Carbon Dioxide on Mars
By Nick Hoffman

As little as 5 years ago, if you had polled Mars researchers about the role of CO2 on the Red Planet you would have got a fairly short answer - that CO2 formed the bulk of the thin atmosphere, and on a seasonal basis CO2 snow and frost coated the polecaps in alternation. Sure, there were some interesting aspects to this story - that over 25% of the atmosphere froze out and sublimed again each year, and that the atmosphere was responsible for the giant dust storms, but that was about it for modern Mars.

However, if you asked the people who modelled what Mars was like in the past, they would have told you that there used to be much more CO2 in the atmosphere and it was responsible, through the greenhouse effect for making conditions warmer and wetter in the past, with the only question being "How warm?" and "How wet?". Some argued that Mars had been warm and wet enough for rivers, lakes, and perhaps even oceans of liquid water. In many ways, the story of CO2 was quite secondary to that of water, simply enabling the conditions for water to perform miracles of erosion and deposition, and forming the environment, potentially, for Life on Mars.

Today, a broad debate is opening in the scientific community where CO2 is gaining a new significance in the past and present surface activity of Mars. Some workers still believe that water played the major role, but even for these researchers, CO2 is becoming far more important. Others take an extreme view that CO2 alone can explain essentially the entire surface evolution of Mars and paint a picture of a very different planet than the one we thought we knew just a few years ago.

New ideas are emerging at a frantic pace and Mars research is struggling to keep up with the possibilities the new CO2 models provide. In this article we attempt to provide an overview of the different ideas being worked on and the groups who are rocking the boat of water on Mars.

A note on the physics of water and CO2

First, however, we will have a quick look at water and CO2 to understand their similarities, differences, and interactions. Both substances can exist as liquid, solid, and vapour we are familiar with these three "phases" of water on Earth. CO2 is more volatile than water (by which we mean that it forms a gas more easily - both at lower temperatures and at higher pressures - so to observe similar effects in carbon dioxide we need to study it under pressure and at low temperature). [phase diagrams here?] At low temperatures, both form solid ices - normal water ice which melts at Zero Centigrade and Dry ice that melts at -56.6 C, although one needs a high pressure chamber to keep the CO2 as a liquid and stop it boiling. Indeed, on Earth, carbon dioxide "boils" before it melts - a process we call sublimation (a solid passing directly to a gas). Dry ice sublimes at -78 C on earth, and in Mars thinner atmosphere at -130 C - the lower the pressure holding it in, the lower is the temperature at which it transforms. On Mars, temperatures are normally too cold for liquid water to form. We would expect the majority of the "water" on Mars to be solid ice in the polecaps and permafrost. For H2O, the permafrost zone extends to the equator, and down to depths of many kilometres. You would have to drill a deep borehole on Mars to find liquid water, unless a hot spring can be found where deep volcanic heat is warming the water and helping it rise near the surface.

In liquid form, both can dissolve other compounds, but very different ones. Water is good at dissolving ionic compounds such as sea salt, while CO2 dissolves light organic molecules that water cannot shift. In theory, therefore, if we examine areas of Mars where liquids have flowed, we should be able to tell chemically whether the dissolved salts are the normal compounds we expect from saline water, or more exotic compounds indicating that liquid CO2 has been in action. So far, however, we don't have definitive evidence, but the Viking and Pathfinder surface chemistry data seem to fall somewhere in between pure water and pure CO2 models.

CO2 and water also interact with each other in interesting ways. CO2 gas dissolves in liquid water - we are familiar with carbonated drinks where this principle is used to give tongue-tingling tastes to the liquids. We don't just feel the bubbles popping on our tongues, but also taste the weak acid formed by dissolved CO2 - scientists call this Carbonic acid and over thousands and millions of years, lightly carbonated water can dissolve some rocks like limestone. Paradoxically, it also reacts with other types of rock to form limestone, so a complex carbonate cycle exists on Earth, and should also on a formerly wet Mars. The most important sign of this is that we expect to find thick deposits of limestone rock on the floors of former lakes and oceans on Mars. For some reason we have not yet found these deposits. They can't be simply buried, because impact craters dot the surface and act as boreholes to probe the subsurface. Perhaps some surface chemistry is altering the carbonates faster than they are being exposed by erosion, or perhaps they simply are not there to be found? The 2001 Mars Odyssey orbiter is carrying the latest in a series of Infra-Red radiometers, looking for the signature of carbonate rock. Perhaps this instrument will succeed where others have failed before, or perhaps this is becoming a futile search and we should try to explain why the carbonate is missing.

At high pressures, if you inject too much CO2 into water, it can't all dissolve, and you get coexisting liquid CO2 and liquid water. The CO2 is denser and tends to settle to the bottom of the pressure chamber, but the two substances are like oil and water and form blobs and globules if you stir them up. This strange realm of behaviour is being studied on Earth as scientists look to the deep oceans for disposal of CO2 greenhouse gas. Compressed CO2 can be pumped out at great depth where it flows away downhill as a stream of shimmering droplets.

At low temperatures, a joint CO2-H2O ice is formed. This has a different crystal structure than normal ice which has hexagonal crystals (remember the shapes of snowflakes?).  CO2 clathrate has an open latticework formed by 44 water molecules which have spaces that trap up to 8 CO2 molecules in small 'cages' within the lattice. We would expect that the polecaps of Mars would contain quite a lot of clathrate, since both CO2 and water molecules are present and clathrate is the preferred form rather than separate ices. Unfortunately, the Mars Polar Lander that would have answered questions like this was lost during re-entry to Mars and scientists will have to wait several years before a new mission can be mounted to replace it.

Clathrate is important on Mars. It is the most common ice we expect to find, especially underground where the pressure of overlying rock forces CO2 molecules into the lattice. Thus, the melting of ice on Mars is likely to release both water and CO2. We will see important applications of this later.

Remember the 44 water to 8 CO2 molecular makeup of clathrate? That means that it doesn't take much CO2 to tie up a lot of water. When we try to calculate how much water and CO2 there should be on Mars (or at least how much there originally was) we get estimates of 3 or 4 parts water to each part of CO2. That means there isn't enough H2O on Mars for all the CO2, and that large amounts of CO2 would be left over after all the water is turned to clathrate. On Mars, we would expect to find this excess CO2 either as layers of dry ice in the polecaps and polar permafrost, or as liquid CO2 under high pressure, deep in the crust. Mars is cold enough for liquid CO2 to readily form everywhere except near the surface in the poles, so if we drill on Mars (looking for water?) we might be surprised to drill into a pressurised pocket of CO2. This CO2 would blow out ferociously, endangering the drilling equipment and crew just like an old-fashioned oil well blowout on Earth. The CO2 would emerge as a jet of gas at the surface - a geyser - which would fountain upwards and disperse into the atmosphere. Sand and rocks from the side of the borehole would be spat out into the air and rain down on the drilling rig, while pieces of drill pipe might be forcibly ejected from the hole and flung out like spears. Not a good scenario when you're wearing a pressure suit!

Signs of water on Mars

The CO2 story on Mars has many facets, but most of them revolve around the evidence for fluid flow on the surface. As we have seen, modern-day Mars is unsuited to liquid water unless one goes deep into the regolith. The main reason for this is temperature which comes directly from Mars' greater orbital distance from the warm Sun. Mars receives 43% of the sunlight that Earth does, and even a warm day on Mars is like being in Antarctica (except without any ice!). Astronomers calculate that the early Sun burned a bit cooler than it does now, giving out only 70-80% of its present day energy, so early Mars would have been even colder.

However, Mars has signs of giant floods on its surface. These are larger than anything ever seen on Earth - they would be 1000 times larger than the Mississippi in flood. Huge channels have been carved across the face of the planet, with giant channels extending for thousands of kilometres, almost as if Lowell's canals had been rediscovered. But these are no figment of blurry telescopes and late nights, these are very much real and any model for Mars evolution has to explain them. Called "Outburst Flood" channels, these are the single largest sign of water on Mars.

The best Earth analogue is the channelled scablands of Washington State, USA, where a giant glacial dam burst and unleashed catastrophic floods across the land. Temporarily, huge cataracts roared and house-sized boulders were tumbled along the scoured channels as the water made its way to the sea. Similar scenarios are envisaged for Mars but rather than a surface lake, there appears to have been subsurface storage of the liquids. Whatever fluid it was has burst out from Under the ground, causing huge collapse zones we call "chaos", from the litter of giant residual blocks up to several kilometres in size. Vic Baker of the University of Arizona did much of the pioneering work in this area and is still an important focus of research.

Proponents of water on Mars argue that the water was deep enough to be liquid and that once it burst out it would keep flowing for the few days needed to reach the northern lowlands. There it would spread out into a lake or perhaps even a northern ocean where it would freeze over and slowly sublime away over the millennia. Modern Mars is dry and desiccated but it need not have been in the past, or so these researchers claim.

There are, of course, a few problems with a water-based model. For one, there just isn't enough available water. The problem is not the total amount of water available on Mars, but how to get it to where it is needed. Unless the subsurface of Mars is riddled with caves large enough to sail the Titanic in, one simply cannot get enough water into the gaps in the rock to account for the volumes of the floods. It appears that each collapse zone flooded not once, but many times - as much as ten times in each location. This is very hard to explain on a planet with a dry surface and despite valiant efforts over the years, no credible answer yet exists.

Planetary Geologists are versatile thinkers and the lack of a complete mechanism for water flow doesn't stop them describing exactly what it was like. A common statement is that "If it walks like a duck and quacks like a duck, it must be a duck". By which they mean, that even if we don't have a full catalogue of the behaviour of these floods on Mars, they are enough like water floods that they must be so.

Some researchers even go so far as to describe oceans on Mars. Tim Parker (JPL) began this area of study about 15 years ago with a very carefully worded paper describing "contacts" at the edge of the northern plains. These appeared to be flat and might have been the shorelines of an ancient ocean. Considerable work has been done on these since and Parker now describes multiple parallel shorelines, using Earth analogues from Lake Bonneville. Other workers have lent support to the oceans hypothesis with studies of the geometry of the margins, but in the last couple of years the pendulum has swung away from shorelines. Most recent  in-depth studies have failed to identify coast-like morphologies.

Another problem is the lack of carbonate rock, which is beginning to be a glaring problem on Mars. This is a particular problem for the ocean hypothesis, but troubles all scenarios where water is active at the surface. Any lake or river cutting through virgin rock should have had carbonate deposits on its banks and bed, but we repeatedly fail to find these. So what is all the CO2 doing on Mars if it isn't locked up in carbonate rock? Where is it and what influence has it had on the history of Mars?

We though we understood modern-day Mars as cold, dry and arid. But last year Mike Malin and Ken Edgett stunned the Mars research community with images of young flow features on Mars. A very large number of gullies showed signs of fluid flow down channels in the centre, leading to deposits at the foot of slopes. Oddly, these "wet" gullies were all located on very cold poleward-facing slopes, often at high altitude in the southern hemisphere - altogether the coldest part of Mars and the least hospitable to liquid water. Nonetheless, Malin and Edgett showed how these features were very similar to features on Earth where water flow was involved.

With this background, let's look at the current research into CO2 on Mars. Who is doing what, and where are the new ideas taking us? We'll look first at the older Outburst Floods, and then at the young gullies.

MEGAOUTFLO

One of the least radical ideas that has emerged in the last few years comes from Vic Baker. His ideas about Outburst Floods are evolving and these days he has an integrated model that involves release of both water and CO2 from subterranean sources. He recognises the stability of clathrate on Mars, and models long-term deposits of CO2 clathrate in the regolith. At particular times and places, volcanic activity warms the ground and leads to breakdown of the clathrate to its constituent components - Water and CO2. Baker agrees that the CO2 component is violently expansive - almost explosive - and is involved in the terrain collapse process. But here the water and CO2 part company. Except for a small dissolved fraction, the bulk of the CO2 goes straight into the atmosphere while the water rushes across the land and carves the floodways. This water would be mildly carbonated, like the best mineral waters, but the CO2 in it would be essentially passive.

The atmospheric CO2, however, helps produce a mild greenhouse atmosphere that warms Mars slightly - enough to get a water cycle going with evaporation/sublimation from the terminus of the floods (a small ocean), and recycling of the water through the atmosphere to fall as snow or light rain and soak back into the ground and recharge the depleted atmospheres. Local glaciers might develop, and the whole scenario is like the drier parts of Antarctica, so although CO2 is involved, it does so to a limited degree. Baker does not discuss in detail the storage of CO2 in the regolith while melting proceeds, but by basic physics this must be as liquid CO2, so Baker is at least recognising the existence of liquid CO2 in the subsurface of Mars, even if only on a temporary and local basis.

Dry CO2 Winds

Several authors have investigated the potential of wind action on Mars to carve the outflow channels. In recent years, Peter Ravenscroft produced a long account of potential wind activity on Mars, based on his observations in the Australian outback. Conway Leovy (University of Washington), a noted Mars atmospheric scientist has produced a much clearer and more scientific description of the processes involved.

In essence, the flood channels are seen as the traces of funneled winds that have followed the same track across Mars for millennia. Over the aeons, the ground has been stripped of fine particles and larger material has been sand-blasted and eroded away. In time, everything is possible and these authors suggest that wind erosion alone is capable of carving canyons many kilometres deep. Of course, the soft and friable nature of Mars regolith helps this, but considerable time is still required.

This idea has a number of problems. For instance the regularity in depth and width of the channels would be hard to achieve with wind alone, and the sharp edges of the channels suggests an unusually focussed wind structure. Furthermore, the evidence at the Viking 1 and Pathfinder landing sites suggests that metre-scale boulders have been rolled into position by the flow. Winds on Modern Mars have difficult moving sand grains, even in the most violent storms, so clearly a very different atmosphere is called for in Mars past.

We are indeed talking about the past, here. The outburst floods peaked about 3 billion years ago, and in many cases the scoured surfaces have been exposed ever since. A steady rain of meteorites has led to a dense pockmarking with impact craters and the winds that are invoked to have eroded the channels in the first place must have stopped working, or these craters would also be removed.  So instead of the full 4.5 billion years of Mars history, only about 1 billion years is available to carve the channels before the winds died out forever. Why they should die is not entirely explained, but as with a wet Mars, the sudden disappearance of an active atmosphere is an unexplained oddity of these models.

There are, to be sure, many moderate and even severe wind erosion effects on Mars, but these are usually local and lead to chains of pits and deflation hollows, to unevenly weathered surfaces, and to stripping of plains material over wide areas. Wind has been important on Mars, but probably not in the way that Ravenscroft and Leovy suggest.

Liquid CO2

Over the years, a number of Authors have discussed literal rivers of liquid CO2 on the surface of Mars. Authors such as Sagan, Milton, Numedahl, and Vlassopoulos have discussed ephemeral flows of liquid CO2. In each instance, although the fluid is superficially attractive, due to its stability in Mars' subsurface at the prevailing temperatures, the ideas have failed due to its lack of stability at the surface. An absolute minimum of 5 times earth's atmospheric pressure is required to stabilize liquid CO2. Since modern Mars has less than 1% of Earth's pressure, the possibility is quickly dismissed. Any liquid CO2 that did appear at the surface would instantly vaporise, and probably do so explosively. No liquid would flow, and yet this idea keeps returning because water models are so hard to justify, and liquid CO2 is so likely underground on Mars.

White Mars

In contrast to so-called "Blue Mars" models of abundant liquid water, Nick Hoffman of La Trobe University (Melbourne, Australia) proposed a radical new model in 1999. Coming from the same starting point as the liquid CO2 models, he took the process one step further. Recognising the explosive degassing of liquid CO2 emerging from the subsurface, he drew analogies between a chaos zone and the throat of a degassing volcano on Earth. Although the temperatures are very different, the processes are similar. On Earth. magma breaks down explosively into volcanic gasses and fine particles - "ash". On Mars, a liquid-CO2 saturated ground would explode into dust and sand grains, and CO2 gas.

On Earth, the cloud of hot gas and ash will often roll down the side of the volcano as a fierce hot blast, faster then any normal wind, and carrying a load of pulverised rock and rubble. Boulders up to a few metres in size are transported down the side of the mountain in a terrifyingly destructive avalanche of heat, gas, and rubble. The town of Herculaneum was destroyed in 79 BC by just such a storm, and in 1929 the entire town of St. Pierre in the West Indies was obliterated. 29,000 people died in seconds, and only two persons survived the onslaught. Another recent example was the lateral blast of Mount St. Helens which developed a huge hot cloud that killed a number of bystanders.

These pyroclastic flows and surges on Earth are gas-supported, yet transport coarse debris. They flow like a fluid, but no liquids are contained in them. Perhaps this is the answer to the paradoxes of Mars? Were the Outburst flood channels carved by huge dry floods of carbon dioxide gas and rocks floating down the channels like a ghostly wave, yet travelling terrifyingly fast (over 150 metres/second  - 330 mph)?

Carbon Dioxide has a number of advantages over water as a mechanism for the floods. For one thing, because it expands to much, each block of rock that collapses in a chaos zone already contains all the volatiles it needs to transport itself all the way to the lowland plains. The reservoirs of liquid CO2 do not need to be replenished like water reservoirs do. Each chaos zone collapses and flows once, then is quiescent once its volatiles are exhausted.

The speed of collapse is awesome in such a model. An entire chaos zone measuring 80 km long by 60 km wide and 5 km deep can be excavated in a matter of hours, or a few days at most. This is the same sort of rate as seen on Earth in the largest-scale volcanic eruptions where an entire magma chamber empties within few days, leaving a giant collapsed hole called a caldera. Flows on Mars are of such a scale that they may dwarf even these giants, and one problem with this CO2 vapour flow model is that we simply have not yet developed the tools to model flows on such a scale so to some degree the theory is unproven. Nonetheless, it is an interesting one.

The White Mars model predicts that Mars has always been cold and dry, with a thin atmosphere. In fact it was probably colder and drier in the past due to the weaker Sun. Other predictions include the lack of  carbonate rock, since water probably never flowed on the surface, and a host of mineralogical predictions based on CO2 solvent action rather than water.

More data and modelling are required to "prove" the White Mars model but so far the model is proving to be at least as robust as conventional "Blue Mars" models.

Mud and Debris Oceans

Heinz-Pieter Jöns of the University of Wuerzburg in Germany has been proposing non-standard flow models for some time. He argues that many esker-like ridges in the polar regions may be frozen "waves" of mud or slurry released in major thaw episodes of CO2. This is akin to the models of Ken Tanaka (USGS) who has long held that the outburst floods were fluid-poor debris flows of some kind. He has worked with colleagues to develop a model for the fill of the northern plains not by conventional sediment at the base of a water ocean but by a single wave of flow of a "Mud Ocean". To explain the rapidity of terrain collapse required, they invoke CO2-driven collapse processes and CO2-charged mudflows with or without liquid water.

Models for modern-day Mars - the Gullies

Two basic models are being floated for the origin of the gullies. The first, originally proposed by Michael Carr of the USGS and Nick Hoffman, but elaborated and championed by Don Musselwhite (University of Arizona), brings White Mars concepts to modern-day Mars. It is proposed that near-surface reservoirs of liquid CO2 are released in the cold gullies, and that small gas-supported flows pour down the gullies and erode them. Thus, there is no liquid water involved and we are dealing with substances wholly appropriate for modern Mars.

Opponents of these models suggest that the temperature conditions in the subsurface are not exactly right for the storage and release of liquid CO2. Additionally, the anti-CO2 faction claim that CO2 would be TOO explosive to make such small and detailed features. Although the explosivity argument can probably be solved by using less CO2, more modelling is clearly needed here before either side can claim their case is soundly proven.

The alternative model for gully formation by CO2 has been put forward by Nick Hoffman again. He notes that similar features on Earth do not involve fluids emerging from deep underground, but are formed by surface melting. He shows that surface thaw of CO2 snowpack in the gullies will lead to shallow avalanches of cold CO2 snow and rocks that have been warmed in the springtime sun. As these materials tumble down the gullies, the temperature will equalise and CO2 gas will be generated, producing a gas-lubricated avalanche or small flow. This model avoids all the difficulties of shallow subsurface storage of liquid CO2 and is an interesting possibility.

Summary

The last few years have seen a new flood of data returns from Mars with the successful Pathfinder and MGS missions, despite the loss of MCO and MPL. This data flood is enabling new observations of the planet and new models to be constructed that are shaking the established concepts of a planet that was once active with water but is no longer. A variety of researchers are beginning to look at the role of CO2 in flow processes on modern and ancient Mars and in the process discovering some interesting answers to previously impossible puzzles.

A wide variety of models are suggesting the role that CO2 has in wind erosion, and in unconventional fluid flows that mimic the action of water. Even water-based models are starting to use CO2 to give them more explosive power. The most radical ideas are coming out of two centres - Arizona where the University of Arizona and its Lunar and Planetary science institute and the USGS field station in Flagstaff have collected a significant body of researchers, plus Australia where some of the most radical ideas are being formulated. In each case, these are dry regions where people are used to dealing with long droughts - but perhaps not as long as Mars has endured!

While most researchers are not yet prepared to adopt a model as extreme as White Mars, aspects of it are being incorporated into most serious studies of flow processes and surface conditions on Mars. Within the next few years, the question on Mars may no longer be "How warm and wet?" but "How cold and dry?" Indeed, the fundamental question of White Mars is "Do ANY features on the planet actually REQUIRE liquid water, or can we use CO2 for everything?"

For those who persist in saying that if it walks like a duck and has a beak like a duck, then it must be a duck, I suggest you read up about the platypus. These marsupials even lay eggs but they don't quack!
 
 

(c) Nick Hoffman. All rights reserved, etc.
 
 
 

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

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