet, based on before-and-after shots taken since 1999 by another orbiter, Mars Global Surveyor (MGS).
MGS saw bright gully deposits that had been thrown down the sides of slopes in a fashion that looked just like a water flow.
But to the keener eyes of MRO's HiRISE camera and its Compact Reconnaissance Imaging Spectrometer, the evidence seems more equivocal, and the new deposits may not even involve water.
"The thing we've found with six examples is that they all occur on some of the steepest slopes, steep enough that dry movement, dry flow, could have been sufficient to explain these deposits," said Professor McEwen.
"I'm confident in the conclusion that a dry flow is probably the leading candidate. We can't rule out that it was water flow, especially a wet debris flow that made deposits look like this.
"We've got to keep observing these, looking for changes. If we can see an example on a sufficiently shallow slope that we can't explain with dry movement, then that would be very important."
...on some steep slopes, tumbling dry material could be responsible
MRO has taken a detailed look, too, at the northern lowlands of Mars where many have suggested a great ocean once stood. If that was the case, the landscape should be covered with fine-grained, silt-like sediments. The power of MRO's HiRISE camera, though, can resolve many large boulders.
That is at odds with the standard Earth model of how such an ocean should look.
It is possible the boulders were carried to their present locations in floods, said Professor McEwen; or as the ocean froze, it formed glaciers which then reworked the surface to produce the large rocks.
"This 'ocean' in the northern plains is a frustrating ocean because we can't find anything to definitively tell us it was really there," he added.
There are aspects of the Martian water story that are being filled out by MRO. Landscapes with branched channels and fan-like deposits typical of liquid flows are seen around several impact craters, including the 60km-wide Mojave crater.
Some have suggested these features resulted from rains that fell in a sustained warmer climate. The MRO data, though, suggests the features may instead be linked to impact events - meteoroids hitting a water-rich crust to release short-lived floods.
Mojave Crater's features may be down to impact-released water
Reviewing the latest findings, Richard Kerr, in-house correspondent at the Science journal, said MRO's findings should come as no surprise.
"It really falls in line with a trend of the last few years in which the surface of Mars is looking drier and drier," he told the BBC.
"Geochemists who've been looking at the colours of the surface of Mars and interpreting its mineralogical composition have been concluding Mars has indeed rusted - that's what gives you the red colour, the yellow-brown colour - but it didn't take much water.
"It was maybe damp rather than what we think of as wet."