Planetary science

Picturing a recently active Mars

Discoveries made with the High Resolution Stereo Camera on the Mars Express orbiter show that, as recently as a few million years ago, the surface of Mars was being shaped by flowing water, lava and ice.

Spectacular ground-based images and chemical analyses of ancient sedimentary rock formations1 leave no doubt that Mars had a watery ancient past. These discoveries by NASA's Mars Exploration Rover Mission apply to the Noachian epoch of Mars history — the first several hundred million years of the Solar System, when the impact rate of meteors and comets was much higher than in the past 3–4 billion years2. Image interpretation has long been key to revolutionary changes in the scientific understanding of Mars. Three papers in this issue3,4,5 add to that tradition, but also add to the new evidence that Mars has been active in the geologically recent past.

The first vidicon camera images and atmospheric data from the Mariner fly-bys of the 1960s led to the view that the planet had been continuously cold and dry6. Thinking changed radically during the 1970s, when pictures of higher resolution and global coverage, taken from the Mariner 9 and Viking orbiters, revealed river-valley networks and huge channels carved by cataclysmic floods7. Nevertheless, in the theoretical syntheses that followed during the 1980s and 1990s, any active hydrological cycling was generally limited to the Noachian epoch. This MIDDEN hypothesis (‘Mars is continuously dead and dry, except in the Noachian’) prevailed for two post-Viking decades, during which there were no further pictures at increased resolution. Advocates of the MIDDEN hypothesis have variously argued that Mars is poorly endowed with water, that it is inefficient at releasing water to its surface, or that most of its water was lost to space.

In the 1980s, NASA planned a Mars geoscience and climatology orbiter, but many scientists wanted its payload to be limited to non-imaging experiments. Some even proclaimed that imaging science was an oxymoron, and that the only use of pictures was for public relations. Thus, a high-resolution stereo camera system was rejected for what eventually became the Mars Observer spacecraft. The miniature, non-stereo, high-resolution Mars Observer Camera (MOC) was eventually added to its payload, but Mars Observer was lost in 1993, and MOC did not arrive at Mars until the Mars Global Surveyor mission of the late 1990s.

Despite being limited to non-stereo coverage of very small areas, the phenomenally high resolution of the MOC images contributed to notable discoveries. These discoveries included very recent, globally dispersed hillslopes, with gullies created by water8, as well as associated lava flows and large channels carved by cataclysmic flooding9. All of these were less than about ten million years old10, as indicated by the densities of impact craters measured from the high-resolution images. These features and many others11 were clearly anomalous in regard to the prevailing MIDDEN hypothesis.

The High Resolution Stereo Camera (HRSC) carried by the European Space Agency's Mars Express Mission is the direct descendant of the camera that was rejected by NASA for the original Mars Observer. Although slightly lower than MOC in resolution, HRSC produces colour and stereo images in the broad areal context necessary for the fully effective application of geological analyses (Fig. 1). Initial results from the camera12 identified two major anomalies in regard to the MIDDEN hypothesis. First, crater counts for calderas on five major volcanoes show that there has been episodically active volcanism during the past 20% of martian history, up to only two million years ago; second, glacial activity occurred in the equatorial regions of Mars, at the base of the Olympus Mons escarpment, as recently as four million years ago.

Figure 1: Martian geology in broader context — a perspective view from the Mars Express High Resolution Stereo Camera of a 3.5–4-km-high mountain massif in the planet's mid-latitude regions.

Debris from the massif is organized into crater-filling flow patterns of parallel ridges that spread out to a width of 16 km in the foreground. The overall landform assemblage is remarkably similar to that of terrestrial, debris-covered glaciers, known as piedmont glaciers. Fretted pits and depressions show that much of the glacial ice has wasted away since the period of maximum activity, which probably occurred very late in the geological history of Mars, several tens of millions of years ago. The image has a 30× vertical exaggeration, and its brownish colour is an approximation to Mars' dusty surface. (Reproduced from ref. 4.)

The three papers in this issue use HRSC data to provide an overwhelming case for new thinking about recent geological activity on Mars. On page 356, Hauber et al.3 document that explosive eruptions, about 350 million years ago, created depressions on the flank of the volcano Hecates Tholus. As recently as five million years ago, glacial deposits were laid down on the floors of these depressions.

Head et al.4 (page 346) use HRSC data to reanalyse controversial landforms that had been recognized as possibly glacial from Viking data. Combined with data from the Mars Global Surveyor, the new images (Fig. 1) show that the landforms are indeed evidence for geologically recent and recurring glacial activity at what are now tropical and mid-latitude regions on Mars. This glaciation may be a response to recent changes in the incidence of sunlight induced by variations in obliquity of the planet's spin axis13. Glaciers form by the net accumulation of snow and its metamorphosis to ice during an Earth-like hydrological cycle.

Early evidence from Viking data for relatively young Mars glaciation14, plus other indications of young flowing15 and ‘ponded’ water activity16, led to a hypothesis in 1991, considered outrageous at the time17, that Mars had been episodically active in a hydrological sense for geologically short intervals of post-Noachian time. In the 1991 hypothesis18, martian climate change is triggered by the outburst of volcanic gases and cataclysmic flood water, with the water ponding in huge seas that subsequently evaporated.

Murray et al.5 (page 352) present striking evidence for just such an event, resulting in a frozen water body approximately the size of Earth's North Sea. The inundation was generated within the past five million years because of cataclysmic flooding that accompanied volcanic eruptions from the fracture system associated with two long and narrow depressions known as Cerberus Fossae. Flood water was probably conveyed to the sea via the Athabasca Valles channel system at peak flows of about ten million cubic metres per second9. This may be a smaller version of what happened when much larger channels delivered peak flows of hundreds of millions of cubic metres per second to indundations on the northern plains of Mars during much earlier periods of post-Noachian martian history18.

What happened to all the water that induced the past, short-term hydrological cycling on Mars? The evidence from HRSC for recent aqueous activity suggests that the water is still present, as ice on the ground and water deep beneath the surface. In early May, Mars Express will deploy its Mars Advanced Radar for Subsurface and Ionosphere Sounding (MARSIS), which should be able to detect subsurface water and ice to depths of several kilometres. Evidence from the latest pictures indicates that the water will surely be there.


  1. 1

    Squyres, S. W. et al. Science 306, 1709–1714 (2004).

    ADS  CAS  Article  Google Scholar 

  2. 2

    Neukum, G., Ivanov, B. A. & Hartmann, W. K. Space Sci. Rev. 96, 55–86 (2001).

    ADS  Article  Google Scholar 

  3. 3

    Hauber, E. et al. Nature 434, 356–361 (2005).

    ADS  CAS  Article  Google Scholar 

  4. 4

    Head, J. W. et al. Nature 434, 346–351 (2005).

    ADS  CAS  Article  Google Scholar 

  5. 5

    Murray, J. B. et al. Nature 434, 352–356 (2005).

    ADS  CAS  Article  Google Scholar 

  6. 6

    Leighton, R. B. & Murray, B. C. Science 153, 136–144 (1966).

    ADS  CAS  Article  Google Scholar 

  7. 7

    Baker, V. R. The Channels of Mars (Univ. Texas Press, Austin, 1982).

    Google Scholar 

  8. 8

    Malin, M. C. & Edgett, K. S. Science 288, 2330–2335 (2000).

    ADS  CAS  Article  Google Scholar 

  9. 9

    Burr, D. M., Grier, J. A., McEwen, A. S. & Keszthelyi, L. P. Icarus 159, 53–71 (2002).

    ADS  Article  Google Scholar 

  10. 10

    Berman, D. C. & Hartmann, W. K. Icarus 159, 1–17 (2002).

    ADS  Article  Google Scholar 

  11. 11

    Baker, V. R. Nature 412, 228–236 (2001).

    ADS  CAS  Article  Google Scholar 

  12. 12

    Neukum, G. et al. Nature 432, 971–979 (2004).

    ADS  CAS  Article  Google Scholar 

  13. 13

    Head, J. W. et al. Nature 426, 797–802 (2003).

    ADS  CAS  Article  Google Scholar 

  14. 14

    Kargel, J. S. & Strom, R. S. Geology 20, 3–7 (1992).

    ADS  Article  Google Scholar 

  15. 15

    Gulick, V. C. & Baker, V. R. Nature 341, 514–516 (1989).

    ADS  Article  Google Scholar 

  16. 16

    Parker, T. J., Saunders, R. S. & Schneeberger, D. M. Icarus 82, 111–145 (1989).

    ADS  Article  Google Scholar 

  17. 17

    Kerr, R. A. Science 259, 910–911 (1993).

    ADS  Article  Google Scholar 

  18. 18

    Baker, V. R. et al. Nature 352, 589–594 (1991).

    ADS  Article  Google Scholar 

Download references

Author information



Rights and permissions

Reprints and Permissions

About this article

Cite this article

Baker, V. Picturing a recently active Mars. Nature 434, 280–283 (2005).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.