Atmospheres and Terraforming

Retroactively added to my series on countering misconceptions in space journalism.

I have been meaning to write a blog about terraforming for many years, but the recent controversy (thanks Elon) about some exciting MAVEN results is the perfect opportunity. MAVEN, or Mars Atmosphere and Volatile EvolutioN, is a satellite orbiting Mars since 2014 specifically to study its atmosphere. Previous Mars missions have studied parts of its surface, and the InSight mission due to land in late November will study, for the first time, the interior of the Red Planet.

Mars’ geological history is an enduring mystery. Much of the surface is heavily cratered, like our moon, meaning that we have a pretty good record of what ancient rocks were doing, because they’re still there. In contrast, Earth’s surface has been worn down and subducted many times through the process of plate tectonics. Yet if we look closely, there are clear signs of flowing and standing water. This is puzzling in many ways. Where did the water come from? Where did it go? Why did it stop flowing? And, given that back when this surface formed the sun was much younger and cooler, how could Mars have been warm enough to sustain liquid water on its surface?

This image shows gullies and deltas in the south west part of Gale Crater, not far from where the Curiosity rover is currently driving.

MAVEN was designed to study the evolution of Mars’ atmosphere, and a series of results have provided exciting new insights. It has long been theorized that Mars’ atmosphere was gradually stripped away by the solar wind. Unlike Earth, Mars lacks a strong gravity well and a big magnetic field, which increases the rate that gases, particularly lighter gases, are stripped from the planet. This accounts for why Mars’ atmosphere is today both thin and mostly of CO2. But since gases are constantly bubbling up from the interior of the planet through volcanoes, we needed to get quantitative data. MAVEN provided that data. Today, Mars loses about 100 g/s, or 3000 T/year, of atmosphere. This sounds like a lot, but the total mass of Earth’s atmosphere, for comparison, is 5*10^15 T. To put 3000 T/year into perspective, a large passenger jet produces about that much CO2 in a week.

It is a popular misconception that for humans to live on Mars, Mars would need a magnetic field. There are all kinds of difficult things about living on Mars (see but building a planetary scale magnetic field would be at another level entirely. Fortunately, Mars does not need humans to built it an artificial magnetic field for any reason! As we will soon see, atmospheric loss rates are vanishingly tiny over the relevant time scales of less than a billion years, and Mars’ current atmosphere provides adequate, if not flawless, protection against space radiation for nearly all cases.

So what’s all the fuss about?

Noted Mars human exploration advocate Robert Zubrin, together with planetary scientist Chris McKay, wrote a series of papers in the early 1990s exploring the idea of terraforming Mars (see eg. Taking Mars Viking data and using rather primitive simulations, they were able to show that, under some reasonable assumptions, there was enough CO2 frozen into the surface of Mars that only a tiny addition of heat, by humans, would be enough to tip the climate away from its current frozen state to a runaway greenhouse. As the planet warmed, it would release more gases, warming further! Eventually, this process would be slowed by the lack of remaining accessible gases. If the atmosphere became warm enough, water would melt and restart the hydrological cycle, creating a much more Earth-like environment, albeit with a poisonous atmosphere of mostly CO2. Still, having to wear an oxygen mask to breath is much less onerous than an entire space suit.

Potential problems included uncertainties about the stability of a hydrological cycle on a planet with such tall mountains, as water would snow there and accumulate in glaciers, removing liquid water from the system and reflecting more of the sun’s heat into space. Additionally, the total reserves of CO2 in the polar caps and under the soil was poorly constrained.

Last week, a Nature paper ( was published by Jakosky and Edwards arguing that, based on data from more recent orbiters, the total near-surface reserves of CO2 are much too low to build up the atmosphere enough for terraforming. What a bummer!

I’m not sure if I find this result surprising. I would think that if Mars were, according to Zubrin and McKay, only a few degrees away from runaway warming, that a sufficiently large meteorite impact or volcanic eruption in the last few million years might have been enough to tip the scales. And perhaps, in the past, it was. There is some evidence for smaller scale water flows in the more recent past, but whatever happened and when, the planet is cold and dry now. According to the principle that our present time is not “special” in any way, that suggests that Mars reasonably robustly and quickly returns to something like its present state. How much meddling is necessary to terraform the planet is another question entirely!

Of course, the usual players got into the act on Twitter:

Ah Twitter, where you can watch your heroes disgrace themselves in real time!

Clearly the only way to fix this impasse is for me to write a blog no-one will read adding almost no substantive information to the deafening screams. But with math!

The Earth’s atmosphere is thickest at the surface, and gets thinner as one ascends until it peters out into space. At the surface, we experience air pressure of 101.3 kPa, or 1 bar. I prefer bar as a unit of measure in this case!

Even though gas is roughly 1000x less dense than condensed matter, it is still affected by gravity. Indeed, the reason that the air is thicker and at higher pressure down low is that all the air above it is pressing down on it. So even if we somehow liquefied all the air (by making the Earth really cold) the pressure at sea level would stay the same. In other words, a phase change doesn’t change gravity! Only now the Earth would be covered not by a gaseous atmosphere, but by a thin ocean of mostly liquid nitrogen and oxygen.

I personally really like this way of thinking about global gas and liquid resources. That is, in terms of a global equivalent liquid layer of some depth.

Let’s start with our home planet, Earth. We’re going to ignore sources of volatile molecules beneath the crust, though there’s a huge amount there too, and focus on surface resources. Earth’s global equivalent depth of water is 2.6 km. The next most abundant molecule is nitrogen, with a global equivalent layer depth of 8 m. Then oxygen, with 2 m. 10 cm argon, 4 cm water, 4 mm carbon dioxide (up from 2.5 mm 5000 years ago), 0.2 mm neon, and a hair each of helium, methane, krypton, and hydrogen. This is a total depth of 10 m, which will be familiar to SCUBA divers computing pressure at depth.

By comparison, Venus has an atmosphere that, if liquefied, would be more than 600 m deep.

For humans to breathe, they need an atmosphere with oxygen, some buffer gases, and not too much of poisonous gases, such as CO2 and CO, among many many others. But how much oxygen do humans need? The partial pressure of oxygen in haemoglobin, the blood’s oxygen carrying molecule, is 130 mbar. This means that at a partial pressure of oxygen below this, the blood is not saturated with oxygen, and less physical activity is possible. For healthy, adapted, young people it’s possible to walk around and live at 5500 m of altitude (ppO2 = 100 mbar), but it’s pretty hard to reproduce over 4000 m (ppO2 = 128 mbar).

There are various tricks people use to exist higher up, such as breathing pure oxygen with positive pressure, but the same general rules apply. Lots of mountain climbers with great gear die all the time. Ideally, an atmosphere with 200 mbar is enough to allow physically active humans to breathe from an unsealed oxygen mask, but without a pressure suit. At the absolute lower limit, the Armstrong Limit of 68 mbar is the vapor pressure of water at human body temperature ( Below this, the body’s surface fluids boil and adequate oxygen cannot be delivered by any means without a pressurized garment.

On Earth, 200 mbar is achieved with a depth of 2 m of liquid water or liquids of similar density. On Mars, the gravity is 37% as strong, so a 2.7x greater depth of fluid is required. Condensed CO2 is about 1.5x as dense as liquid water, so taking these numbers into account, a layer of dry ice 3-4 m thick on Mars is adequate to produce an atmosphere with oxygen mask enabling pressure.

Is there 4 m of CO2 mixed into Mars dirt within, say, 200 m of the surface?

The next step is to examine the known resources on Mars. Mars’ atmosphere is so thin it’s about 1/100th as thick as Earth’s. That is, if it were condensed, it would be about 10 cm thick on the surface. In terms of its composition, that tiny crust is 9.6 cm CO2 (+- 2 cm seasonally, as some freezes out in winter), 2 mm argon, 2 mm nitrogen, 0.15 mm oxygen, and less than a hair of carbon monoxide.

Despite the fact that Mars’ atmosphere is almost totally CO2, it is thought that Mars’ volcanic processes have produced significantly less CO2 over geological history than Earth, because Mars’ small size and resulting reduced pressure in the mantle is unable to produce much CO2 from various metamorphic transitions at depth.

So ideally a runaway greenhouse terraforming first effort on Mars would increase the volatilized supply of CO2 from its current level of about 10 cm global equivalent depth to 4 m, an increase factor of 40.

At this point I would like to point out why it is that losses due to solar wind don’t matter. If 4 m of CO2 is produced in, say, 100 years, then that’s a rate of 4 cm/year. Even if global losses increased from 3000 T to 300,000 T/year due to an increase in cross section, the annual loss rate is an equivalent depth of 2 nm. It’s in the noise.

Zubrin and McKay (1993) attempted to estimate reserves of CO2 available. The two main sources they considered were the south polar cap and frozen into the regolith, with 100 mbar and 400 mbar of CO2 respectively. This is equivalent to 1.8 m and 7.2 m global equivalent depth respectively, or easily enough to get to a warm, wet, oxygen-mask requiring atmosphere, even if only half the CO2 was readily accessible.

25 years later, we know a lot more about Mars, so let’s check how Jakosky and Edwards have refined these early estimates. They estimate a total resource of 10 cm of CO2 at the south pole (which turned out to be mainly water ice), and 1.8 m absorbed on minerals in the regolith. They also provide an upper limit on total CO2 in clathrates (150 mbar, 2.7 m) and near-surface carbonate rocks (150 mbar, 2.7 m), despite admitting that there is little evidence for either, especially anywhere near that much of them! In total, there doesn’t seem to be enough CO2 remaining on Mars to warm the atmosphere.

In response, Robert Zubrin tweeted:

Both Zubrin and Jakosky estimate 1% soil mass fraction of absorbed CO2 by weight, though Zubrin goes down to 200 m instead of 100 m. Thermal diffusion goes as the square root of time, so doubling the depth will require four times the time, at the same thermal gradient. Reducing his estimate to 100 m, 150 mbar of atmosphere is still a lot more than Jakosky’s estimate of 40 mbar. What’s going on here?

I don’t know for sure how they did their math, but it seems that Jakosky has assumed a regolith density equivalent to water, whereas rock is typically 3x as dense as water. Zubrin’s estimate would call for a global equivalent layer of 4 m of CO2 in 200 m of regolith, since condensed CO2 is about 2x less dense than rock. 4 m of CO2 under Mars’ gravity produces a pressure of 220 mbar, which while 25% less than Zubrin’s estimate, would be adequate to walk around in.

That said, 1% w/w is as much of a guess as 200 m effective absorption depth. Even on Earth we don’t have a good understanding of the water fraction of permafrost, or its thermal conductivity, or extent, and permafrost is a reasonably good analog for Mars perma-CO2-frost.

As far as I can tell, there just might be enough CO2 on Mars to begin a greenhouse warming effect, but even in Zubrin and McKay, they recognize that uncertainties in regolith affinity for desorbed CO2 could shift their equilibrium a lot. In other words, it really doesn’t hurt to tip the scales solidly in the desired direction. We know this is possible, since with a 1 mm addition of CO2 to Earth’s atmosphere we’ve warmed the planet considerably!

A wide variety of terraforming methods have been proposed over the years, some less practical than others. The overarching goal is to increase the total amount of heat energy on the planet. The baseline heat source is the sun, which delivers about 10^16 W (10 petawatts!) to the surface of Mars.

In order from least sensible to most sensible:

Nuking the poles

The Earth has about 15,000 warheads with a total yield of 6500 MT of TNT, which is 2.7×10^19 J. This sounds like a lot, but is equivalent to only 45 minutes of sunlight. In principle, hydrogen bombs can be made arbitrarily large, but in practice even launching the largest possible warheads from Earth would barely shift the scales. Even if the bombs were carefully landed and detonated under the ice caps, instead of exploded while flying by, the net contribution to the heat budget is negligible. Even if all the uranium in the crust of Earth AND Mars were enriched and allowed to produce heat in reactors, it wouldn’t be enough to make a difference. The sun is THAT powerful, which is why solar energy is so exciting.

Crashing comets

Mars is pretty dry, though the exact quantities of water left frozen under the surface is not very well constrained. It could be between 20 cm and 200 m global equivalent layer. Even if it turns out there’s no water or gas left, the thinking goes, comets or small moons could be brought in from the outer solar system and crashed into the planet to thicken the atmosphere and warm things up a bit. Unfortunately, the sheer quantity required to do this exceeds known reserves of comets! Not to mention the fact that landing enough of them to make a difference would completely resurface the entire planet and kick up enough dust to ruin things for centuries. Finally, the energy and time required to move icy bodies from the outer solar system inward is quite prohibitive, compared to the requirements of other methods listed below. Even a dry Mars has much more water than all the comets combined.

Giant mirrors

The fundamental advantage of mirrors is that they can be made extremely thin, and thus with relatively small quantities of material. Additionally, a mirror can be made in a curve to concentrate sunlight on a particular area, which is necessary to either melt down to subsurface resources or vaporize carbonate rocks. A series of mirrors 100 km on a side could be made with as little as 40 T of material, so in principle launchable on a Falcon Heavy then flown using solar wind pressure to Mars, before finding a stable equilibrium between Mars gravity and solar pressure, from where to gently barbecue the planet. Alternatively, one could mass produce billions of small, independently operated solar sails using cell phone technology.

The last two options are concerned less with adding more heat to mars, but preventing heat from escaping once it arrives. They require interventions on the surface.

Black dirt or lichen

The amount of light reflected by an object in space is the albedo, and varies between 0 (perfectly black) and 1 (perfectly reflective). Mars’ albedo is 0.15, meaning that 85% of the sun’s visible light that falls on it is absorbed. If we could decrease the albedo a bit more, the planet will absorb more heat energy from the sun, and warm up. Mars’ albedo increases during the southern winter, as a thin layer of CO2 freezes out of the atmosphere, covering large areas with reflective white snow. The best way to reduce this effect is to blacken the snow, such as occurs in the ski fields of Europe after an Icelandic eruption. Mars is already quite dusty, but the distribution of black dust over the surface would increase the absorption of solar energy. Even better, dark colored lichen, genetically engineered to survive on the harsh surface, would propagate itself over the planet “for free”, without humans having to go out and paint 145 million square km of landscape.

Greenhouse gases

As far as CO2 goes, it’s a pretty good greenhouse gas, blocking emitted thermal radiation across a wide range of wavelengths. It does, however, have some holes where it is transparent and allows heat to escape through the atmosphere and back into space. This blog ( has a great explainer on the matter, while (Wordsworth et al. 2017 discusses the implications for the early Mars climate.

Human-produced greenhouse gases can be selected to block the gaps in the CO2 absorption spectrum. The best gases for this purpose are perfluorocarbons, or PFCs. On Earth they’re generally used in relatively small quantities in chip manufacture, medicine, and the production of Teflon. But on Mars, provided adequate supplies of fluorite mineral from which to derive fluorine were located, rock-eating PFC factories would dump gases like CF4, C2F6, and C3F8 into the atmosphere as fast as they could be built.

(Marinova et al. 2005 estimates that the addition of 0.2 Pa of the best gas mixture is adequate to trigger runaway warming. 0.2 Pa is a global equivalent layer of 6.6 microns, but since they’d have to be produced at discrete locations, a consideration of the total mass required is in order. 0.2 Pa is equivalent to a total production of 7.8×10^9 kg, or 0.958 km^3 (in the condensed state), of which about 80% would be extracted from fluorite, and 20% from CO2 in the air. This sounds like a lot, but in 2017 we managed to extract 5.4 km^3 of oil from the Earth. If fluorite were discovered in dry lake beds at 10% ore concentration, then an area only 10 km x 10 km x 100 m would need to be excavated, comparable to the largest open cut mines in the world today. Of course, this process would occur not at one huge hole but numerous sites selected for natural resource abundance, and over many decades.

Production would, of course, require massive automated digging machines powered by nuclear power plants, but the net heat retention for the planet per joule of uranium used would be millions of times greater than detonating it in an atomic bomb. Once PFC levels were high enough, runaway warming using frozen CO2 would be triggered, eventually resulting in a warm, wet, though poisonous atmosphere.

Finally, let’s examine long term prospects for Mars

Using variants of the above processes, Mars could likely be given a warm, wet atmosphere in a small number of centuries. But even if humans can walk around outside wearing only an oxygen mask, it’s not quite terraforming if plants and animals can’t easily exist. To do this, a large fraction of the CO2 atmosphere will need to be converted to oxygen. For humans, a global equivalent layer of about 3 m of oxygen is needed. This is more oxygen, on a per area basis, than Earth’s atmosphere, because of Mars’ lower gravity. If Mars’ atmosphere has stabilized with, say, only 6 m equivalent depth of CO2, converting half of this to oxygen will severely damage the planet’s greenhouse. More worrying, most plants and animals cannot tolerate high levels of CO2, no matter how much oxygen is available.

In the Mars Trilogy, author Kim Stanley Robinson elides this difficulty by proposing that organisms on Mars wear a CO2 filter mask, or get a genetic modification to increase their CO2 tolerance, as some diving animals like crocodiles can on Earth. Such a future is so far away at this point that I can offer little better than science fiction myself! Perhaps ongoing production of PFCs, plus orbital mirror blasting of carbonate and nitrate rocks will produce enough atmosphere that an Earthlike existence will be possible. In any case, careful ongoing and energy intensive management will be necessary to restabilize the climate at any desired level.

I think it’s very exciting that our ongoing robotic missions to Mars have enabled such an interesting conversation, and that powerful dreams for the future continue to inspire hope in new generations. Sending humans to Mars is within our technological capabilities. Sustaining them on the surface indefinitely is possible. And, eventually, making Mars more like Earth is a worthy challenge.

6 thoughts on “Atmospheres and Terraforming

  1. “if we somehow liquefied all the air (by making the Earth really cold) the pressure at sea level would stay the same”
    Tiny nitpick. Only if the Earth was perfectly spherical in a vacuum.
    If you snapped your fingers and froze the earth, the atmosphere would rain out and coat the whole surface about 10m deep in liquid air. Then that liquid would flow down to the ocean and form a layer about 13m thick and provide a bit over 1 bar at the old ocean surface.
    Beyond that, the water in the ocean would expand as it turned to ice. So the old sea level would be under 10s-100s of metres of ice.


  2. You mean it can’t happen in seconds like in Total Recall? lol

    More seriously, will something like lichen survive at 1% atm composed mostly of CO2? It’s a tough plant, but it’s still a plant.

    Also, if we do happen to find microbial life on Mars, all these plans will be on hold until we study it and decide what’s more important there.

    Midterm, I think the best answer is putting a giant plastic sheet over a canyon and pressurizing the whole canyon (with anchor points, like your blog on a better alternative to the dome). You can fit a lot of people and hydroponic farms into an area a kilometer wide and tens of kilometers long, especially with structures 3 or 4 stories tall.

    Then a century from now, we can talk about how best to terraform the entire surface. What I like about what is presented here is that it is a lot less destructive than bombarding the surface with hundreds of comets – you could have people living there while the terraforming is going on.


  3. Imagining a huge constellation of independently adjustable mirrors being focused to vaporise surface ores, or with less annihilation, boost solar incidence around habitats… Love the blog btw, I almost always end up thinking about something I’d never considered before.


  4. I think your method of describing atmospheric mass in terms of equivalent fluid levels is helpful for understanding such large numbers. But it wasn’t clear if you are always talking about that depth taking into account the planet the fluid is on. So 10m on Earth is significantly more than 10m on Mars because of the differences in diameter of the planets. That undermines the explanation technique a bit since you can’t readily directly compare the two depths to understand the quantities being discussed.


  5. Around 5 million cubic kilometres of water ice have been identified on Mars. If it is possible to build TW scale fusion reactors in the future, we could use electrolysis to break this down into hydrogen and oxygen. The hydrogen would rapidly escape into space. The oxygen will accumulate in the atmosphere. That is engineering on a huge scale. But with huge scale comes huge scale economies.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s