You have to

• Increase its atmosphere to nearly three times the Earth’s atmospheric mass per square meter (10*9.807/3.711) or about 26 tons per square meter. That’s because with its low gravity, Mars needs much more mass per square meter to achieve the same atmospheric pressure. Mars almost certainly doesn’t have enough carbon dioxide for that.
• Remove nearly all the carbon from the atmosphere leaving oxygen because carbon dioxide at 1% or above is poisonous to humans. If you do this using photosynthesis then it will take about 100,000 years even if you have as much vegetation as Earth, because the light levels are half those on Earth and because there is so much more atmosphere you have to convert to get decent oxygen levels. Oxygen is added to Earth’s atmosphere and removed slowly and even if we lost all our vegetation, the levels wouldn’t change noticeably for thousands of years. You have to remove vast amounts of carbon, enough to cover the surface to depths of meters. Remember there are 26 tons per square meter of atmosphere. Of that, 23.2% by mass needs to be oxygen or about 6 tons per square meter. So the amount of carbon to be removed, remember there is one carbon atom paired to two oxygen, oxygen atomic number 8, carbon atomic number 12, So you need to remove 6*12/16 = around 4.5 tons of carbon from the atmosphere per square meter in organic compounds with hydrogen and other atoms in the form of wood, peat, soil or whatever. Clearly at the end of that process the planet will be covered in several meters thickness of carbon based organics extracted from the atmosphere by photosynthesis. It’s not surprising that Chris McKay estimated this would take 100,000 years.
• Make sure the atmosphere has about 20 tons per square meter of nitrogen - we need nitrogen as a buffer gas and again to achieve Earth like pressures you’d need much more than for Earth for the same pressure. Mars has some nitrates, and we don’t know how much. It might have vast quantities of nitrates sequestered below the surface from its early atmosphere. But you’d need some way to split those nitrates and return the nitrogen to its atmosphere, if it has enough of them. You could use another inert gas in place of nitrogen, but there’s no obvious candidate that’s abundant enough on Mars to take its place, and nitrogen is best for nitrogen fixation - you probably want a fair bit of it anyway.
• Warm it up in some artificial way - because Earth’s atmosphere would not be a warm enough “blanket” at the distance of Mars. It would be too cold for trees even at the tropics. There are two main proposals here. The first, most favoured, is to add powerful greenhouse gases, which involves mining cubic kilometers of fluorite ore on Mars per century and building 500 full scale nuclear power stations on Mars to provide the electricity needed to make the greenhouse gases. The other option is to build thin film mirrors in orbit around Mars which may sound easier until you realize that the thin film mirrors need to have a total area of a thin film mirror about the size of the planet itself. The greenhouse gases approach is thought to be the “easier solution”.
• You need to do something to prevent solar storms from stripping its atmosphere, as has happened in the past. This is not an immediate problem, but would strip away its atmosphere on timescales of millions of years. Or maybe you keep replenishing its atmosphere - it probably needs to be maintained in some technological way for the foreseeable future anyway as the green house gas production would have to continue endlessly replenishing the gases as they are lost or it will cool down to much to be habitable.
• You need to add vast amounts of water somehow, perhaps from comets, because though it has some ice, remember that its ice caps are much smaller than Earth’s ice caps with total ice far less than Antarctica, and no oceans and the tropical areas dry to considerable depth. So melt the ice and any in the tropics will first just drain into the desert or migrate to the poles to form ice caps, so it’s not clear you’d have any seas or much standing water unless you add lots more water to it somehow.
• Make sure that meanwhile you don’t also add vast amounts of sulfur dioxide or hydrogen sulfide or methane or set up some kind of biological cycle that will add those gases to the atmosphere - at least if you want it habitable for humans.
• Then you need to make sure it has an ozone layer to protect from UV radiation. It’s got no magnetic field to protect the upper atmosphere from solar storms which may be a problem as solar storms will damage the ozone layer.
• You have to set up the various cycles to return carbon dioxide to the atmosphere, to cycle nitrogen etc. These have to work nearly three times more efficiently than on Earth with half of the light levels to have the same effect, because of the reduced gravity. They also have to do that with less than half the light levels (unless you have large thin film mirrors directing extra sunlight to the planet). So for an exact duplicate of Earth you’d need the cycles to work nearly six times more efficiently than they do on Earth.
• Problem returning carbon dioxide to the atmosphere after it is sequestered in chalk, limestone etc. There is no continental drift so volcanic activity is minimal, it’s not totally inactive on geological timescales but none known at present (by contrast Earth has numerous active volcanoes). This means there is no way to return carbon dioxide to the atmosphere once it gets sequestered as limestone, chalk etc into its lakes and seas. So we need another way to return the carbon dioxide to the atmosphere. There are microbes that eat chalk and limestone and turn it back into CO2 so you’d need to set up some new cycle based on those microbes, and it’s not at all clear how easy that would be to do. This is for the long term slow carbon dioxide cycle. So - it’s not a problem on the thousands of years timescale - but presumably you want to set up a planet that’s going to work indefinitely into the future, if so it is a problem.
• We don’t know what effect the reduced gravity would have on animals and humans, for instance can they still give birth to healthy offspring and can they grow up normally?
• And - on top of all that - how do you make sure that the terraforming planet ends up in the desired end state? What if it starts to evolve towards a hydrogen sulfide, or methane rich or sulfur dioxide rich atmosphere? What if the amount of carbon dioxide increases no matter how much you try to get rid of it? What if it evolves microbes and higher creatures that eat the oxygen producing algae and plants before they can convert the carbon dioxide into oxygen? What if it evolves pathogens that are harmful to humans and animals, or our plants?

On the oxygen levels - you could try to make a thinner breathable atmosphere with only 10% oxygen, pure oxygen without the nitrogen, which is simpler but is a fire risk. You can deal with that in spacecraft and spacesuits by making sure there is nothing flammable but how do you make an entire planet non flammable? The thing is that the nitrogen actually acts as a fire retardant by absorbing the heat from the flames, and if you have only oxygen, its no more flammable, but you’ve left out the flame retardant in the atmosphere, so fires start more easily. So that doesn’t seem practical and I haven’t seen anyone suggest we attempt terraforming with a nearly pure oxygen Mars atmosphere.

The Mars society very optimistically suggest that we could get to a planet warm enough for trees with a carbon dioxide atmosphere in 1000 years - and then build up an oxygen atmosphere after that. At that point humans could get around unprotected with air breathers but would need scuba diving style gear with their tanks filled with full atmosphere, not just oxygen, because the carbon dioxide would kill you eventually if you just used an oxygen mask. And there would be no animals or birds at that point (unless you somehow equipped them with scuba gear too!).

It’s a vast mega engineering prospect that has to be sustained for probably at least 100,000 years before it’s habitable at levels similar to Earth, and then sustained indefinitely with hundreds of power stations producing greenhouse gases constantly to keep the planet warm - or planet scale thin film mirrors to warm it up, with comets or other sources of volatiles used to replenish the atmosphere over millions of years.

I can’t see anyone doing this myself, when you think how much you could benefit Earth with a similar amount of mega artchitecture. Certainly at present, I don’t see how anyone could commit to such a project for themselves and their descendants for a thousand years, so over 30 generations. The Mars colonists would not be able to go it alone,, can’t see it working without support from Earth.

Of course that’s based on present day technology. Who knows, with nuclear fusion maybe small scale nuclear fusion, 3D printers that can replicate almost anything, with a magical level of technology as it would seem to us today, it might be possible. But that is so different from what we have today and we can’t even imagine such a future clearly, any attempts are bound to be as far from the mark as early C20 projections of the twentyfirst century. So I don’t think we should colonize Mars setting out on the basis that we know what technology we’ll have even 50 years from now and that we know what our priorities are.

The obvious thing to do instead is paraterraforming. That is, you cover regions with domes to build enclosed cities. But that’s as easily done on the Moon, indeed the lunar caves may form pre-fabricated vast tunnels kilometers in diameter and over hundred kilometers long and it may well have abundant ice and other volatiles at its poles. As well as that, then enthusiasts like Paul Spudis have worked out ways that Lunar colonization could be commercially viable. The bulk of books by Moon colonization advocates consist of discussions of how it can be made commercially viable. While Mars colonization advocates use a very broad brush treatment, saying that somehow it will work, but when pressed for details they come up with unconvincing ideas. Mainly they focus on the idea that the colony would pay for itself by export of intellectual property to Earth. Given that they would have the highest technology colony ever and be dependent on Earth for multi million dollar spacesuits, hundreds of millions of dollars habitats, each needing to be replaced frequently, environment control etc - I can’t see how they would have export of intellectual property to Earth. It would surely be the other way - that they rely on intellectual property exported from Earth for nearly everything they do.

Then - there’s the planetary protection issue. Mars may have present day life there, and may have past life. Even if it has no life, it may have habitats and be the only place in our solar system to find out what a planet like Mars is like whether it has inhabited or uninhabited habitats. If we introduce Earth life there, we may lose this opportunity to make major discoveries about what could easily be one of the more common types of planet in our galaxy. There is no other Mars we can explore if we mess up our Mars in this solar system. And we might if we are lucky find some form of life that’s evolved independently or some very early form of life that is long extinct on Earth.

There are no issues like that with the Moon.

Chris McKay has suggested we could “Marsform Mars” (my own word for what he describes) - turn the clock back on Mars to recreate the conditions that may have existed there billions of years ago. That’s much more feasible, since after all Mars was like that originally. But it’s probably lost a lot of its volatiles since then so it would still involve finding those volatiles somehow. It only seems to have enough CO2 to more or less double its current very thin atmosphere. Still if we found a way to do that, it might be of interest as it would make it marginally more habitable to any native Mars life.

That life might be of great interest to us, it might teach us much about biology, biochemistry, medicine. Not at all guaranteed that we’d be able to cultivate it in vitro, there are many microbes that exist in the wild on Earth that we have not yet, to this day, been able to cultivate in the laboratory. Who knows, it might also have products of great value to Earth, products of the Mars biology.

At any rate, I think we have to study Mars first, before we think of attempting to introduce Earth life there, and indeed also before we attempt to “Marsform it” if that was the decision.

And if we do try to terraform in some way - well - clumsy attempts at introducing Earth life right now could set Mars off in a direction of change that we would later wish to reverse and find we can’t do . It might have many possible end states. How can we hope to direct an entire planet to a desired end state with present day technology? We find it hard enough to so something about carbon dioxide levels at a fraction of a percent on Earth so how could we hope to keep the Mars atmosphere on track if it deviates in some unexpected direction? And even more so, how could we make sure the biology and evolution there goes in a direction that is to our liking? Remember Mars conditions are very different from Earth and microbes can evolve rapidly. Maybe we can learn enough to do this from our experiments with closed system habitats on the Moon or elsewhere, or maybe we make contact with extra terrestrials who advice us from their own experience of attempting something like this.

Elon Musk brushed over all this with CGI showing a planet revolving and turning green as it rotated behind him - but that’s brushing a lot of details under the carpet. I think few of those who are so keen on colonizing Mars have realized quite how different it is from Earth and how much would be involved in terraforming it.

Note also that on Earth also it took hundreds of millions of years to develop to the modern atmosphere. So even if we had an Earth clone, same size, maybe with an early Earth atmosphere not yet habitable to modern Earth life it would be a huge acceleration to make it habitable in as short a time period as 100,000 years.

We have no practical experience of terraforming, and I’m not sure it would even be possible so quickly with an early Earth clone, never mind Mars, and the idea that we could direct it to a desired end state seems ambitious in the extreme. But for Mars the situation is far harder than for an early Earth clone because it is so much further from the Sun, has no magnetic field and the gravity is much less.

We don’t need Mars as a “backup”. Because there is nothing that could make Earth as uninhabitable as Mars would be even after a thousand years of terraforming. It’s clear I think that Earth is where we make our stand as the only easily habitable planet in our solar system for our form of life.

I think we need to move forward with a positive vision for the whole of Earth, not an escape plan to try to get away from it. Humans in space can protect Earth from asteroids, mine rare minerals and metals that are getting in short supply here, move polluting industry into space, and make discoveries that could expand our understanding of our biology and of the universe hugely. We can explore, have destinations for explorers and tourists, there are many things we can do in space. But our best backup is on Earth. The Moon could be useful as a backup of knowledge and a seed library because of the very stable geological conditions there. But for living humans, Earth is the best place for our backup. Earth is its own backup. There’s no disaster, not an asteroid impact, not supernovae, not gamma ray bursts, that could make Earth uninhabitable to humans not in the present day solar system, though there were billions of years ago. Even if 97% of species went extinct, then Earth would still be far more habitable than Mars, amongst the 3% there would be plenty for a versatile tool using omnivore like ourselves to cultivate and life on. We’d surely be amongst that 3%.

But there are no nearby stars that can go supernova (we now know) and we also know all the 10 km or larger asteroids that do regular flybys of Earth and none of them will hit us for several centuries - and only 1 in 147 of flybys at present are by comets, and the only nearby candidate for a gamma ray burst is pointed away from us.

It’s true that 500 million years from now or later, Earth may begin to become uninhabitable to humans. But that’s so far into the future that humans could evolve from the very fist multicellular microscopic life forms a second time. And anyway a temporarily terraformed Mars, even if it worked would probably not last that long, before it needs to be replenished again by comets etc. Maybe Mars will be just what our descendants or whatever species follows us needs when the Sun goes red giant billions of years from now. But it’s far too soon to try terraforming it for that distant species, which might not even be oxygen breathing, and would find our attempt at a terraformed Mars with its atmosphere long gone, all its volatiles lost to space, and harder to terraform than it is for us now. They might also be able to move Earth to a more distant orbit from the Sun or protect it with sun shields, or make asteroids and even planets into a Dyson sphere or who knows what else in that distant future.

Meanwhile if we can develop self sustaining closed ecosystems in space, there’s enough material in the asteroid belt to build space habitats enough to house trillions of people, total surface area (not of the asteroids but of the habitats that could be made of them) a thousand times that of Earth, and far more if we also include materials beyond the orbit of Jupiter, or ideas of using materials from moons and our Moon to make into habitats.

See also my

• Imagined Colours Of Future Mars - What Happens If We Treat A Planet As A Giant Petri Dish?
• Trouble with terraforming Mars
• OK to Touch Mars? Europa? Enceladus? Or a Tale of Missteps?
• Case For Moon First
• Why Humans on Mars Right Now are Bad for Science. Includes: Astronaut gardener on the Moon
• Asteroid Resources Could Create Space Habs For Trillions; Land Area Of A Thousand Earths