This story has recently hit the news. I think it is reasonably clear he was not putting forward a serious worked out future plan for Mars. But is there any potential in the idea?

He talks about it here at 2 minutes in to this video. He just says "The fast way is to drop thermonuclear weapons on the poles": 
He talks about it here at 2 minutes in to this video. He just says "The fast way is to drop thermonuclear weapons on the poles": 

I found it surprising how seriously his rather off the cuff, and joking, remark has been taken by some reporters, reported as a major news story by many journalists. For instance, the Independent reports Want to make Mars hospitable? Drop nuclear bombs, says Elon Musk and many other headlines of that nature today. Also we've been getting several serious questions about it on quora today as a result, basically social media is buzzing with it today.

And it's an interesting question, so, given that there is so much buzz about it today, I think it is worth looking at it anyway, in detail :).

In response to all those criticisms, Elon Musk later clarified it by saying that he meant nuclear weapons detonating continuously over both poles of Mars to create two new mini suns - a far future sci. fi. scenario.

"What I was talking about," said Musk, "was having a series of very large, by our standards, but very small by calamity standards--essentially having two tiny pulsing suns over the poles.”

Elon Musk Clarifies His Suggestion To Terraform Mars By Nuking It

Which takes it into the realm of science fiction, a factory of some sort making so many nuclear bombs that you can explode them constantly. A nuclear explosion lasts less than a minute

After the lapse of 10 seconds, when the fireball of a 1-megaton nuclear weapon has attained its maximum size (5,700 feet across), the shock front is some 3 miles farther ahead. At 50 seconds after the explosion, when the fireball is no longer visible, the blast wave has traveled about 12 miles. It is then moving at about 1 ,150 feet per second, which is slightly faster than the speed of sound at sea level.

CHAPTER II: Descriptions of Nuclear Explosions

To keep two such suns going constantly you need to make around 3000 nuclear bombs every day, in a production line, constnatly transporting them to Mars and then exploding them over the poles. This is just way way into the realm of Sci. Fi. And even then not at all clear it would do much to Mars.

There has never been any paper suggesting nuclear weapons could terraform Mars, either quickly or slowly, not on their own anyway, AFAIK.

Giant space mirrors, yes, impacting comets yes, artificial greenhouse gases much more powerful than carbon dioxide or water vapour, yes, these have all been suggested, but nukes, they are just not powerful enough. Anyway the atmosphere still would need continual extra warming even if you did liberate the carbon dioxide (CO2),  (that's the reason for the giant space mirrors). or it would just freeze out again pretty quickly.

And anyway, as we'll see, it wouldn't work. Not make it habitable as in Earth habitable or even close. You could however use a nuke to melt a lake in the polar ice caps. It would soon melt over but it would remain liquid for some time beneath the ice. That's about all you could do with nukes on Mars.

Still it may be fun to find out why not, and I'll look at some other ideas for terraforming Mars and also the ethical dilemmas that it would involve if we did ever find a practical way to do it.

For newer version of this article

- Nukes Can NEVER Terraform Mars - A Million Tsar Bomba's, Largest Nuke Ever Made, Can't Do It - Elon Musk's Sci. Fi Teeshirts

What about Kim Stanley Robinson's "Mars" trilogy?

Many of you may be influenced by Kim Stanley Robinson's "Mars" trilogy which gave the impression that we can terraform Mars in a few generations.

This is a series of three books, Red, Green and Blue Mars that came out in the mid 1990s together with a book of short stories with the history as backgrop called “The Martians”. In this he envisions Mars terraformed and developing a planet spanning civilization in a couple of centuries. The main focus is on social issues but he has a backdrop of terraforming with plausible sounding science, and this has influenced many people to think that terraforming Mars would be easy and accomplished as soon as a couple of centuries. See Mars trilogy - Wikipedia

The Martians cover

Kim Stanley Robinson himself says that it would take far longer than his trilogy suggests, which is based on 1980s ideas. He says that we can’t use Mars as a ‘backup planet’. We have to fix our problems on Earth to have any hope of surviving on the timescales of the book. See the podcast here and summary on Io9 here here. His "moholes" wouldn't work at all - it would soon get back into equilibrium, perhaps warm it up by a fraction of a degree for a short while. They don't come from any scientific paper. It's the same for the windmills idea.

It is a science fiction book written for entertainment, and he says his main aim was to reflect on conditions on Earth. But it is not a scientific blueprint for terraforming Mars, and was never intended as such.

See my answer to In Kim Stanley Robinson's "Mars" trilogy, how long does it take to terraform the planet and how scientifically accurate is he?

This article goes into a lot of detail about the nukes idea, so a short summary may help. 


First, nukes are nowhere near as powerful as a comet impact. A large comet would release about as much energy as hundreds of thousands of the Tsar Bomba, at fifty megatons, the most powerful nuclear weapon ever tested.

A single large comet impact is equivalent in yield to around 625,000 of these - the most powerful nuclear weapon ever tested

Then, comet impacts also are not powerful enough either, as they haven't terraformed Mars yet.

Comet impacts haven't terraformed Mars

It is true that Mars can be pushed into a runaway greenhouse effect, where all the dry ice gets evaporated into the atmosphere.

However, to get to that point you need to have an atmosphere that is about 10% of the atmospheric pressure of the Earth at sea level. Mars current atmosphere is 1% so you need to somehow release nine times the current amount of CO2 in its atmosphere. 

That amounts to around 140,000 cubic kilometers of dry ice you need to liberate somehow to get a runaway greenhouse effect (one cubic kilometer of dry ice masses about 1.6 gigatons and the Mars atmosphere is about 25 terratons, and you need nine times that amount to reach 10% of the Earth's atmospheric pressure, and 9*25,000/1.6 = 140,000 cubic kilometers approx).

From this graph, Chris McKay, a leading planetary astrogeophysicist, deduces that to get Mars to go into a runaway greenhouse effect using just CO2, it needs about 100 millibars - or same pressure as 10% of Earth's atmosphere, to be released into the atmosphere.

It is not clear whether Mars has that much dry ice. There is at least 10,000 cubic kilometers, so enough to nearly double its atmosphere. But to end up with ten times its current atmosphere is less certain. 

If you don't release enough CO2, then it will revert to its current state because a denser atmosphere would be out of equilibrium with the vapour pressure of the dry ice at its poles, so it would all condense back to its poles as dry ice

The amount of CO2 that could be released by the largest nuclear bomb ever exploded, at fifty megatons, is tiny, even if you dropped it directly on a known dry ice deposit at the poles and all the energy went into liberating CO2, and the dry ice didn't need to be raised in temperature, just sublimated - it would liberate less than a cubic kilometer of dry ice.

To double the amount of CO2 in the atmosphere you need 20,000 of those, with all the energy put into turning the CO2 into gas.

To increase it to ten times the current atmospheric level, if there is that much CO2, you need 180,000 of them. But in practice, given that the CO2 is sure to be patchy, sometimes thin layers, spread out, mixed with the soil, and that only a fraction of the energy of the explosion goes into sublimating the CO2, you are talking about millions of them.

Now, though using nuclear explosions is surely impractical, in princtiple, you might be able to do it with powerful greenhouse gases.

Unlike the nuclear bomb idea, there are some papers on this topic and some think is the easiest way to do it. But this is still is a large scale project.

Based on the published research on this topic, you'd need to mine over 11 cubic kilometers of fluorite ore on Mars. More than that if you have to use ores with lower concentrations of fluorine. Then after that, it requires the output of over 200 nuclear power stations on Mars, working on creating these gases 24/7 for a century before you can push it into the new state.

And then, even a thick CO2 rich atmosphere on Mars is not enough to keep it warm enough to be habitable for much more than microbes and hardy lichens and such like. That's because Mars gets only half the sunlight of Earth. Sunlight warms planets relative to absolute zero, for instance if it is 0 C on Earth that's 273 C above absolute zero, most of that warming due to the sunlight.

On Mars then with half the sunlight, if you do a simplified calculation of the temperature, without greenhouse effects, it would be -64 C average compared to -18 C for Earth.

Table from here: Predicted Planetary Temperatures - American Chemical Society

This is ignoring effects of any internal heat - just looking at the balance of the incoming and outgoing sunlight based on the albedo. The second from bottom row, with temperatures in Celsius in brackets shows the calculated temperature. The bottom row shows the observed average temperature, which is increased by the greenhouse effects

So, our atmosphere which warms Earth up to tolerable temperatures would warm Mars up only to temperatures well sub zero. It is just not possible to have a warm Mars with flowing water, oceans, etc and warm enough for trees, animals etc with an Earth like atmosphere.

After this runaway greenhouse, if you could get it to work, it would be warm enough for liquid water on Mars occasionally, but most standing water would soon turn to ice, and it's not likely to get hot enough for trees to grow.

You'd need constant resupply of greenhouse gases, or space mirrors or such like to keep it warm enough for habitability. And then it would take about a hundred thousand years of photosynthesis to get from that to an oxygen rich atmosphere.

Also, once you get an oxygen rich atmosphere, you then lose the warming effect of the CO2 (because oxygen and nitrogen are not greenhouse gases - water vapour is, but you don't have enough of that). 

So, you would need to increase production of the greenhouse gases when the CO2 is converted to oxygen - or build more space mirrors to keep the planet warm or it would plunge back to its present cold state. 

The natural average surface temperature of Mars, if you gave it an atmosphere like Earth, is around -50 �C. (By comparison, the same figure for Earth is 16 �C).

Fluorine based gases could be used to heat up the atmosphere enough, but it would take more than 11 cubic kilometers of this ore, or more if it has a lower proportion of fluorine, to make enough powerful greenhouse gases to warm up Mars enough, and require output of over 200 nuclear power stations for a hundred years.

There are many more issues with terraforming Mars, which I touch on - some of those issues are covered in more detail in my Trouble with Terraforming Mars.

I also look at Chris McKay's idea, which we could call "Mars forming" Mars - that if we find interesting life there, to try to return it to the conditions of early Mars instead of attempting an Earth like atmosphere.

I also look at the potential for using our resources and the solar system resources for kilometer scale space habitats instead. This is similar to terraforming, but an easier project. They could be completed in decades rather than millennia and we can respond quickly to fix mistakes which on planets would take centuries to fix or not be possible to fix at all. And I look at how you could explore Mars arguably more thoroughly, and in a more interesting way also for the explorers, and more safely, from its moons Deimos or Phobos or from orbit.

Exploring Mars via telepresence from a base on Phobos, or in this illustration, Deimos

I also look at the potential of a comet impact - or indeed could be nuclear bombs - to create lakes in the polar regions that could remain liquid for thousands of years - this probably happens naturally from time to time. But they would be anaerobic lakes on Mars, perhaps CO2 rich but with no oxygen.


If you went all out and exploded thousands  of 1 gigaton bombs on Mars, what it could do is briefly release a lot of dry ice (CO2) into the atmosphere as CO2, also water vapour, which would make the Mars atmosphere briefly a bit thicker. But nowhere near as thick as Earth. This happens naturally whenever large comets hit the Mars poles, and as you can see, they haven't made it into a habitable planet.

The other effect would be to create a lake at the poles. Which would soon freeze over. But below the insulating covering of ice, it would remain liquid for a while.

But it would be much better to deflect a large comet to hit Mars if you want that effect.

Even quite small bolides hitting Earth atmosphere release the equivalent of small nuclear bombs in terms of the amount of energy released into the upper atmosphere when they create bright fireballs briefly brighter than a full moon.

To interpret this table, a kiloton is about 4,184 gigajoules. So the ones labeled as 10,000 gigajoules are in the kilotons range TNT equivalent

Mars  is closer to the asteroid belt and gets hit about ten times more often than Earth, and with its thin atmosphere, then even small asteroids get all the way to the surface. On Mars these wouldn't be air bursts, they'd be surface hits, and there would be ten times as many of them as for Earth. 

It gets hit by meteorites large enough to create a small crater 200 times every year.

One of many fresh impact craters on Mars spotted from orbit. It gets hit by 200 meteorites large enough to create small craters like this every year (much higher frequency than Earth because closer to the asteroid belt, and also thinner atmosphere). See Pow! Mars Hit By Space Rocks 200 Times a Year

It gets hit by the equivalent of a 1 megaton impact every 3 years estimated . On Earth the equivalent figure is once every 15 years.

It gets hit by a 10 megaton impact every century. See Earth and Mars crater-size frequency distribution and impact rates:theoretical and observational analysis

So, it gets hit by impacts with the equivalent of multi-kiloton nuclear weapon impacts frequently and multi-megaton level impacts would be common. And on Mars with its thin atmosphere, these make it all the way to the ground, pretty much the full yield is released at ground level.


The largest nuclear weapon ever detonated on the Earth, the  Tsar Bomba, had a yield of 50 megatons (or possibly 57)

This is its bomb casing

And the fireball:

The Tsar Bomba fireball, about 8 km in diameter.

A 130 meter diameter meteorite hitting Earth has a similar yield to the Tsar Bomba at about 60 megatons (though in the case of Earth much of it would be dissipated in the atmosphere on the way down).

Conceivably we could build a 1 gigaton nuclear weapon in the future with 20 times the yield of the Tsar Bomba.

But a 1 km diameter comet has a yield of around 100 gigatons. 

A 10 km diameter comet has a yield of around 100,000 gigatons. 

On Mars, the impact velocity is roughly halved for short period comets (not so much for long period comets) - that still makes the yield around 25 gigatons and about 25,000 gigatons respectively.

So you'd need 25,000 gigaton bombs each one 20 times more powerful than the largest nuclear weapon ever built, to equal the effect of a single one 10 km diameter comet hitting the poles of Mars. And that's a size of comet that would hit Mars from time to time.

I got those figures from Defending Planet Earth - page 36 - by the Space Studies board. Other sources give somewhat varying figures, but whether it is 60,000 or 100,000 gigatons is neither here nor there.


We can also come at this another way. The latent heat of melting for ice is quite high. 334 Kj/kg.

A kiloton is about 4,184 gigajoules, so that's enough to melt 4184000000/334 kg or about 12,500,000 kg. Or about 12,500 cubic meters.

A Tsar Bomba has a yield of 50 megatons. So, if somehow all the energy could be directed into melting ice into water and none into heating up rock or dissipated into the atmosphere could melt 50,000 times that, so given that there are a billion cubic meters in a cubic kilometer, it could melt 0.625 cubic kilometers of water (625 million tons = 50,000*12,500). That of course is for ice that is already at 0 C. It would also heat up the ice from temperatures far lower than that to melting point too, but lets ignore that for now.

Spread over the entire surface of Mars, surface area of 144.8 million square kilometers, then that melt water becomes a global equivalent layer of 0.0000043 meters (1000*0.625/144,800,000), or about 4.3 microns of water.

The water currently in the Mars atmosphere is the equivalent of a global layer of 10-20 microns of water - and in the northern summer, reaches a maximum level of 60 - 70 microns. Scientists 'map' water vapor in Martian atmosphere

This plot shows how the amount of water vapour in the atmosphere of Mars varies seasonally, and by latitude, from 5- 10 microns to a maximum of 60 - 70 microns. A single Tsar Bomba would increase this by 3.6 microns momentarily - and that is if you could find a way to focus all the energy from the bomb into melting ice and ignoring the heat needed to raise it some tens of degrees to 0 C first, and assuming that somehow all the water remains in the atmosphere and doesn't fall out as rain or ice, as it surely would - the Mars atmosphere doesn't have a high capacity for water vapour and already reaches 100% humidity every night.

So one Tsar Bomba, if all the energy went into melting water at 0 C, and if all that water then evaporated into the vacuum of the atmosphere, didn't just fall back and freeze over - it would increase the amount of water vapour temporarily roughly by a tenth of the difference between the lowest and highest figures in the seasonal variation.

This is an over estimate of course. In practice only a tiny fraction of the energy would end up as water vapour. I'm just trying for a first ballpark figure here to show the issue.

If we used 25,000 gigaton bombs (500,000 Tsar Bombas), the equivalent of our 10 km diameter comet, this could melt the equivalent to a layer of about 2.2 meters of water over the surface of Mars. 

So this would melt a fifth of the estimated 11 meters of water equivalent over the surface of Mars thought to be available on Mars according to one estimate.

But - in practice nothing like that amount of energy would go into melting the ice. And any water that got into the atmosphere would almost immediately condense out again as ice, or fall back as water and freeze over - or if it did get further afield, would penetrate into the dry equatorial sands and be lost. The equatorial region is as dry as the Sahara desert to a depth of at least a hundred meters, possibly hundreds of meters (except for a few patches of rather surprising anomalous subsurface regions of ice).

Then much of that water would remain on the surface at the impact point as a lake that would freeze over. At least, that's what happens with models of giant impacts on Mars.

The idea I think is to try to kickstart a greenhouse effect, not make significant changes to Mars right away. I can't find much about this, as it's not the most usual suggestion for a way to heat up Mars. If anyone knows of a good cite on this do say in the comments.

Michio Kaku briefly mentions nuclear bombs here:

But he gives it just as a passing remark, I doubt if he means that nuclear bombs by themselves could do it.


The situation is much the same for dry ice, latent heat of melting / sublimation is 184, if all the energy went into sublimating dry ice, then each Tsar Bomba would liberate 50,000*4184000/184, or 1.137 billion tons of CO2, just a tiny fraction of the 25 trillion tons (teratons) of the Mars atmosphere. 

So, that's only 0.00455% of the existing atmosphere for each Tsar Bomba - and to do significant warming even by a few degrees you'd have to thicken the atmosphere considerably.

Dry ice has a density of 1.6 tons per cubic meter. So that 1.137 billion tons corresponds to 1.137 / 1.6 or 0.71 cubic kilometers.


You could try to double the atmospheric pressure. It's thought this happens from time to time naturally so permitting liquid water to lie on the surface of Mars and to flow in the deeper regions. Some of the gullies are thought to be formed by liquid water at times in the past when the atmosphere was thicker.

The debate has gone back and forth for several years on the gullies on Mars. Originally thought to be possibly water carved, then new ones were seen to form, clinching the case for dry ice in their formation. But now there's another emerging hypothesis that's gaining favour that the more recent smaller ones made from dry ice and the older ones made from flowing water at a time when the Mars atmosphere was a bit thicker, as recently as half a million years ago. If this hypothesis is correct, then  in paleolithic times, our first tool using ancestors may have lived in a solar system with liquid water flowing on Mars.

If so, then liberating the 10,000 cubic kilometers of buried dry ice in the Martian South Pole could potentially return Mars briefly to that past era of flowing water - but would the dry ice stay in gas form if we did that - or would it require mega-engineering such as orbital mirrors or greenhouse gas factories on Mars to warm the planet?

Though Mars may not have enough dry ice for a thick atmosphere of carbon dioxide (there is considerable uncertainty here), there is probably enough dry ice on Mars to approximately double its atmospheric pressure to around 2% of Earth's pressure.

That's as a result of discovery in 2011 of 10,000 cubic kilometers of buried dry ice at its poles - so about 16 teratons. It would be enough to warm it up so that you get liquid water stable on the surface.

But, even with this modest goal, to thicken Mars' atmosphere to the extent that water can survive on its surface without instantly boiling at just a degree or two above zero, it doesn't seem too likely that nukes could do the trick.

Discovery of resources of 10,000 cubic kilometers of dry ice at the Martian South Pole - which suggests that at times its atmosphere may get thick enough for liquid water to be stable on the surface - this may happen naturally at times of high axial tilt, when the polar ice caps evaporate naturally. See also the Scientific American article about it.

In this image, Red corresponds to about 600 meters or yards thick; yellow to about 400; dark blue to less than 100, tapering to zero.

So you'd need hundreds of thousands of these Tsar Bombas to start to make a significant difference, with all of the energy somehow focused into liberating dry ice into CO2.

Or, given that only a tiny fraction of the released energy would go into sublimating dry ice, and most go into heating the underlying rock or radiated into space - more likely millions or tens of millions of them.

And when you do, the carbon dioxide would most likely just condense back to the surface as dry ice. You probably need some long term change such as change in eccentricity of the Mars orbit (so its perihelion is closer to the sun and aphelion more distant than it is now) or change in its tilt (obliquity) to keep it as gas. Or your megatechnology mirrors of course.

Also, there doesn't seem to be enough left there anyway for significant warming of Mars or a thick atmosphere, at least on the basis of what we know so far, though there's always scope for new deposits to be found. 

If you think liberating a small amount of water or dry ice from the ice caps like this could cause a runaway greenhouse effect of some sort - well why hasn't it happened already with comet impacts?


When Comet Sliding Spring was first discovered, then for a time they thought it was going to hit Mars, and then could have done this experiment for us naturally, but it missed.

Artist's impression of Comet Sliding Spring's close flyby of Mars in Autumn 2014. At one point it had a chance of hitting Mars and if it hit the higher latitudes, it could have done this experiment for us, and we could have seen if it created a long lived lake, or had any effect on the atmosphere etc. Impacts like this can happen on Mars from time to time, and modeling suggests they create subsurface lakes, but they haven't terraformed Mars, and the energy released dwarfs anything we could do with nuclear weapons. Image credit NASA.

The subsurface lake would be a temporary habitat for life on Mars a bit like the subglacial lakes in Antarctica. This happens from time to time and is one of the possible habitats in present day Mars for life there.

The models suggest that if a comet a few kilometers in diameter hit a high attitude region of Mars, the result would be a crater 30 - 50 km in diameter and an underground hydrothermal system that remains liquid for thousands of years. 

It would be a habitat for microbes.

It would be an anaerobic lake, so no oxygen, even if you have hydrothermal vents form in the lake. The Antarctic lakes are nitrogen and oxygen rich, even hyperoxygenated, through the effect of dissolved air getting into the lakes as the snow at the surface turns to ice, then the ice at the bottom of the covering layer of ice melts, a continual downward conveyer of Earth atmosphere into the lakes. But this is just a way of forcing the existing atmosphere into the lakes at high pressure. On Mars with a CO2 atmosphere, the same process if it happened at all, would make the lakes CO2 rich, not oxygen rich.

For more about this, see Lakes on Mars (Google eBook),


As it turns out, some scientists have actually worked on this, especially Chris McKay, NASA astrogeophysicist, special interest in the origin of life, who has been working on this since the early 1990s.

What he found out is that the Mars atmosphere has a "tipping point" that if you can increase its CO2 levels from its current 1% of Earth's to about 10% of Earth's, then it will go into a runaway greenhouse effect and all the CO2 will go into the atmosphere - it will no longer be stable on the surface as dry ice. But if you are not able to increase it by that much - say - that you increase it only to 5% of Earth's atmospheric pressure - then eventually the CO2 will all condense back at the poles returning it to its current state.

The comet impacts and the nuclear bombs don't get you anywhere near this required level of CO2. But powerful greenhouse gases can. So also can orbital mirrors - by directly heating Mars so changing the parameters directly. But he focuses most on greenhouse gas production. So let's just take a quick look at what he found out - for the techy details see his papers.

This shows the dynamics of the situation on Mars. The solid curve shows the greenhouse temperature / pressure relationship for a CO2 atmosphere on Mars. The atmosphere has to maintain an equilibrium between the vapour pressure of the dry ice at the Mars poles and the pressure of the CO2 in the atmosphere.

It is currently at point A, a point that depends on the current conditions on Mars (axial tilt and orbital eccentricity), and any attempt to perturb it by releasing CO2 puts the caps out of equilibrium with the atmosphere. The dry ice then condenses back to the polar caps, returning the system to its current state.

To get it to a runaway greenhouse effect by releasing CO2, we have to move it to point B. This involves making the atmosphere about ten times more dense than it is now, up to about 10% of the Earth's atmosphere. At that point, if there is enough CO2 in the Martian regolith, we might get a runaway greenhouse effect warming the planet until all available CO2 is released. Nuclear bombs, or a large comet impact, can't release anything like enough to do this however.

(Figure from The physics, biology, and environmental ethics of making mars habitable (page 103))

Chris McKay puts it like this in his earlier 1997 paper on the subject (with other collaborators), Making Mars Habitable

"From our analysis, one could propose the following sequence of events: production of CFCs (or other greenhouse gases) starts on Mars and the surface temperature warms by ~ 20 K. The regolith and polar caps release their CO2 and the pressure rises to 100 mbar.

One of two things could then happen. If there were large regolith and polar CO2 reservoirs, the pressure would continue to rise on its own. If these were absent, the CO2 pressure would stabilize, and additional CO2 would have to be released from carbonate minerals. At this stage (perhaps between 100 and 100,000 years), Mars may be suitable for plants.

If there was a mechanism for sequestering the reduced carbon, these plants could slowly transform the CO2 to produce an O2 rich atmosphere in perhaps 100,000 years. If sufficient N2 could also be released from putative soil deposits, and the CO2 level kept low enough, then a human-breathable atmosphere would be produced.

Continued production of CFCs that absorb radiation across the whole spectrum would be required to maintain the warm temperature. Destruction of ozone by these CFCs would probably require these gases be made in sufficient amounts (considerably in excess of current terrestrial production rates) to constitute an ultraviolet shield."

[I've added extra paragraphing and rephrased his 10^5 years to 100,000 years for easier readability by non technical readers]

He advocates greenhouse gases as the easiest way to do it. In his later 2004 paper (with another collaborator) he suggests using non chlorine or bromine greenhouse gases, so you no longer have that issue with the ozone layer.

In his 2004 paper with other authors, he also attempts an estimate at how much power is generated to make those greenhouse gases (at forty billion metric tons it is just not practical to transport them from Earth).

He reckons that if you had 245 power stations on Mars each generating half a gigawatt of power continuously, that would be enough power to supply factories on Mars that could generate enough greenhouse gases to reach this tipping point in a hundred years - mining resources on Mars to make the gases.

In a 2005 paper with Margarita Marinova as the principle author, they do detailed modeling of various greenhouse gases, and come up with a mixture of gases that can tip it into a runaway greenhouse at 0.2 pA. Since one pA is about a thousandth of the current Mars atmosphere (100,000 Pascals to a bar), that's 0.025 teratons, or 25 billion tons of the gases, a somewhat smaller figure.

This shows the effect of introducing the greenhouse gases. This changes the greenhouse temperature / pressure relationship. To start with you get a small amount of warming, but eventually as the temperature gets high enough, you reach a point where at all points along the curve, the balance is in favour of the dry ice  subliming into the atmosphere, and you get a runaway greenhouse effect. That then would move Mars into a new state with a CO2 dense atmosphere and release whatever quantities of CO2 are available on Mars in the polar caps and the regolith, ending in a situation with CO2 no longer stable on the surface in a solid state long term.

They found that compounds of fluorine gave the best results. You also need carbon and sulfur for his optimal mix of 15% C2F6, 62.5% C3F8 and 22.5% SF6 (I make that mixture 83% Fluorine by weight). (for details see his paper, section 5.3).

He doesn't go into this, but with 25 billion tons of greenhouse gases to produce in that century, you'd also have to mine many cubic kilometers of fluorite ores (or whatever is needed for your gases) to produce them as well. At the normal density of fluorite ore, of up to 3.5 or so, and given that fluorite, CaF2, is 50% fluorine by mass, and if the greenhouse gases are 83% fluorine by weight, that would be over eleven cubic kilometers of ore required for the project. If the fluorine was present in lesser concentrations in the Mars deposits than it is in fluorite ore, you'd need more material than that to make the gases.

In any case, it's clear that a significant level of mining activity would be needed to mine the fluorine on Mars.

He estimates that you would need 25700 times Earth's current yearly production of these gases during this phase of the project. Once you reach the desired concentration, you need about 3 times Earth's current yearly production to maintain the concentration, although it might be that the gases persist longer on Mars than on Earth which could reduce that "top up" requirement.

Fluorite ore (photo by Rob Lavinsky). To generate enough greenhouse gases to raise the temperature enough for a runaway greenhouse effect - you need to mine enough fluorine and other materials to make around 25 billion tons of greenhouse gas, which you could get by mining 7 cubic kilometers of fluorite ore if it exists on Mars. This would trigger a runaway greenhouse effect, end state currently not known as it depends on the amount of dry ice in the Martian regolith.

What happens after that depends on how much CO2 is available as dry ice - as you might then need to find a way to extract CO2 from carbonate rocks if most is now in the form of rock.


You might think that another way to achieve the runaway greenhouse effect is to extract CO2 from carbonate rocks, which on Mars, on the surface anyway, occur mainly as magnesite rather than limestone. Mars seems to be rich in sulfate deposits after all, compared to Earth, seems likely to be plenty available.

So what about mining those deposits to make sulfuric acid and using it to liberate the carbon dioxide? Well I haven't seen it suggested as a way to terraform Mars, and if you think it through, you can see why. The problem with this scenario, as you have the equivalent of well over 100,000 cubic kilometers of dry ice to liberate - so, probably no need to do the calculations, it's going to be roughly 100,000 cubic kilometers of sulfuric acid that you need to create and of sulfate rocks you need to mine (within an order of magnitude anyway). That's a far larger mining operation than the one needed to mine and process 11 cubic kilometers of fluorite deposits.


Without greenhouse gases, you need at least 100 millibars available to tip it into the runaway greenhouse state in the first place, or all your COwill end up at the poles eventually over long enough timescales, and it will return to its original state. 

With enough of these very potent greenhouse gases, they do the warming by themselves, so all the CO2 will go into the atmosphere, even if it doesn't have much. But you still need to get it to above that 100 millibars point if you want the atmosphere to be long term stable in its new state without need to keep manufacturing greenhouse gases on Mars. 

The end result depends on how much CO2 there is on Mars. So far only a bit less than 1 millibar has been definitely confirmed, as far as I know, those ten thousand cubic kilometers of dry ice at the South pole. That wouldn't be enough to tip it into a stable end state with a denser atmosphere without some extra forcing such as continual production of greenhouse gases or orbital mirrors.

Mars may have large quantities of dry ice deposited in the regolith, but it might also be that all the missing CO2 from the early atmosphere has been transformed into carbonates through chemical reactions in the past involving water. The surface of Mars is rich in sulfates and not in carbonates. But SO3 is rapidly consumed by the surface layer of soil when present - and then CO2 is not able to condense in sulfate rich soil because it is highly pH sensitive, and so, since the surface of Mars is sulfate rich, any carbonates will have to form at a lower layer in the soil. (See pages 16 - 17 of Geochemical Reservoirs and Timing of Sulfur Cycling on Mars).

Surely there must be large deposits of carbonates, or dry ice ,or both at depth - because where else could the ancient atmosphere have gone to? CO2, unlike H2O is not easily lost from the Martian atmosphere into space. It would only require a small percentage of it to be available as dry ice to make this scenario work. But how much is there? Various estimates. Zent et al in 1995 estimate 30 - 40 mbar which would not be enough unless there are substantial not yet detected deposits at the poles.

If anyone knows of any more confirmed deposits of dry ice on Mars, or estimates for the total amount of dry ice in the regolith, do let me know!


If we do find viable Mars life on Mars, especially if it has a different biological basis from Earth life, Chris McKay in his environmental ethics advocates restoring the past ecology of Mars - making it habitable for Mars life, in interests of maintaining and increasing diversity of life. Perhaps in analogy to "terraforming" we can call this "mars forming" Mars - restoring conditions of past Mars.

"A further conclusion from these tenets of deep ecology is that indigenous martian life, different from Earth life, should be the first aim of restoration ecology on Mars, thereby increasing the diversity of life. Even though it is unlikely, it may be possible to find viable martian life or revive dead but frozen martian microbes. If life on Mars did experience its own separate genesis, then restoring that life to global diversity would have to be the best possible option for Mars."

(from page 105 of "The Physics, Biology, and Environmental Ethics of Making Mars Habitable")

Recent developments make the possibility of present day Mars life more likely than before  though whether it would be similar to Earth life or have a different biological basis is a big question - exobiologists of course hope that it will turn out to be different as that would be of immense interest to biology.

Here is Nilton Renno talking about a recent discovery that suggests that "swimming pools for bacteria" may be widespread on Mars. He was co-investigator for Phoenix, and is co-investigator for Curiosity - and he is responsible for the Curiosity REM "weather station on Mars" which reports surface conditions there including the ionizing radiation levels.

(click to watch on youtube)

There are many other suggestions for habitats on present day Mars, after the surprising discoveries of the Phoenix lander in 2008, especially droplets of what seem to be water on its legs perhaps from deliquescing salts - and its isotopic evidence from the atmosphere that liquid water in geologically recent past, either sporadically or ongoing, is involved in extensive interactions with the atmosphere.

See my Are There Habitats For Life On Mars? - Salty Seeps, Clear Ice Greenhouses, Ice Fumaroles, Dune Bioreactors,... for an overview of this rapidly developing field, which has many citations to the academic literature to find out more

This goal of "Mars forming" is easier to achieve, and may be possible quite quickly, with greenhouse gases. If you can trigger the runaway greenhouse effect, no more transforming of the atmosphere is needed, because a CO2 atmosphere is your desired end goal. 

However - for it to remain in this new region of stability by itself, after you warm it up, you still need a hundred millibars of CO2 on Mars, and this still needs to be confirmed that it is, as far as I know.

Maybe somehow we could even import enough volatiles to recreate early Mars? Or is this atmosphere there, in the deep carbonate rocks, and could it be released in some way?

Now that would be quite something!

It seems easier to foresee the consequences here, with less by way of ethical dilemmas. There seems less potential for "terraforming going wrong" as you are not attempting to make it go into a new Earth-like state that it has never entered before, you are restoring it to a state that it seems it has already been in in the past.

So, if it does have enough carbon dioxide to get into this stable warmer state, we might have a reasonable practical expectation that the end result would be stable at least for millions of years, without need to extrapolate far beyond the knowledge of Mars we have already.


Wonderful though it might be to restore an early Mars like that, by releasing CO2, still, it might not be long term stable on geological timescales. There are a few things we would need to look into.


Eventually on the millions of years timescale, it's going to lose its oceans and atmosphere again. The water will eventually be dissociated and hydrogen lost to space.

The COwill be converted to carbonate rocks in any oceans, or adsorbed onto mineral and soil surfaces in the soil zone (as happens on Earth, an important sink for CO2) or into the deeper regolith (which is thought to be very porous on Mars), or buried as organics.

Unlike Earth, Mars has no continental drift to subduct buried deposits of carbonates or organics and then return them to the surface in volcanoes, so there will be a constant gradual loss of carbon from the system to carbonates or buried organics, unless we can devise some biochemical process to return it.

There are ideas for dealing with this in a terraformed Mars however, which we could also use for our "Mars formed Mars".


One way, the carbonates could be turned back to CO is by introducing acidified thermal water into the carbonate deposits. This could also be a method of getting CO from the carbonates in the early stages of Mars Forming. Dan Popoviciu in his "Some Ideas Regarding the Biological Colonization of The Planet Mars", mentions this idea and also suggests Matteia sp, a cyanobacteria that not only photosynthesizes, and fixes nitrogen, but also dissolves carbonates.

He got this idea from Friedmann, so for the details, see Friedmann et al's Terraforming Mars: dissolution of carbonate rocks by cyanobacteria (1993)

On the problem of carbon lost into limestone deposits Dan Popoviciu just says this.

"Unlike the situation on Earth, the martian lithosphere is unitary. There are no tectonic plates, rifts os subduction zones. This prevents the efficient recirculation of the chemical substances in the lithosphere, mantle and atmosphere. Apparently, this contributed, in the past, to the loss of most of the planet?s atmosphere. For example, carbon (of biological origin or not) fixed, as carbonates or hydrocarbons, in sedimentary rocks, would "sink" into the crust, being unable to come again to the surface. The effects of this would be visible only after very long geological eras."

"The solution that mankind will find to this problem cannot be anticipated. It would be possible the creation of biochemical cycles that would replace the geological ones."


Apart from the problem of CO2 being absorbed in the soil and deeper in the regolith, and deposited in oceans as limestone, you also have the problem of photosynthetic life.

On Earth, then an oxygen rich atmosphere is a benefit, it cools it down and it stopped it turning into a copy of Venus. With a CO2 atmosphere our oceans would boil and that would create a runaway greenhouse effect through the water vapour, with a Venus like planet with surface temperatures of 400 C a likely end point. Without photosynthetic life, that is probably what would have happened on Earth.

But on Mars we have the opposite problem. It needs greenhouse gases to keep warm. So if Mars has photosynthetic life already, or it gets introduced accidentally - then without mega-engineering - as soon as it starts to convert the CO2 into oxygen, the planet cools down. Once the partial pressures of CO2 get to less than the equivalent of 10% of the Earth's atmospheric pressure - then unless you have other greenhouse gases such as methane to compensate, the atmosphere is no longer stable, and will revert back to its present state with the dry ice at the poles.

So, before "Mars forming" Mars, if we ever did it, we'd also need to think through implications of what photosynthetic life might do to our new biosphere long term,


Here we intend to leave Mars to its own devices, and it might not develop a way of returning the CO2 to its atmosphere by itself (especially if you think there is an element of luck involved in the way weak Gaia finds solutions to keep planets habitable).

Once that happens it would have lost the remaining 30 meters or so of water and its remaining CO2 and become a totally lifeless rock.

It managed to keep liquid oceans in the early solar system - but that was with a hotter sun - and we don't know if it had photosynthetic life at that stage, or at all.

If you introduced more gases with comets however, you could end up with a Mars with more volatiles than it has today. But you would still have the question - is this long term stable, and if not, can we make it so - or is it just stable for a few million years? 

In a future with mega technology you could continually restore the volatiles with comets.

Or perhaps, if it doesn't find a solution by itself to return the carbonates from limestone to the atmosphere, we might by then know enough to be able to introduce some new lifeforms to set up a cycle that constantly returns the deposited limestone as CO2, and that is stable long term? And that creates abundant methane and in some way regulates the photosynthetic life so it is not a problem? Or could that happen naturally?

Perhaps we could get insight into this from studying expolanets, and from smaller scale experiments with kilometer scale Mars simulation habitats in space?

Even "Mars forming" Mars, trying to return it to an earlier stage, seems to be not without its risks and potential to go wrong.


Returning to the idea of terraforming Mars, rather than "Mars Forming" - there is a lot more to it, beyond just getting the atmosphere thick enough.

I go over this in more detail in Trouble with Terraforming Mars. So if you've read that, I may be going over things you've read already. But for those who are new to it, in brief, and with some new information as well:


First, Mars gets half the sunlight levels of Earth.

The first thing that would happen if you add lots of water to Mars somehow is that its polar ice caps would get larger. They are tiny even compared to Earth's and it is further from the sun.

Animated fly around of the Mars polar ice cap by ESA, using images from the ESA's Mars Express

Then - even if you can manage to get the atmosphere as thick as it was, say, in the Hesperian age on Mars (the age of floods, when there was still water flowing on the surface, often in large rivers, but often for short time periods then drying up) - and an atmosphere of entirely CO2, greenhouse gas, then it would not warm it up sufficiently by itself to keep ice free liquid oceans year round.

A CO2 atmosphere by itself just doesn't provide enough warming for that.

In some of the models of the later stages of early Mars it's thought that the oceans melted only at times of high eccentricity of the planet's orbit, and at times in the orbit when Mars was closer to the sun. Since Mars is currently in a near circular orbit, in those models the ice wouldn't melt at all.

With the recent discovery that early Mars had oceans, this is a matter of ongoing debate in scholarly papers, whether they were covered with ice most of the time, or ice free. And then, how Mars kept warm enough for liquid water if they weren't covered in ice, given that the CO2 by itself doesn't seem to be adequate to do that.

For an example of recent research on this, published in June 2015:

"In the warm and wet scenario, an anomalously high solar flux or intense greenhouse warming artificially added to the climate model are required to maintain warm conditions and an ice-free northern ocean. Precipitation shows strong surface variations, with high rates around Hellas basin and west of Tharsis but low rates around Margaritifer Sinus (where the observed valley network drainage density is nonetheless high). In the cold and icy scenario, snow migration is a function of both obliquity and surface pressure, and limited episodic melting is possible through combinations of seasonal, volcanic, and impact forcing."
Comparison of "warm and wet" and "cold and icy" scenarios for early Mars in a 3-D climate model

See also: The role of obliquity, water vapor and trace gas
greenhouses on the early martian climate

And if you ever get to an oxygen rich atmosphere, you lose the CO2 anyway, so at that point anyway, you have to find some other greenhouse gas to replace it. The fraction of a percent levels that are needed for human habitability are far too low to make a significant difference on Mars.

Terraforming suggestions typically add large mirrors in space, or greenhouse gases to compensate for that.

In Chris McKay's analysis he assumes continued production of those very potent fluoride based greenhouse gases once you get to an oxygen rich atmosphere. Once the CO2 is gone, the natural equilibrium average temperature of a  Mars even with a thick Earth like atmosphere is around -50 C. That's because nitrogen and oxygen do almost nothing to help retain heat. (Chris McKay et al in Making Mars Habitable estimate -55C, and that includes a greenhouse effect due to CO2 set to 1% so much more than for Earth, especially as that also means three times the column mass as well in the lower gravity).

Earth average temperatures are 16C which includes a greenhouse effect of about 33C, much of that due to water vapour. But on Mars at below -50 C the amount of the greenhouse effect due to water vapour would be small.

So your mirrors or greenhouse gases have to raise the temperature by about 60 C, and they have to do this in perpetuity for as long as the planet remains habitable.


Then, most people don't know this, but actually, CO2 gas, though fine at the usual concentrations, is poisonous to humans, at levels above 1%. Even if you have plenty of oxygen, you'll die if you have too much carbon dioxide in the air you breathe. Above 10% it can kill you rapidly with convulsions, coma and death. The usual guidelines for permitted exposure levels at work are a maximum of 0.5% for an eight hour period.

So somehow you want to convert all that CO2 into organics and oxygen, or the most you can have live there are trees. But to do that takes a huge amount of energy, or else, you use photosynthesis. But to generate all that oxygen you need to take the carbon out of the atmosphere permanently. It is not enough to just circulate it around as plants usually do - eventually nearly all of it returns to the atmosphere as they decay.

Carbon cycle - normally carbon dioxide cycles around and back to the atmosphere through microbial respiration and decomposition. So the net terrestrial uptake each year is tiny. On Mars, with half the sunlight, and a third of the gravity - the atmosphere needs three times the mass per unit area for the same surface partial pressure.

Our oxygen on Earth is the result of many thousand years of carbon removed from the atmosphere - if we cut down all the trees, and destroyed all the vegetation, the air would remain breathable for millennia. On Mars, then it would take around 100,000 years to build up an oxygen rich atmosphere using similar processes.

And that of course is very fast compared with the hundreds of millions of years it took to form an oxygen atmosphere on Earth originally in the Great Oxygenation Event.

Graph of the oxygen levels in the Earth's atmosphere for the last 3.8 billion years by Heinrich Holland. From The oxygenation of the atmosphere and oceans

Mars surface is probably already oxygenated from Early Mars which analysis of the Mars meteorites suggests, had an oxygen rich atmosphere,. So the long billion years timescale for absorbing oxygen into the soil and ocean beds in stage 3 of this graph have probably already happened on Mars.

So the timescale is mainly for stage 3, where you see the current oxygen levels reached in a period of less than half a billion years. The idea would be to speed that up to a hundred thousand years if we go with Chris McKay's estimate.

So if you get as far as a thick carbon dioxide atmosphere - you need to grow enough plants for year after year, forming layers of peat or similar - to cover the entire surface of Mars with a layer several meters thick of organics to sequester all that carbon. This part of the process, Chris McKay estimated would take about 100,000 years. The availability of sunlight is the bottleneck here - because it has about half the light levels of Earth. Also, with a third of the gravity, it needs to have three times the mass of oxygen per square meter as Earth does, for the same atmospheric partial pressure.

Then, humans don't do well in an environment of pure oxygen long term at normal atmospheric pressure - we suffer from oxygen poisoning unless the atmosphere is very thin. A pure oxygen atmosphere is also an increased fire risk as both the US and Russia tragically found out in the early years of spaceflight. It is okay if you have a very thin pure oxygen atmosphere, which is still breathable - but - that still leads to an increased risk of fire (nitrogen as buffer gas also helps to suppress fire as it absorbs heat without taking part in the reaction). In the other direction, in a denser atmosphere, if you can increase the nitrogen enough to have oxygen percentages below 13.9% though same partial pressure as on Earth, then fires would no longer be able to spread - the atmosphere itself would be a fire suppressant.

In the ISS they use an Earth like mixture of oxygen and nitrogen, 80% nitrogen, 20% oxygen. In Skylab they used an atmosphere with 75% oxygen and 25% nitrogen. With Apollo they used a low pressure atmosphere of oxygen. In spacesuit EVAs they use oxygen at low pressures (requiring a long period of pre-adaptation in the ISS before they don the suit).

Perhaps living in a low pressure oxygen rich atmosphere with not much buffer gas is possible though with elevated fire risk. I haven't seen a suggestion to do this for Mars however - if anyone knows do let me know.

In Making Mars Habitable Chris McKay suggests aiming for the same column mass as Earth, for radiation shielding. That could be achieved with 390 mbar of total pressure, though for an air like mixture of gases, he suggests 500 mbar, with 200 mbar of oxygen based on high elevations where people are able to live long term on the Earth.

Normally it's assumed that you'd want a buffer gas like nitrogen, Chris McKay says about this, "An atmosphere suitable for humans to breathe on a long-term basis will probably require a buffer gas to prevent oxygen toxicity and spontaneous combustion".

Mars has some argon but not huge amounts. It originally had lots of nitrogen but it is gone, perhaps turned into rock?

So you have to somehow liberate or find huge quantities of nitrogen to use as a buffer gas for the atmosphere. Chris McKay in his paper assumes this nitrogen is available in the soil, which it may be - after all they think that Mars did start off with a nitrogen rich atmosphere and it is not easy for it to escape from Mars, unlike water which splits into hydrogen and oxygen and the hydrogen is easily lost.

And it probably doesn't have enough water either, 11 meters global equivalent layer of ice - which may sound a lot but remember, you've got equatorial regions that are dry to depths of over a hundred meters, probably hundreds of meters - melt that ice and most of it would just sink into the equatorial dry sands.


As Chris McKay says in one of his papers, we can't hope to direct the evolution of the ecosystems of a terraformed Earth, but would rely on Gaia, as in the weak Gaia hypothesis to bring it to a point of balance at some point - many interactions of the ecosystem coming together.

But how much can we rely on Gaia to do this? Do you just need to introduce life and a planet will automatically go into whatever is its best end state for life, to make that planet more habitable and keep it habitable long term? Or is it at least partly a matter of luck, that it worked on Earth?

David Waltham argues strongly in his Lucky Planet book (continuing a line of investigation begun in Peter Ward and David Brownlee's Rare Earth Hypothesis) that there is a large element of luck involved.

Lucky Planet - why Earth is Exceptional and What That Means for Life in the Universe

One of the things he talks about there is the Great Oxygenation Event. This cooled down the Earth at a point where the sun was beginning to get too hot for life. Without that, the Earth would be pretty much uninhabitable for complex life by now. But it is hard to see that event as a response by Gaia to a warming sun, rather it seems like a lucky coincidence, that photosynthetic life developed at just the right time to cool down the atmosphere by removing CO2 and making the atmosphere oxygen rich instead.

Greening of Mars by Chroococcidiopsis. This would remove carbon dioxide out of the atmosphere, so cooling it down. On Earth, this cooled the planet down at just the right moment, when the sun was getting too hot. Was that just pure luck?

On early Mars the same process would cool it down so much it would plunge it into a greenhouse phase (there is no suggestion currently that early Mars did have photosynthetic life, but if it did, it might not have helped in terms of habitability).

If that had happened on ancient Mars for instance, it would have plunged Mars into a permanent snowball phase which it would never have been able to escape from because it needed the CO2 to stay warm, and again, hard to see how "Gaia" would have interfered in that process and stopped the oxygenation of the early Mars atmosphere if that had happened due to photosynthesis - what feedback mechanism could have done that?

Also Earth itself has gone through several snowball phases - or at least slushball phases - when it was only barely habitable at least for surface life. So Gaia seems to be rather unreliable. He also points out that it has actually cooled long term as the sun has got hotter. A feedback effect could keep the temperature stable, or prevent it from getting too much hotter as the sun gets hotter, but it's hard to see a feedback effect actually cooling the Earth as the sun gets hotter.

Whether or not Gaia works on Earth as normally understood, or whether there is also an element of luck involved as well -it seems not something you could be sure of on basis of current knowledge, that it would also work on Mars.

Then, a part from the possibility of the feedback forcing it into a snowball Mars phase, there's also the possibility of different end states from Earth. Particularly, Mars surface is sulfate rich, so you can easily imagine an end state that has levels of hydrogen sulfide or sulfur dioxide, that are fine for whatever life evolves there, but poisonous to humans. Or high levels of methane, a greenhouse gas, which could help keep the atmosphere on Mars warm without need for artificial greenhouse gases - but the result would not be habitable for humans. These though would be perfectly valid end states and for any organisms living there, if they eventually developed intelligence, they would see themselves as living in an environment in which everything was tuned to keep it in a state that is perfect for them.


I'll just touch on this, go into it more in my other articles on the topic. Though it is perhaps the issue that comes up most often in internet discussions, it's only one of many. But important.

The lack of a magnetic field has two effects mainly

  • No protection from solar storms
  • Atmosphere loses water vapour - but very slowly. It would lose the water its atmosphere maybe over hundreds of thousands or millions of years. They'll know more there as they learn more about the past evolution of the Mars atmosphere.

The first, shielding from solar storms - that could be done by local magnetic generators, massive ones shielding an entire habitat - or just shielding your habitats with several meters thickness of regolith over the top - or living in caves, and doing everything by telepresence on the surface.

Any rovers would need solar storm shelters and you make sure you don't get so far from the rover during an EVA that you can't get back into your storm shelter if necessary within a few hours.

Ideally whenever you leave your rover, you'd stay within reach so you can get back into your solar shelter in minutes. The largest solar storms, though they occur only every few decades, can happen almost any time in the solar cycle without much warning.

The second reason - well to combat that you need some way of generating a magnetic field that will remain in place even in some distant future a million years from now or more when the inhabitants of Mars have lost their technology, if that happens.

Though that distant future might seem remote - I think it is a significant thing. 

Terraforming a planet is like having a child. If you make the decision to do that, I think that ethically you commit your civilization to seeing it through so that the planet has a reasonable future, as good as any other planet.

It doesn't seem ethically responsible to start off a planet that you know has an ecosystem that will fall apart a few hundred thousand or million years into the future. Those future beings who live there are only separated from us by time, and it is because of us that they face those issues whatever they are. And maybe a somewhat wiser us, even a century later, could have arranged a far better method of terraforming or maybe done something completely different.

So I think we need a solution to the long term loss of water vapour too. It is a bit hard to think of any technological solution lasting for millions of years. So, that suggests some serious level of mega-engineering, e.g. impacting giant asteroids or even impacting Mercury into Mars if you could somehow use that to start up the movements in its interior that generate the magnetic fields. Not at all impossible (e.g. sending large asteroids on repeated flybys of Mercury to change its trajectory) - but well beyond near future planetary engineering.


There are ideas for all this, but it involves megaengineering - importing water from comets. Could get the nitrogen that way too using methane rich comets from the outer solar system. Large mirrors in space. Etc etc.

Then you have the problem that the planet you just terraformed in this way - with its greenhouse gases or mirrors, needs continual support by high technology to keep it running. If you can terraform it as quickly and easily as that, it can unterraform just as quickly.

And - it is easy to make mistakes in calculations. When they built Biosphere 2, they didn't take account of certain reactions in the concrete that removed oxygen from the atmosphere inside, which was the main reason it failed. What effects on Mars would we be missing in our calculations for this multimillenia terraforming project?

And long term it loses water vapour from the atmosphere. You've still got the solar storms so it would be hazardous to be out of doors and far from your shelters especially in early days when the atmosphere is thin. You've also got the much higher impact rate because it is closer to the asteroid belt.


And that's just a start of the problems. Particularly, what is life on Mars going to do? It's a very different planet from Earth. The Earth ecosystem transported to Mars wouldn't work. Remember, plants have got to generate three times as much oxygen for the same partial pressure, using half the sunlight, to start with. So the lower gravity can have more side effects than you'd expect.

In the very long term, the carbon dioxide would get trapped into limestone in the seas, assuming it has shallow seas. With no continental drift to subduct it, turn it back to carbon dioxide, and return it to the surface, it just stays there as rocks.

Then, the water vapour in the upper atmosphere gets dissociated by solar radiation without protection of a magnetic field. Somehow Mars has already lost probably hundreds of meters of global equivalent layer of water. It might not take much for it to lose the remaining 30 odd meters it has left (including water in the rocks).

Mars has lost hundreds of meters of water in the past, and has only 30 meters left global equivalent (one estimate). Terraforming attempts risk losing the little water that's left.

The end state eventually, in the far future, might well eventually be a planet rich in carbonate rocks, and lacking any ice or water at all. There are ideas for fixing all this - but with no previous experience of terraforming, they are not hugely convincing, that they would work.

And why would life evolving on Mars as we attempt to terraform evolve towards Earth like conditions? What if it gets taken over by organisms, microbes say, that produce large quantities of hydrogen sulfide, or methane, to take an example? Or just aerobes that proliferate, able to use small concentrations of oxygen, that remove oxygen from the atmosphere as fast as we try to add it? Or life that in one way or another evolves to be hazardous to humans, an allergen for instance, or with byproducts that are harmful to humans, or even diseases of humans such as Legionnaires that also infect microbes - and then evolve in the Mars environment of high radiation and UV and return to humans. Or the microbes poison the water supply or the air for us?


And with such a long term project - how do we know that this is what our descendants would want a thousand years from now when the project nears completion of its first phase (this is the usual quoted time to get as far as an atmosphere suitable for trees, without oxygen - that is with megaengineering)? 

Would they want a terraformed Mars, or would they already have plenty of habitats in the asteroid belt, or indeed, have only a small population on Earth and no interest in Mars at all by then?

With the pace of technological innovation and changes, it is probably harder for us to imagine Earth a thousand years from now, than it would be for a medieval person to imagine present day Earth accurately.

And - would we sustain this project for a thousand years, anyway, hundreds of billions of dollars a year most likely? We are lucky to keep a technological mega project going for 30 years at present.

And - perhaps Mars is needed, but not now, maybe it needs to be terraformed 500 million years from now when Earth becomes uninhabitable because of the warming sun. So doing it now - unless we can do it sustainably - may mean it is no longer available in the future when it is most needed.

At some time in the far future, Earth will end up almost in the atmosphere of the sun as it goes red giant. Some time long before then, perhaps half a billion years from now, Earth will be no longer habitable (unless our descendants or other future civilizations on Earth can move it). If so, Mars might be just what we need. Maybe that is when these ideas for terraforming Mars will be most useful? Perhaps it may even become habitable naturally, especially if it is hit by giant comets at some point in the process.

I think it is great to look into and work on these ideas. Can learn a lot about exoplanets this way, and about how Earth works for instance. But we don't have to actually do the terraforming, not yet anyway.


And it's a responsibility also. Like having a baby. But a baby that has a 100,000 year gestation period. And again I don't think myself that humans as a technological species are quite at the stage where we can take on overseeing the birth and first baby steps of a baby planet that needs to be monitored closely for 100,000 years.

But great and fun idea to think about and it may lead to new thoughts and approaches maybe in unexpected ways :).


With the level of technology needed for terraforming Mars - why not just move all those comets and other resources into close to Earth orbit instead of just impacting them into Mars - and build space habitats? 

And use the technology for mirrors around Mars to build thin film mirrors to concentrate sunlight on solar panels in orbit around Earth and beam the energy to Earth and solve our energy problems?

Space Canada's SPS Alpha project. One of several that work by using thin film mirrors to concentrate the sunlight - a technique you can't use on the Earth as winds and weather would tear the thin films apart. And of course in space you get the power pretty much 24/7 (depending on the orbit|).

If you can build giant mirrors to terraform Mars, why not instead use the same technology to solve the Earth's energy problems?

It doesn't seem to make a whole lot of sense to use that technology to pour water onto a desert planet as part of a hundred thousand year project. 

Well not yet anyway, we are at the stage, surely, of attempting multi-decade rather than hundreds of thousands of years projects.


This is something I often encounter as a programmer. Just dive in without planning carefully and you can get a disorganized architecture for your code that is hard to update or modify - and may even lead to you having to bail out and start again with a new program. So a good programmer will spend a lot of time thinking about how to organize the code before they start on a complex project.

It's the same with interstellar travel. If the time it takes a robotic probe to get to Alpha Centauri goes down so rapidly with developing technology that as time continues your arrival date there gets earlier and earlier the longer you wait - then it is not time to launch yet. If the arrival dates start to get later and later, then it may be time to launch, so long as there is not likely to be a sudden new development, or if the time needed has got down to a low value like a few decades.

Similarly with terraforming. Suppose, using Chris McKay's estimate, it takes 100,000 years to full oxygen atmosphere like Earth - and an unstable ecosystem that needs to be constantly maintained including constant creation of greenhouse gases or maintenance of space mirrors - and longer term, for the water, importing material from comets - and lots to go wrong along the way. 

But a century from now, maybe with nanotech, maybe fusion power, or some other approach, someone finds a way to do it in 5,000 years. Two centuries from now they have it down to 1,000 years. Then eventually, say three centuries from now, someone finds a way to do it in a century.

Well, leave it three centuries before you start the project, and you complete it within 400 years instead of 100,000 years. For as long as the end date is continuing to get earlier, it is best to postpone starting on the project.

That's just an example, with made up figures there but it shows the general idea that especially in long term multimillenium projects, then it may actually pay huge dividends in the time needed for the project, to delay the start.

And the thing is - it would of course be different if it was a progressive thing where anything we do now is bound to help towards the goal, if only by a little.

But - our failed primitive attempts on Mars right now could easily mess up the planet in such a way as to make future terraforming or whatever else we want to do there much harder. It may not be possible to bail out and start again.

And there is no hurry to get started, especially for as long as it remains a multiple millenia project. If we wait long enough, maybe we can device some way of terraforming Mars that nobody has thought of, that results in a planet that is sustainable long term right through to the red giant phase of the sun. Or devise reliable tech that is self maintaining that can be relied on to keep it habitable for the indefinite future.

Or we might decide that something else is a better future for Mars than that.


And as for Mars - for now anyway - why not build settlements in orbit around it, instead of on the surface? These could use resources from Deimos, which has already been suggested as a possible source of water for Earth orbit, with Kuck's idea of the Deimos Water Company - so you've actually got an economic reason for being there too, which is very hard to supply for Mars. Of course, after proper scientific investigation of Deimos first and assessing the effects of such activity.

The delta v is not huge to get the water back to Earth - that is assuming it does have water which is not confirmed yet, but it is a type of asteroid in composition that typically often does have substantial amounts of water.

There are enough resources in just Deimos alone to build space habitats around Mars with total living area of 100,000 square kilometers. That's more than twice the size of Switzerland and roughly the same area as Colorado or Oregon.

Artist's impression of human explorers on Deimos - outermost moon of Mars and an ideal outpost for exploring Mars by telepresence. It has a crater at its South pole almost permanently shadowed, amongst the coldest place in the solar system, as well as natural cooling for rocket fuel that needs to be kept at cryogenic temperatures - and many other advantages. Just close to those places of permanent shadow you have a site of continuous sunlight to place your solar panels. No dust storms or night time to interrupt your power. Many other advantages over Mars, and potentially a type of asteroid likely to be rich in resources useful to humans.

Phobos also has several advantages too e.g. its Stickney cater, Mars facing, with natural protection from more than 90% of cosmic radiation and only 40 milliseconds round trip time for telepresence operation on the surface.

Then from Mars orbit, you can explore the surface by telepresence. See my To Explore Mars With Likes Of Occulus Rift&Virtuix Omni - From Mars Capture Orbit, Phobos Or Deimos

Or we could just continue to explore it from Earth - with higher bandwidth we could increase the rate of exploration hugely. I plan a new article on this soon, on Why are our Mars Rovers so Slow? - early version on quora here: my answer to Why don't Mars rovers move faster?. But if you want the excitement and probably faster pace of exploration from a human on site presence, and can find the budget for it, well telepresence from orbit seems a great next step to aim for.

It's a much safer next step anyway and potentially a spectacular mission with orbital views of the whole of Mars - it looks quite Earth like from orbit - and direct telepresence enhanced vision operating of rovers on the surface, with haptic feedback. Which would also drive technological innovation in telepresence technologies.

The spectacular orbit of the Molniya sun synchronous slowly precessing orbit suggested for the early HERRO mission to explore Mars via telepresence (the spacecraft in this video is not meant to represent any particular idea - it is a far future spacecraft in the orbiter program I used to make this video to illustrate the orbit in their papers).

For more about this, again see my To Explore Mars With Likes Of Occulus Rift&Virtuix Omni - From Mars Capture Orbit, Phobos Or Deimos

Safer for the astronauts, who can explore via telepresence, don't have to continually monitor oxygen levels and worry that a sprained ankle would mean they can't get back to their refill supplies in time, or a fall could break their visor or tear out an air hose. And can return to Earth easily in an emergency, and don't have to do the risky landing on Mars.

And safer for Mars too, in the near future certainly we want to keep Earth microbes away from the "special areas" on Mars until we can study them to see if there is life there. But with global dust storms, and microbes imbedded in a grain of dust able to be carried anywhere on Mars, protected from the UV radiation by the dust storm and iron oxides in the grain of dust - how can you guarantee to localize contamination to the human landing site. (This is a point that was made by Carl Sagan decades ago). And how can you guarantee that Mars won't be globally contaminated by Earth life during a hard landing, a Space Shuttle or early Soyuz type accident - which surely would be inevitable at some point with multiple human missions and new technology? In the longer term future if we did decide to terraform Mars using the methods of ecopoesis, you might want to start with algae, as in early Earth - but not introduce aerobes or creatures that eat the algae. With a planet previously seeded with a random mix of Earth life from human landings,, that might be impossible.

Or if we find Mars life there, we might decide to leave Mars to the Martians. Chris McKay has argued strongly that we should be sure to explore Mars reversibly until we know more about the planet, that any contamination we introduce to it can in principle be removed if needed. The surest way to do that is to keep humans in orbit for now.


The one thing we can do on Mars is to create large domed cities - or live in caves. Eventually you end up with large parts of Mars covered in domed habitats.

If you do that, of course you have none of the problems of terraforming inside your habitats - you use artificial methods to heat them, using power stations or whatever. You have the same issues of a space habitat - high maintenance and expensive to construct compared to an Earth home. But perhaps really large habitats, kilometer square, can be easier to maintain per volume enclosed than a small habitat? If you double the radius of the habitat, surface area enclosed goes up as the square, and if you have multiple levels inside, the volume goes up as the cube, and only the outer shell is hard to maintain.

It's the same equation with space habitats of course. I think only experience can tell us which is easier to construct.

Space habitats have the advantage that in the early stages you can move large masses around with almost no power. To move a thousand tons, you just need to give it an initial shove, then another shove to stop it when it gets to the right place - while in a gravity field you have to use power continuously to move things around.

You might think that the resources in space are spaced far apart, so could be a problem - but that's not so difficult as you'd think. What matters is the delta v, not the distance.

Some aspects of the mining operation are easier in space. You can use metal carbonyls to extract metal - first nickel, which is easy to extract at low temperatures, just enclose your small iron meteorite in a bag of carbon monoxide, warmed by the sun, this extracts the nickel, and then print out using a high temperature 3D printer attached to your bag. Then iron and platinum can be extracted by a similar process but at higher temperatures.

You have plenty of energy in the form of 24/7 solar power.

And in space you can use low mass, large thin film mirrors to concentrate that power for your mining operations either directly as heat, or to generate lots of electricity.

It is at least possible that making a space habitat will be easier than making a domed city in vacuum conditions or near vacuum. Probably we can only know for sure with practical experience doing it.

If you can make a space habitat more easily or as easily as a domed city, it has many advantages - you can construct it anywhere convenient in the solar system with the available materials, not limited to a particular planet. It has solar power 24/7. You can construct it to any climate you like and any gravity level you like with artificial gravity.

And, you just need a 300 meter asteroid for a kilometer scale Stanford Torus. The volume goes up as the cube of the diameter, you can figure out that there is enough material in just the asteroid belt for habitats equivalent to domed cities covering and equivalent area to a thousand times the surface of Mars. Even in Deimos, there is enough to make habitats equivalent to domed cities covering the entire area of Arizona.

For more on this, see Asteroid Resources Could Create Space Habs For Trillions: Land area of a thousand Earths.

The surface domed cities anyway have issues of planetary protection on Mars. Depend on you not finding life you want to protect on Mars, where of course the outcome everyone hopes for is that we do find interesting present day life on Mars.

They would introduce life to Mars, in the wastes from the domed cities and the melted ice around them - but if there are habitats on Mars, then widespread on Mars.

The problem then is you are doing accidental planetary ecoengineering there - which may lead to unexpected consequences, not necessarily making Mars easier for humans, at least, if there is a fair measure of luck involved in making a planet habitable. All those possibilities such as life evolving to be hazardous to humans and returning to the habitats, or poisons the water supplies for humans, or whatever.

Also you can introduce life that might fight against any future attempts to terraform Mars. Life that would consume algae, consume oxygen, or whatever. The whole of Mars is interconnected via the global dust-storms, for microbes at least, able to survive for long periods of time in radiation resistant and UV resistant dormant states, and revive when they encounter a suitable habitat.

So they need thought on Mars I think. Perhaps in the future we can start by building kilometer scale domed cities on the Moon, eventually - and at the same time - space habitats from the asteroid belt, and find out which of those work well.

And for Mars enthusiasts, perhaps we can start by building domed settlements on Deimos or Phobos. With its low gravity, it still has the advantage of a "down" side while many of the advantages of an asteroid in terms of easy to move the material around, so is in between a planet and a space habitat.

It would be useful for plant life anyway (plant life seems to have rather low minimum gravity requirements). With the low gravity it would also be possible to export plants easily.

For humans, it rather depends on what level of gravity is needed for human health. That's the same for habitats on Mars or the Moon. Any of those, or all, or none, might need artificial gravity to "top up" the gravity to tolerable levels for humans living there long term. Which then depends on what spin motions humans can tolerate in space conditions (they are very different from Earth conditions so we may be able to tolerate high levels of spin without nausea - experiments on Skylab with the rotating litter chair suggests this possibility though they were not intended at the time for testing effects of artificial gravity spins on humans).


Just to say, many of you may think we need to go to Mars as a backup of Earth. That's getting rather a long way away from the topic of this article, so, I'll touch on it briefly here. I know that several brilliant people have suggested this should be one of our main objectives for sending humans into space, not only Elon Musk. Perhaps most notably, Stephen Hawking. But being brilliant doesn't mean you are always right in everything you say, and that what you say can't be questioned by anyone else!

First, there is no known hazard that is at all likely to make all humans on Earth extinct instantaneously. A very large asteroid impact could, but none of the planets or moons in the inner solar system inside of Jupiter have been hit by anything that big throughout its cratering record, for getting on for four billion years (in case of Venus the cratering record doesn't go back far as it got resurfaced in a global event, we think, a few hundred million years ago, but no reason to suppose it is different).

Some humans would certainly have survived the Chicxulub impact at the end of the Cretacious period. Small mammals did, also birds, turtles, many creatures did, and with our technology we have capability to survive the same things they can survive.

Then, so long as some humans survive on Earth, it remains the very best place in the solar system to rebuild civilization.

It is hard to beat having a breathable atmosphere, and being able to go outside your habitats without a spacesuit. Or having liquid water, or atmosphere and magnetic field protecting from solar storms and cosmic radiation. So, of course, we would rebuild civilization on Earth anyway, not on Mars.

Anyone on Mars would come back here, if they could, as the obvious place to try to rebuild civilization - so why continue on Mars when there is an Earth to rebuild?

As for hazards that arise due to technology - well space colonies would be the very place where such hazards would most likely arise. Conditions never encountered before where new diseases can develop. Advanced technology, requiring technology to survive, such as 3D printers - if we have nanotech in a big way it will probably be used most in space. Once you have millions there, then it is as easy to get hold of a spaceship with capability far greater than any ICBM than it is to get hold of a plane on Earth now. And your space habitats explode, surely, if hit by a spaceship at full speed, never mind nuclear bombs. It must count as one of the worse places for technology hazards to develop and in a future with fast spaceships, then you'd have as much interplay between habitats and Earth as we have between countries.

Anyway, there is only a one in a million chance of an impact by a giant impactor in the next century. And amongst all the natural hazards we face, impacts by giant asteroids are the ones most easy to predict, and about the only ones that we could actually prevent by deflecting the asteroid.

We have already found nearly all the one kilometer or larger asteroids in orbits similar to Earth. We are well on the way to finding the remaining 10% in the outer solar system, hope to find nearly all of those within another decade. And we have ways to deflect asteroids. Especially if you find out about them long in advance, if it has several flybys of Earth with small "keyholes" it has to pass through first, as is the usual situation if you catch them well in advance - a change of delta v of a millionth of a meter per second may well be enough to deflect an asteroid so it doesn't hit Earth.

If you have billions of dollars, say, for your attempt at a backup - if your aim is to prevent human extinction - why not instead use that funding for a massive search to find all the 1 km sized asteroids right out to the Kuiper belt, and beyond? It's the most cost effective way of using funds at the moment to prevent extinction events, and ends up saving the whole world, not just hoping to save a small band of "survivors".

Certainly by a thousand years from now, long before you could make Mars habitable for humans, if we continue the level of technology that would be needed to terraform Mars - by then we will surely know the position of every single rock in the solar system, probably down to ones as small as just a few meters across or less.

So - I don't see any future in this idea of "backing up" myself. Not now. Half a billion years from now this may be exactly what is needed when Earth becomes uninhabitable. But doesn't seem too likely that we can devise a backup now that will be useful half a billion years into the future - indeed an attempt to terraform Mars now, would probably make it less habitable for the far future if it leads to it losing its volatiles eventually.

We can go into space to help the Earth, to identify hazards to Earth and deal with them, to return resources to Earth, to explore, to advance our understanding of biology and our universe, for tourism, for adventure, as poets, authors, musicians, artists. I see many reasons for going into space. But I can't see any future with present day technology in going into space to escape Earth. And if we do develop technology, with 3D printing or self replicating machines, which makes it easy to live in space, it would still be far easier to live on Earth, even in our deserts, or in cities floating on the seas or in the skies, than in space.

Such magical technology would transform all our lives, not just the people in space, and I think we just can't plan at present to take account of the effects. That would be like a medieval person trying to start on a project to benefit people in the twentieth century.

If you are interested to find out more about this perspective , see my "Why we can't back up Earth on Mars, the Moon, or anywhere else in our solar system."


Do you want to download this as an ebook, to read on your kindle, or kindle app (available for most operating systems)?

This is the 2015 article. You can get it here:

Elon Musk's Mars Nukes Idea: Feeble Compared To Natural Impacts On Mars (Kindle book, Amazon)

It's the same article with  addition of table of contents, so you can jump easily to any section, title page and formatted as a book. Text, images etc are the same.

Estimated length - equivalent to 60 pages in a printed book


I talk about the Mars impact lakes and many other potential habitats for present day Mars life in Are There Habitats For Life On Mars? - Salty Seeps, Clear Ice Greenhouses, Ice Fumaroles, Dune Bioreactors,...

As usual, if you spot anything at all to correct in this, however minor, don't hesitate to say in the comments below :). Including typos.

Regular readers of my blog may be interested to know that I've listened to those of you who suggested that some of the posts here are a bit long to read as a single scrolling page. So I've started to convert them to kindle books. Just $0.99, the minimum price, so you can download them onto your kindle. I'll do just about all the longer posts eventually. If you have any favourites you want turned into an ebook that aren't there yet, do let me know, as it doesn't take that long to do, now that I've got it all figured out how to do the conversions. 

If you are interested you can get them here, the ones I've done so far, about 20 booklets so far.

Robert Walker's kindle books on

Here are the ones related to this topic of terraforming
Trouble with Terraforming Mars (Amazon) - equivalent to 43 printed pages

Terraforming Mars - far into Realms of Magical Thinking (Amazon) - equivalent to 19 printed pages

Imagined Colours of Future Mars: Should We Treat a Planet as a Giant Petri Dish? - equivalent to 29 printed pages

Also you might like:

Ten Reasons NOT to Live on Mars - Great Place to Explore - equivalent to 33 printed pages


Asteroid Resources Could Create Space Habs For Trillions: Land area of a thousand Earths - equivalent to 36 printed pages

and if you are interested to find out more about the impact lakes, and other habitats, with many cites to the scholarly literature if you want to go back to the sources:

Also see my

For my Science20 articles here

The only difference is that the books have a table of contents, cover image and title page, and are formatted for kindle. The text and images and everything else is the same.