Recent research has confirmed the observation that extensive areas of Mars are dominated by layered terrain. Layers exist everywhere - even within craters. This poses two main problems, depending on the interpretation of the origin of the layers.

If the layers are sedimentary, then it implies an active early sedimentary history for Mars, involving a far stronger hydrological cycle than we have evidence for. If early Mars were this active, then all the early craters should have been wiped away, yet they are largely unaffected. In addition, layering exists within many unbroken craters, without sign of an entry or exit channel. It is as if the layers had fallen from the skies.

If the layers are lavas, then it implies a much greater early volcanic history than we see evidence for. Again, many of the large early craters should have been destroyed or buried, and we should see evidence for many more volcanic vents and fissures than we do in fact see. In one way, the volcanic model can explain the layers in enclosed craters. Massive volcanic eruptions could have formed ash clouds that rained down into the craters, yet we don't see the vents.

Here are the seeds of yet another Mars paradox. The evidence is internally contradictory. We see layers, yet no sign of the processes that caused them or the likely side effects of those processes.

If we look carefully at the available images and think wider about processes, we can begin to understand what's really going on. Look for example at a recent MGS image release of a crater in Nepthenes Mensae, with layers exposed in its walls.
 

On the left is the MGS image at reduced scale. Clicking on it will access the full-scale version. Layering is clearly developed in the  walls of this ~3 km young crater. We are seeing several hundred metres thickness of layered material. If we look at the context image on the right, we see that the crater is developed in the floor of a larger (~25 km) and older crater with no obvious entrance or exit. Therefore this layering must have filled the larger crater somehow without breaching it. The epoch of layering can also be deduced to straddle the formation age of the two craters.

Note that there are several nearby craters of intermediate age, particularly a ~12 km crater to the left. It is immediately obvious that at least some of the layering must be impact ejecta from the 12 km crater. Thus we have the beginnings of a model for the layering on Mars. It is the accumulation of many sheets of impact ejecta from many craters, large and small.

Why don't we see similar layers on the Moon? There the impact layers are more jumbled. What is different about Mars?

The answer is our old friend, Carbon Dioxide.

When a large meteor hits the Moon, it blasts debris around, but there is no real difference between the debris from one impact or another. All end up in an undifferentiated vertical stack. On Mars, the target is regolith (crushed rock and dirt) plus ice. Water ice is relatively refractory. Some of it melts or vaporises in the impact, but most is blasted out as unmelted grains. The CO2, on the other hand, is very volatile. It readily forms a gas with much lower energy requirements. Therefore, every time a large impact crater was formed on early Mars, a large volume of CO2 gas was produced which temporarily formed a huge cloud - perhaps even a planet-wide atmosphere in the case of basin-scale impact. The ejecta from the impact would be deposited on a timescale of seconds to minutes (or perhaps as much as a few hours for a really big impact). The temporary "bubble" atmosphere would persist for days before it cooled down and froze out as an independent layer. Thus each impact would lay down an ejecta sheet, and then a layer of CO2-rich ices. In this way the ubiquitous layering of Mars was formed by repeated impacts, and is anticipated to be composed of alternating rock-rich ejecta sheets and ice-rich "atmospheric" layers.
 
 

The process of collapse of these temporary atmospheres would have been spectacular. Perhaps it is appropriate that the earliest epoch on Mars is termed the Noachian, after Noah and his flood. Although most of the atmospheric precipitation would have been CO2 snow and ice, there would also have been water precipitating as snow, hail, and even as rain in warm areas. Local erosion in warm areas would have been intense during these episodes, allowing for rapid erosion of the dendritic valley networks (although in other cooler zones, the CO2 ice and snow would have formed a protective icy carapace over the ground).

And a final, intriguing, possibility:- after a basin-scale impact, enough CO2 could have been vaporised by the impact to generate a temporary atmosphere for Mars of greater than 5 bars pressure. In these circumstances, liquid CO2 would have rained out of the condensing atmosphere and flowed as rivers across the Martian surface!
 

      Created: May 2002
      Last modified: May 2002
      Authorised by:  Head, Earth Sciences

      Maintained by: Nick Hoffman
      Email: nhoffman@unimelb.edu.au