The reason we know so little about the polar regions is not lack of interest. The polar regions are of great interest, for instance the Martian dry ice geysers in Richardson crater, one of the most interesting dynamic processes on Mars and the polar regions also have astrobiological interest too. There are potential habitats there that might even have fresh liquid water within 7 cms of the surface, beneath clear ice - of all the unexpected things to find on Mars with its near vacuum atmosphere!

As far as I know the only these are the only proposals for naturally occuring on or near the Martian surface. It is a process that happens beneath clear ice in Antarctica, with subsurface layers of ice melting due to a solid state greenhouse effect. Möhlmann's models show it should happen in the Martian ice sheets too, so long as there is similarly clear ice there.  It would be isolated from the near vacuum atmospheric conditions, and insulated from the cold of the surface due to the excellent sealing and heat trapping properties of clear ice. The models predict a layer of fresh water only a few cms thick, 10 to 20 cms below the surface in transparent ice. Thin though that layer may be by Earth standards, it is of extraordinary interest on Mars as a potential habitat for micro-organisms, in an environment where any fresh water on the surface would evaporate almost immediately.

The main potential habitats, which I’ll look at in detail in this answer, are:

  • Flow like features in Richardson Crater that form after the Martian dry ice geysers have erupted (not the same as the ones in the northern hemisphere or the ones in Russell’s crater - there are three different similar looking features that form in different conditions - only the ones in Richardson Crater are of special interest for astrobiology)
  • Liquid water forming around sun warmed grains in snow or ice
  • Perchlorate salts lying on layers of ice forms liquid water droplets in tens of minutes
  • Liquid water can exist permanently below 600 meters of ice (100 meters of rock) kept warm by the heat of Mars itself, if it once forms, e.g. after an impact
  • Ice fumaroles can mask the heat signature of venting of hot moist gas and make good habitats
  • Anywhere there is clear ice in polar regions, then fresh liquid water can form at a depth of around 6.5 cms by the solid state greenhouse effect. For a few days in summer it should reach up to a 10-20 cms in thickness.

So it’s exciting for astrobiology, also for geology too, but they are also habitats the Earth microbes could contaminate and by the Outer Space Treaty we have an obligation to prevent “harmful contamination” in the words of the treaty. It also just makes sense. If you are searching for native life on Mars, and most people agree that is one of our top science objectives there, the last thing you want to do is to just find life you brought there yourself.

Before we developed this modern understanding of the potential vulnerability of the polar regions to Earth microbes, NASA made two attempts, the Mars Polar Lander which crashed, and Phoenix which succeeded. However it was as a result of unexpected observations by Phoenix that scientists were lead to the realization that actually there could be habitats there for modern native Mars life - and so since then any landers sent there have to be sterilized to a high standard.

We could not send Curiosity there, or a second copy of Phoenix either, because it is now not thought to be sterilized sufficiently. Hopefully Phoenix has not contaminated the region of Mars around it with Earth life, but I think the Phoenix landing site might be a great site to visit to get ground truth on how effective our planetary protection measures have been on Mars - but with an appropriately sterilized lander of course.

How to read this off-line

This article runs to about 100 printed pages (not including the images - word count over 45,000). If you want to read it offline:

  • You can use Pocket - an application that helps with off line reading of web pages.
  • You can also save this page as a pdf. On windows you can use print (Ctrl + P) then change the printer to the pdf printer. The Print button then changes into a Save button and you can then save it where you want.

    Note: the embedded YouTube videos won't work if saved as a pdf, but for this page I've added "(Click to watch on YouTube)" links below each one and those will work in the pdf.
  • You can also save it as a web page. Most web browsers have an option to save a page as "Web page complete" which includes all the images in the page saved on your computer for offline browsing. Be sure to give it time to download the page before you close the browser tab.
  • I plan to make it available on kindle soon.

You can use the same methods for any of my longer articles here.



The current “gold standard” for Mars is set by the Viking landers.

Viking Lander being prepared for dry heat sterilization – this remains the "Gold standard" of present-day planetary protection.

After preliminary cleaning similarly to the levels used for Curiosity, they were then heat-treated for 30 hours at 125 °C

Five hours at 125 °C would be enough to reduce the population of microbes by ten, so this was enough for a millionfold reduction - that’s including enclosed parts of the spacecraft. It would still have a maximum of 30 spores and so several thousand dormant microbes as the spore count used undercounts the number present by a factor of a hundred or so. But in addition the numbers are reduced by the journey out there, the harsh conditions on Mars, and then a microbe would have to be pre-adapted to the conditions there to have a chance of surviving once there.

They didn’t achieve certainty but to a high chance no microbe from Viking was able to replicate and spread on Mars.

According to modern planetary protection rules then you could send a spacecraft sterilized like this to the Phoenix landing site.

But the problem is that modern equipment is much more miniaturized than for Viking, and made up of thin layers only a few atoms thick and delicate materials including epoxy attachments. Even when space hardened, it tends to be more sensitive and so would not stand being baked in an oven for days like Viking. The components would come unglued and instruments also would go out of alignment.


It’s not all bad news however, for heat sterilization. Since Viking, while commercial equipment for most purposes have got more sensitive to high temperatures, we have also had many advances in high temperature technology too. The commercial equipment is not built to withstand high temperatures not because it can’t be, but because it doesn’t need to be.

High temperature electronics and instruments are used where they are needed and are more capable than in the 1970s. We have them for oil wells as they drill deeper to regions where the temperatures go above 200 C. For planes where they can reduce weight by putting sensors closer to the engines, and for electric cars for similar reasons.

NASA has also been working for some time to develop a rover able to withstand Venus surface conditions and drive around and study the surface. With high temperatures, high pressures and sulfuric acid too. Very sterilizing for Earth life.

In 2007 they developed a silicon chip capable of 17,000 hours of continuous operation at 500 °C.

For their Venus rover, we need cameras to operate at high temperatures, we need mechanisms, we need instruments such as a Raman spectroscopy, we need communications and so on. In their 2010 study they thought all of those were possible for the future. Though they couldn’t build it yet, they saw a way to it as a future roadmap.

If the aim is to reach a high temperature for sterilization, the job may be easier to some extent, as the instruments don’t have to actually function at those high temperatures. They have to withstand being heated to high temperatures for a considerable period of time - but will then operate at normal temperatures.

So, if you choose the right components for your lander / rover, we actually have the capability to go beyond what they could in the 1970s and I do think that if we went all out with a major program, as for the Venus rover, and perhaps as a partnership with the Venus project - that we could design a 100% sterile lander in the near future. It would probably need to use RTGs for the power source - and perhaps also as the heat source for sterilization during the journey to Mars, as these have no problem working at high temperatures. Heat your lander at 500 C for six months on the voyage out to Mars and there would be no life left on it at all. Nothing viable. You can also use techniques like CO2 snow which could be done on the surface of Mars to remove even the dead organics from the outside of the lander. This would work if the lander has already been cleaned well already by other methods, is less effective if it has a fair amount of organics on it already.

There is one plan already for a sterile probe to descend into the Europan ocean by Brian Wilcox.

I think myself that designing a 100% sterile rover / lander should be a top priority. It would be expensive to start with, but well worth it.

Once we have built the first one and developed the understanding we would have a basic design there that could be used to explore regions such as the subsurface oceans of Europa and Enceladus and the sensitive sites on Mars even if they have cms thick liquid water or more, and yet not have any concerns about introducing Earth life.

The long term pay off would be huge.

It would obviously take a lot of ingenuity for the astrobiologists, to redesign instruments to be able to be heat sterilized. They did however succeed for Viking, at the temperatures used there. With the Viking sterilization, tenfold reduction every 5 hours, at a dry heat of 125 °C, in theory you wouldn’t need to continue for that long to have pretty much 100% certainty that there is no life left at all.

If anyone knows of any work on this apart from Brian Wilcox’s proposed mission, do say!


Anyway the current rules are not as strict as that. But they do require a lander to be sterilized to Viking levels or higher if they target regions where there is ice within 5 meters of the surface. The reasoning is that a crash could end up melting the ice.

So first here is a map of special regions as updated in 2016, but they also decided that even outside of those regions you need to do case by case studies before landing there.

There Are Regions On Mars That It's Forbidden To Explore


This is a surprising development over the last decade, starting only ten years ago in 2008.

First the Phoenix lander actually spotted droplets forming on its legs.

Unfortunately, it wasn't equipped to analyse them but the leading theory is that these were droplets of salty water. They were observed to grow, merge, and then disappear, presumably as a result of falling off the legs.

Nilton Renno, who was on the team for Phoenix and also runs the REM “weather station on Mars” for Curiosity was one of several who investigated various ways for those droplets to form.

He found that liquid water can form very quickly on salt / ice interfaces when the salt is on top of the ice. By “salt” there he means calcium perchlorate salts similar to the salts they found in the Phoenix site.

Within a few tens of minutes this salt on top of ice formed droplets of liquid brines in Mars simulation experiments. This is striking as it could open large areas of Mars up as potential sites for microhabitats that life could exploit. The professor says

"If we have ice, and then the salt on top of the ice, in a few tens of minutes liquid water forms. Our measurements clearly indicate that. And it's really a proof that liquid water forms at the conditions of the Phoenix landing site when this salt is in contact with the ice.

"Based on the results of our experiment, we expect this soft ice that can liquefy perhaps a few days per year, perhaps a few hours a day, almost anywhere on Mars. So going from mid latitudes all the way to the polar regions.

" This is a small amount of liquid water. But for a bacteria, that would be a huge swimming pool - a little droplet of water is a huge amount of water for a bacteria. So, a small amount of water is enough for you to be able to create conditions for Mars to be habitable today'. And we believe this is possible in the shallow subsurface, and even the surface of the Mars polar region for a few hours per day during the spring."

(transcript from 1:48 onwards)

That's Nilton Renno, who lead the team of researchers. See also Martian salts must touch ice to make liquid water, study shows . He is a mainstream researcher in the field - a distinguished professor of atmospheric, oceanic and space sciences at Michigan University. For instance, amongst many honours, he received the 2013 NASA Group Achievement Award as member of the Curiosity Rover " for exceptional achievement defining the REMS scientific goals and requirements, developing the instrument suite and investigation, and operating REMS successfully on Mars" and has written many papers on topics such as possible habitats on the present day Mars surface.


This is another suggested habitat for life in the Mars higher latitudes based on processes that happen in the Antarctic ice. Dust grains in the ice often produce tiny melt ponds around them in the heat of the summer sunshine. The dust grains absorb the heat (preferentially over the ice), and so heat up and melt the surrounding ice. Then this heat gets trapped because of the insulating effect of the solid state greenhouse effect, because ice traps heat radiation, so forming tiny melt ponds of a few millimeters thickness or more. This could happen on Mars too, so is another possible habitat with fresh water.

It's just a few millimeters of fresh water, but that could be significant on Mars. Another example of this process, then meteorites in Antarctica are often found associated with gypsum and other evaporates - minerals that can only form in the presence of liquid water and must have formed after they fell in Antarctica. Sometimes the researchers find capillary water, or thin films of water, and sometimes they even find evidence of a rather large meltwater pond which formed around the meteorite, or find the meteorites in depressions filled with refrozen ice.

A similar process could be at work in the Martian icecaps too. This process could melt the ice for a few hours per day in the warmest days of summer, and melt a few mms of ice around each grain. Indeed, if I can venture a speculation of my own, perhaps just as in Antarctica, there could be larger melt ponds around meteorites embedded in the ice too - as Mars must have many meteorites embedded in the polar ice sheets.

This could explain another puzzle. Particles of gypsum (the same material that is used to make plaster of paris) have been detected, first in the Olympia Undae dune fields that circle the northern polar ice cap of Mars, See this paper for details. Later on, they were detected in all areas where hydrated minerals have been detected, including sedimentary veneers over the North polar cap, dune fields within the polar ice cap, and the entire Circumpolar Dune Field. There's strong evidence that the gypsum originates from the interior of the ice cap. See this paper for details. Gypsum is a soft mineral that must have been formed close to where it has been discovered (or it would get eroded away by the winds) and as an evaporite mineral, it needs liquid water to form. Opportunity later found veins of gypsum in the equatorial regions, in 2011, a clear sign of flowing water on ancient Mars. But these polar deposits are more of a mystery because they are found in the dust dunes on Mars, so must be produced locally, but where?.

Losiak, et al, modeled tiny micron scale dust grains of basalt (2-2 microns in diameter) exposed to full sunlight on the surface of the ice on the warmest days in summer, on the Northern polar ice cap. They found that these tiny dust grains were large enough to provide for five hours of melting which could melt six millimeters of ice below the grain. They say that with pressures close to the triple point, on windless days, you should get a significant amount of melting. They speculate that this might possibly explain the deposits of gypsum in the polar regions. Could it have formed in a similar way to the gypsum that sometimes forms around Antarctic meteorites?

Möhlmann did a similar calculation. This time he was looking at the possibility of liquid water forming inside snow on Mars. The snow would be exposed to the vacuum, but as the ice melted it would plug all the pores in the snow and eventually form a solid crust of ice on the snow, and so protect it from further evaporation. It would trap the heat as well and so encourage melting. This could happen anywhere between a few centimeters depth down to ten meters below the surface.


This is an interesting suggestion by Möhlmann in an article in Cryobiology magazine, that life may be able to make use of thin film monolayers of the " ULI water" (Undercooled Liquid Interfacial water) wrapped around a microbe, even in tiny nanometer scale layers of liquid water only two monolayers thick.

"In view of Mars it should be mentioned, that there is water ice in the permanent polar caps. At mid- and low-latitudes, ice can form, at least temporarily, via adsorption and freezing in the soil. There, the adsorbed and frozen water overtakes the role of ice, as described above. So, ULI-water can be expected to, at least temporarily, exist also in martian mid- and low-latitudinal subsurface soil. A similar environment can be expected to exist in isolation heated parts of icy bodies in the asteroidal belt, and analogously in the internally heated icy moons of Jupiter and Saturn. It is thus a current and challenging question if ULI-water can act as supporting life in environments with temperatures clearly below 0 °C by delivering that water, which is necessary for metabolic processes, and by permitting transport processes of nutrients and waste. It is the aim of this paper to demonstrate the potential importance of ULI water in view of the possible biological relevance of nanometric undercooled liquid interfacial water."

He cites research suggesting life can remain active in the presence of just two monolayers of water wrapped around a microbe.If there is just a small thermal gradient in the ice, of one degree centigrade per meter, then enough liquid water will form to fill a micrometer sized microbe once a month. Enough will form to fill it once a day if there is a locally steeper gradient of one degree centigrade per 10 cm. This can lead to a constant transport of fresh water to bring fresh nutrients to the microbe, and to remove wastes. The main question is whether this is a sufficient flow of water to sustain life. For more details of this intriguing idea, see his article.


There are two main types of these flow-like features. For a technical overview of them, see the Dune Dark Spots section in Nilton Renno's survey paper. These ones in the southern hemisphere which form in Richardson crater are particularly promising because all the current models involve liquid water in some form and what's more, in the models, these features start off as fresh water trapped under ice.

The more interesting ones, for habitability, are in the south. The southern ice cap consists mainly of dry ice. It is colder, and higher up (at a higher altitude). It stretches as far as forty degrees from the pole in winter (so spanning over 4,700 km), but it reduces to just 300 km across in summer, Richardson's crater is 17.4 degrees from the south pole (that's over 1,000 km).

So though the features resemble each other in appearance, the conditions in which they form are very different and not directly comparable. The southern hemisphere features from at much higher surface temperatures than the northern hemisphere features, and they appear late in spring, after the rapid disappearance of a vast and thick layer of dry ice that covered the entire southern polar region, and beyond. In the summer then surface temperatures at Richardson crater can actually get above the melting point of ice at times in daytime, as measured by the Thermal Emission Spectrometer on Mars Global Surveyor. (See figure 3 of this paper)..

This map shows where the crater is. It is close to the south pole - this is an elevation map showing the location of Richardson crater in Google Mars, and I’ve trimmed it down to the southern hemisphere. You can see Olympus Mons as the obvious large mountain just right of middle, and Hellas Basin as the big depression middle left. Richardson crater is about half way between them and much further south.

Here is a close up - see all those ripples of sand dunes on the crater floor?

Link to this location on Google Mars

Well it’s not the ripples themselves that are of special interest, Mars is covered in many sand dune fields like that planet wide. What interests us are some tiny dark spots that form on them which you can see if you look really closely from orbit.

And, would you ever guess? Although it's one of the colder places on Mars, there's a possible habitat for life there in late spring? It is due to the "solid state greenhouse effect" which causes fresh water at 0°C to form below clear ice in Antarctica at a depth of up to a meter, even when surface conditions are bitterly cold.

The Warm Seasonal Flows often hit the news (probable salty brines on sun facing slopes). But for some reason, the flow-like features in Richardson crater are only ever mentioned in papers by researchers who specialize in the study of possible habitats for life on Mars.

I first learnt about them in the survey of potential habitats on Mars by Nilton Renno, who is an expert in surface conditions on Mars (amongst other things, he now runs the Curiosity weather station on Mars). You can read his survey paper here, Water and Brines on Mars: Current Evidence and Implications for MSL. The models I want to summarize here are described in his section 3.1.2 Dune Dark Spots and Flow-like Features under the sub heading "South Polar Region". But it's in techy language so let's unpack it and explain what it means. I will also go back to the papers he cited, and some later papers on the topic.

In the case of Richardson's crater, both models involve liquid water in some form, and also potentially habitable liquid water. One of the two main models involves relatively thick layers of fresh water below optically clear water ice, up to tens of centimeters thick, and so is very promising for microhabitats. The other model involves microscopically thin layers of fresh water that join together to make a larger stream and pick up salts on the way out. That's very promising too. So let's now look at these two ideas in detail.

First, early in the year, you get dry ice geysers - which we can’t image directly, but see the dark patches that form as a result and are pretty sure this is what happens:

Geysers which erupt through thick sheets of dry ice on Mars. Clear dry ice acts as a solid version of the greenhouse effect, to warm layers at the bottom of the sheet. It is also insulating so helps keep the layers warm overnight. Dry ice of course at those pressures can't form a liquid, so it turns to a gas and then explosively erupts as a geyser. At least that's the generally accepted model to explain why dark spots suddenly form on the surface of sheets of dry ice near the poles in early spring on Mars.

So that would be cool enough, to be able to observe them, video them and study them close up. I hope the rover would be equipped with the capability to take real time video. These geysers are widely known and many scientists would tell you how great it would be to look at them up close, and see them actually erupt.

But most exciting is what happens later in the year, when it is getting too warm for the thick layers of dry ice needed for geysers. These layers of dry ice vanish rather quickly in spring. You would think that the dark spots that you get in the aftermath of the geysers would just sit there on the surface and gradually fade away ready to repeat the cycle next year. But no. Something very strange happens. Dark fingers being to form and creep down the surface as in this animation. Very quickly too (for Mars). I haven't been able to find a video for this, as the papers just use a sequence of stills, so I combined together some of the images myself into an animation to show the idea:

Flow-like features on Dunes in Richardson Crater, Mars. - detail. This flow moves approximately 39 meters in 26 days between the last two frames in the sequence

All the likely models for these features, to date, involve some form of water. Alternatives that one might try to use to model them might include a second ejection of material by the dry ice geyser, or dust deposition, but researchers think these are unlikely to produce the observed effects.


The Richardson crater flow-like features should not be confused with two rather similar looking features, the dark streaks in Russell crater, 55 degrees from the south pole (compared to 17.4 degrees for Richardson crater).

These are braided, divide, recombine and cross each other's tracks. They flow down the slopes channeled by wind formed ridges in the dunes, and most distinctive of all, they are able to rush up over small features of up to two meters high and down the other side.

These seem to be dry features associated with defrosting and small dust avalanches as they are episodic, moving rapidly at speeds of 2-4 meters per second like an avalanche. The authors call them "dark flows". For details see this paper.

They also should not be confused with the Flow-like features in the Northern polar dunes

The two Martian ice caps are rather different. The northern cap is low lying, mainly ice, with a thin layer of dry ice that disappears in summer. The flow like features in the northern hemisphere form at 12.5 degrees from the pole at surface temperatures of about -90°C, which is low enough for dry ice to be stable on the surface. Their models involve either extremely cold salty brines or dry ice and sand. These features are far too cold to be habitable to Earth life and may not even involve liquid water

They are easily confused because they are so similar in appearance, and because both are referred to as "flow like features".

These are thought to form at much lower temperatures. Some of the models for these also involve liquid water but there are other hypotheses as well, some of them involving dust and ice slipping down the cliff faces.

Perhaps one reason the Richardson crater flow-like features get so little attention is that it is easy to confuse them with these other features and assume they have been proved to be dust flows or to form at temperatures to low to be habitable.

But they form in different conditions at different temperatures and the explanations used for these other features don’t work for them. Currently the only models for them involve fresh liquid water beneath the ice, either as layers cms thick, or as thin undercooled liquid water layers, then combining with salts to form the flows on the Martian surface.


So, these southern hemisphere flow like features seem very promising. That’s not as surprising as you might think. The same thing happens in Antarctica - if you have clear ice, then you get a layer of pure water half a meter below the ice.

The water is trapped by the ice so stays liquid. And what’s more, if they model it assuming clear ice like the ice in Antarctica they find that the ice there gets enough heat from the sun in the day to keep it liquid through the night to the next day so the layer can actually grow from one day to the next (ice is an excellent insulator). Also the Mars atmosphere is so thin that it doesn't matter at all that the air above the ice is very cold in these regions. The atmosphere is a near vacuum and works as a great insulator. Better in some ways than Antarctica.

Inuit village, Ecoengineering, near Frobisher Bay on Baffin Island in the mid-19th century - ice and snow are very insulating on Earth or on Mars. Just as you can be snug and warm inside an igloo, a layer of fresh water can stay warm a few tens of cms below the surface, warmed by the sun every day beaming through the clear ice. The near vacuum of the Mars atmosphere helps if anything.

Möhlmann's model is pretty clear (abstract here). If Mars has transparent ice like the ice in Antarctica, then it should have layers of liquid fresh water 5 - 10 cm below the surface and a couple of cm in vertical thickness in late spring to summer in this region. His model doesn't involve salt at all, so the water would be fresh water.

The only question here is whether clear ice forms on Mars in Mars conditions and whether the ice is sufficiently insulating. We can’t tell that really from models, the only way is to go there and find out for ourselves.

Blue wall of an Iceberg on Jökulsárlón, Iceland. On the Earth, Blue ice like this forms as a result of air bubbles squeezed out of glacier ice. This has the right optical and thermal properties to act as a solid state greenhouse, trapping a layer of liquid water that forms 0.1 to 1 meters below the surface. In Möhlmann's model, if ice with similar optical and thermal properties forms on Mars, it could form a layer of liquid water centimeters to decimeters thick, which would form 5 - 10 cm below the surface.

In his model, first the ice forms a translucent layer - then as summer approaches, the solid state greenhouse effect raises the temperature of a layer below the surface to 0°C, so melting it.

The melting layer is 5 to 10 cm below the surface. In the model, then the ice below the surface is first warmed up in the daytime sunshine, due to a greenhouse effect, the infrared radiation is trapped in the ice in much the same way that carbon dioxide traps heat to keep Earth warm. Then because the ice is so insulating, the heat is retained overnight, and the water remains liquid to the next day. To start with it would be only millimeters thick but over several days, gets to thicknesses of centimeters.

He found that subsurface liquid water layers like this can form with surface temperatures as low as -56°C.


This should happen on Mars so long as it has ice with similar properties to Antarctic clear ice.

If there is a layer of gravel or stone at just the right depth, the rock absorbs the infrared heat and that can speed up the process. In that case, a liquid layer can form within a single sol, and can evolve over several sols to be as much as several tens of centimeters in thickness. That is a huge amount of liquid water for the Mars surface.

The fresh water of course can't flow across the surface of Mars in the near vacuum conditions, as it would either freeze back to ice, or evaporate into the atmosphere. But the idea is that as it spreads out, it then mixes with any salts also brought up by the geyser to produce salty brines which would then remain liquid at the much lower temperatures on the surface and flow beyond the edges to form the extending dark edges of the flow-like features.

Later in the year, pressure can build up and cause formation of mini water geysers which may possibly explain the "white collars" that form around the flow-like features towards the end of the season - in their model this is the result of liquid water erupting in mini water geysers and then freezing as white pure water ice

This provides:

  • A way for fresh water to be present on Mars at 0 °C, and to stay liquid under pressure, insulated from the surface conditions.
  • 5 to 10 cm below the surface, trapped by the ice above it
  • Depending on conditions, the liquid layer is at least centimeters in thickness, and could be tens of centimeters in thickness.
  • Initially of fresh water, at around 0°C.

They mention a couple of caveats for their model, because the surface conditions on Mars at these locations is unknown. First it requires conditions for bare and optically transparent ice fields on Mars translucent to depths of several centimeters, and it's an open question whether this can happen, but there is nothing to rule it out either. Then, the other open question is whether their assumption of low thermal conductivity of the ice, preventing escape of the heat to the surface, is valid on Mars.

The process works with blue ice on Earth - but we can't say yet what forms the ice actually takes in these Martian conditions. The authors don't go into any detail about this, but ordinary ice can take different forms even in near vacuum conditions. As an example of this, the ice at the poles of the Moon could be "fluffy ice"

"We do not know the physical characteristics of this ice—solid, dense ice, or “fairy castle”—snow-like ice would have similar radar properties. [then they give evidence that suggests fluffy ice is a possibility there] "
(page 13 of Evidence for water ice on the moon: Results for anomalous polar)

That's the main unknown in their model, whether the ice is blue ice like Antarctic ice, or takes some other form. The ice should at least be in the same hexagonal structure crystalline phase as ice is on Earth - Mars is close to the triple point in this ice phase diagram

Phase diagram by Cmglee, wikipedia. Ice outside of Earth can be in many different phases. For instance in the outer solar system it is often so cold that it is in the very hard orthorhombic phase, where it behaves more like rock than what we think of as ice. However ice on Mars is likely to be in the Ih phase similar to Earth life. The Mars surface is close to the triple point of solid / liquid / vapour in this diagram.

So, the ice is likely to be of the same type as the blue ice in Antarctica. Not likely to have bubbles of air in it. But it could still take a different forms. The model shows that Mars should have layers of liquid water ten to twenty centimeters below the surface if there are any areas of clear blue ice as in Antarctica.

This solid state greenhouse effect process favours sun facing slopes (equator facing). Also, somewhat paradoxically, it favours higher latitudes, close to the poles, over lower latitudes, because it needs conditions where surface ice can form on Mars to thicknesses of tens of centimeters. (The examples at Richardson crater are at latitude -72°, longitude 179.4°, so only 18° from the south pole. There is no in situ data yet for these locations, of course, to test the hypothesis. Though some of the predictions for their model could be confirmed by satellite observations.


Another model for these southern hemisphere features involves ULI water (Undercooled Liquid Interfacial water) which forms as a thin layer over surfaces and can melt at well below the usual melting point of ice. In Möhlmann's sandwich model, then the interfacial water layer forms on the surfaces of solar heated grains in the ice, which then flows together down the slope. Calculations of downward flow of water shows that several litres a day of water could be supplied to the seepage flows in this way.

The idea then is that this ULI water would be the water source for liquid brines which then flow down the surface, mixing with dust, to form the features. That would still be interesting as you end up having flowing liquid water on Mars, several litres a day what’s more. Here is a paper from 2016 describing the idea.

See also Möhlmann's paper The three types of liquid water in the surface of present Mars

Those are the only two models so far. So it does seem very likely that there is liquid water here, and even with the interfacial liquid layers, the water starts off as fresh water beneath the ice, or possibly salty (in either model) if there are salt grains in the ice for the water to pick up. Either way the features start out as a flow of fresh water trapped beneath a layer of ice. This is one of the least publicized types of habitat on Mars, seldom mentioned outside the specialist literature. Yet in some ways it's one of the most interesting, if it exists, because of the potential for fresh water at 0 °C.

This liquid water is hard to observe because the features are so small, beyond the resolution of CRISM. However, analysis of the larger spots, at around the spring equinox, produced a signal that just possibly could be liquid water, where the ice is in contact with the dark material of the dune spots.

" In the gray ring area the water ice 631 surrounds darker surface, where liquid interfacial water layer or brine (Möhlmann 2004, 632 2009, 2010) may form. We found no firm evidence for the presence of liquid water in near-IR 633 spectra, although linear unmixing results show that the data are not inconsistent with a 634 possible slight contribution (a few %) of liquid water in the dark core unit." page 26 of this paper.


Möhlmann has also suggested that his process could be a more widespread phenomenon in the Mars ice caps, not just associated with the geysers, as for Antarctica. Just more noticeable for the flow-like features because of the conditions in which it forms there.

Liquid water could form at a depth of around 6.3 cm wherever there is optically clear ice on Mars in snow / ice packs, just as it does in Antarctica. In summer, it could form layers from centimeters to tens of centimeters in thickness.

Results of Mohmann's modeling of the solid state greenhouse effect in clear ice on Mars. The plateaus show temperatures that get above the melting point of water regularly every Martian sol, at depths of about 6.3 cms. L here is 11.4 cm. Ice at this level will melt periodically, and especially in summer can stay liquid overnight, leading to subsurface liquid water in layers of from cms to tens of cms in thickness. This should happen on Mars not just in the flow-like Features of Richardson crater, but also, anywhere where there is optically clear ice.

In another paper he writes "This liquid water can form in sufficient amounts to be relevant for macroscopic physical (rheology, erosion), for chemical, and eventually also for biological processes. "

His models seem clear enough. The air temperature hardly matters, because the Mars air is so thin it's a near vacuum, insulating the ice, like a thermos flask. The only unknown here is whether Mars does have optically clear ice like this, which is common on Earth in cold conditions like this in Antarctica.

Before I go on to the last couple of examples of possible habitats in the polar regions, let’s just revisit the Phoenix lander site. I think it would be a great place for a mission that’s both interesting for astrobiology and also for ground truth for planetary protection.


So, this is another suggestion, that we could find habitats on Mars inside ice fumaroles. It's a nice idea, and perhaps ice fumaroles do form on Mars from time to time. So far we haven't found any on present day Mars. But it may well be worth keeping a look out for them, as it would be a very interesting habitat if we find one, or one of them starts to form, around a volcanic vent on Mars. If Mars does have any volcanic vents which vent water rich gases through a fumarole, they are likely to form ice towers like this, as happens in Antarctica.

Let's look at the idea in some more detail. This photo shows an ice fumarole - an ice tower that forms around a vent of volcanic gases in the extremely cold conditions right near the top of Mount Erebus in Antarctica.

One of the numerous Ice Fumaroles near the summit of Mount Erebus in Antarctica. If these also occur on Mars, they could provide a habitat for life, and would be extremely hard to spot from orbit due to the low external temperatures. Image credit Mount Erebus Volcano Observatory

For more photos of ice fumaroles see "Ice Towers and Caves of Mount Erebus",

They were originally discovered by the Antarctic explorer Shackleton during his 1908 Nimrod expedition, when he and a few others set out to climb Mount Erebus.

Photograph from Shackleton's Mount Erebus expedition with a fumarole in the background

He described them like this.

"The ice fumaroles are specially remarkable. About fifty of these were visible to us on the track which we followed to and from the crater, and doubtless there were numbers that we did not see. These unique ice-mounds have resulted from the condensation of vapour around the orifices of the fumaroles. It is only under conditions of very low temperature that such structures could exist. No structures like them are known in any other part of the world."

Ice caves form below the fumaroles, and these are especially interesting as a habitat for life.

Entrance to Warren Cave on Mount Erebus. Credit Brian Hasebe. Volcanically heated, the temperatures inside their three study sites were 32, 52 and 64 degrees Fahrenheit (2,11 and 18 degrees Celsius), far warmer than the surroundings.

These ice caves on Erebus are of especial interest for astrobiology, as analogues for habitats outside of Earth, because they are so biologically isolated. Most surface caves are influenced by human activities, or by organics from the surface brought in by animals (e.g. bats) or ground water. These caves at Erebus. are high altitude, yet accessible for study. There is almost no chance of them being affected by photosynthetic based organics, or of animals in a food chain based on photosynthetic life. Also there is no overlying soil to wash down into them.

As described in this paper, these ice towers eventually collapse and then rebuild themselves, but though temporary features, they persist for decades. The air inside has 80% to 100% humidity, and up to 3% CO2, and some CO and H2, but almost no CH4 or H2S. Many of the caves are completely dark, so can't support photosynthesis. Organics can only come from the atmosphere, or from ice algae that grow on the surface in summer, which may eventually find their way into the caves through burial and melting. As a result most micro-organisms there are chemolithoautotrophic i.e. microbes that get all of their energy from chemical reactions with the rocks. They don't depend on any other lifeforms to survive. They survive using CO2 fixation and some may use CO oxidization for their metabolism. The main types of microbe found there are Chloroflexi and Acidobacteria.

This makes them very interesting as an analogue for Mars habitats. If Mars is currently geologically active, then in such cold conditions, it may well have ice fumaroles around its vents, and if so they would be only a few degrees higher in temperature than the surrounding landscape and hard to spot from orbit. We haven't found these yet. The closest we have got so far is that the silica deposits in Home Plate which Spirit found, might have been formed by ancient fumaroles on Mars, (not necessarily ice fumaroles) though they could also have been formed by hot springs or geysers.

This article Martian Hot Spots in NASA's Astrobiology magazine presents Hoffman's ideas. He explains that ice fumaroles on Mars could be up to 30 meters tall in its lower gravity and 10 to 30 meters in diameter, circular or oval in shape. So, potentially these things could grow to be huge on Mars, as high as a nine story high skyscraper, and potentially some of them could be as wide as they are high.

He suggests searching for them on Mars from orbit, and he wondered if some temperature anomalies in Hellas Basin could be ice fumaroles. They wouldn't need to be in polar regions because the fumaroles themselves would bring large quantities of water vapour to the surface to keep replenishing the ice towers as they sublime away in the thing Mars atmosphere. They might be quite easy to spot as white circles or ovals, probably in permanently shadowed regions, and they would be slightly warmer than their surroundings. This shows one of his candidates.

Daytime infrared from Odyssey IR

Anomalous warmth in infrared at night as well on all nine infrared bands, so not a chemical signature.

That candidate is in Hellas Planitia and is from 2003. Despite a search of high resolution visual images they were unable to find anything visual corresponding to them, they were only visible in infrared. But it shows the sort of thing they would be looking for. Lots of small dots around 10-30 meters in diameter each, clustered around a potential fracture. For details see their paper.

The idea is that just as on Earth, volcanic action could bring water vapour and other gases from below. The water vapour, as in Antarctica, would freeze out to form these ice towers. If these environments do occur on Mars, they would provide a warm environment, high water vapor saturation, and some UV shielding. The ones we have on Earth don't have significant amounts of liquid water. However, as they have close to 100% humidity inside, that doesn't matter. They sustain microbial communities of oligotrophs, i.e. micro-organisms that survive in environments that are very poor in nutrients. The same could be true of Mars.

Though we haven't found ice fumaroles on Mars yet, we have found recently formed rootless cones, which are the results of explosive contact of lava with water or ice. This shows that ice (or water) and lava were in close proximity as recently as around ten million years ago.

This shows rootless cones on Mars (to the left) and in Iceland. They are the locations of small explosions of steam, when lava surges over the surface over water or ice. These rootless cones on Mars formed around ten million years ago which shows that Mars has had ice and lava in close proximity very recently. They range in diameter from 20 meters to 300 meters.

So, could there be other ways that volcanic processes on Mars produce habitats by interacting with ice, such as the ice fumaroles? From this 2007 paper:

Hoffman and Kyle suggested the ice towers of Mt. Erebus as analogues of biological refuges on Mars. They combined the idea of still existing near surface ice deposits with the assumption that there is still some localized volcanic activity on Mars today.

There are several examples from Mars that show a direct interaction between lava and ice in the geological history of Mars. The most obvious cases are the rootless cones seen in the northern lowlands. HRSC images show direct and violent interaction in the relatively recent geological history, for example at the scarps of Olympus Mons. Mars today is in relatively dormant phase, and any interactions which might be occurring today are presumably on a much less dynamic scale. Nevertheless, they may be driving local hydrothermal systems. Studying the geothermal processes in the first few tens to hundreds of meters below the surface of Mars today might thus uncover a wide variety of new habitats where biological activity may survive on this cold and dry planet.

For more about this topic see Volcano-Ice Interaction as a Microbial Habitat on Earth and Mars. These ice fumaroles would be of great interest, but of course, being open to the surface, would easily be contaminated by Earth life from surface explorers or brought in to them through dust from the Martian storms.

So far we've been looking at habitats deep below the surface of Mars, though perhaps connected to the surface. But what about habitats on the surface itself? They would make planetary protection even more of an issue, so it's important to look at the possibility. First we need to look at the question, is surface life possible there at all. Just a decade ago, most scientists (with the exception of Gilbert Levin) would have answered with a resounding "No". But that's all changed.

There might also be habitats for native Mars life below the surface similar to lake Vostok in Antarctica - well within reach of drilling. Searches so far have turned up a blank but they could still be there if the lakes are small up to a few kilometers in size. They could be as close to the surface as only 100 meters deep below rock, or 600 meters deep below ice and remain liquid indefinitely


When comet Siding Spring was discovered in 2013, before they knew its trajectory well, there was a small chance that it could hit Mars. Calculations showed it could create a crater of many kilometers in diameter and perhaps a couple of kilometers deep. If a comet like that hit the martian polar regions or higher latitudes, away from the equator, it would create a temporary lake, which life could survive in.

Artist's impression of Mars as seen from comet Siding Spring approaching the planet on 9th October 2014. It missed, by less than half the distance to our Moon. But sometimes comets will hit the Mars ice caps or higher latitudes. If that happens, it will create lakes and hydrothermal systems that last for thousands of years.

These lakes can last for a surprisingly long time, insulated by the ice and heated from below by the rock. The models suggest that large craters of 100 - 200 km in diameter in the early solar system would have made lakes that stayed liquid for as long as one to ten million years. This happens even in cold conditions, so it is not limited to early Mars. A present day comet a few kilometers in diameter could form a crater 30 - 50 km in diameter and an underground hydrothermal system that remains liquid for thousands of years. The lake is kept heated by the melted rock from the initial impact in hydrothermal systems fed by water from deep underground.

Also, there's another way to keep water liquid. Any ice deep enough below the surface, only 100 meters deep, can actually stay liquid indefinitely if covered by an insulating layer of gravel. There'd be enough heat from below, just from the heat of Mars itself and enough insulation above from the gravel, to keep the water permanently liquid. See section 2.2.3 of Niton Renno's article. This is also one theory for the Martian "dry gullies" that they formed through liquid water suddenly flowing out of a subsurface aquifer like this. This was the most popular theory for them at one point, though there are other explanations for them now.

It's much harder to keep water liquid below ice, since rock is much more insulating than ice. It's especially hard for water to form below an ice sheet. If the ice cap was four to six kilometers deep, then you'd expect the base of it to be liquid water, melted from below just through the heat of Mars itself. Though Mars does have ice at both poles, its ice sheets aren't quite as deep as that. But it could still have liquid water at the base of its ice sheets, if there's localized geothermal heating from below.

Also, if a lake formed, originally by geothermal melting or a meteorite impact, it's much easier to keep the lake liquid than it was to melt the water in the first place. In one model, then if a lake forms at a depth of over 600 meters below the ice (originally open to the surface) then it can remain liquid indefinitely from the heat flux from below, even without local geothermal heating.

We'd be able to detect this water using ground penetrating radar because of the high radar contrast between water and ice or rock. MARSIS, the ground penetrating radar on ESA's Mars Express is our best instrument for the job. After several searches, it hasn't found anything yet. See page 191 of this paper. Their resolution isn't that great, however, around a kilometer.

From the searches done to date, we can say with reasonable certainty that Mars doesn't seem to have an equivalent of our Lake Vostok (250 km by 50 km by 0.43 km deep) beneath its ice caps at present. It could however still have small subglacial lakes of up to a kilometer or so in diameter. They were looking for water liquid through geothermal heating, but their search would surely have found impact lakes too.

So, Mars doesn't seem to have any large lakes created from impacts just now. Nor does it have any major lakes formed through geothermal activity below glaciers or ice caps, though it could have smaller lakes.

So in short there are lots of exciting prospects to explore in the polar regions for astrobiology

So far we haven’t even made a start at looking for life there. Or anywhere on Mars except briefly in the 1970s with the Viking landers which produced ambiguous results and have never been followed up.

See also my