Mars Society and Robert Zubrin estimate 1,000 years to establish large plant life followed by a few more millennia to get the oxygen levels up to Earth levels.

Chris McKay in one of his papers estimates 100,000 years for the complete process.

This is assuming a mega technology generating greenhouse gases etc - but basically it's done using biology. Could be speeded up a bit with giant mirrors in space.

If you look at the rate at which plants and algae can create oxygen then you can come up with a far more optimistic centuries long scale for making an oxygen rich atmosphere once you release the CO2.

But the problem with that is that it's seasonal and generally as much oxygen gets taken out of the atmosphere as gets added to it each year by plants.

So what you really need to look at is carbon capture. The reaction of photosynthesis splits water to create oxygen, but it also takes up CO2 to form an organic biproduct. So depends on taking CO2 out of the atmosphere permanently to work.

You need to see how much carbon is taken out of the atmosphere permanently, from the CO2, and then that tells you how much the O2 increases.

If you do the calculation that way - then look at typical rates of carbon deposits formation in the past - e.g. coal, peat, oil, oil shale, and look at the fastest formation rates, you end up with your tens or hundreds of thousands of years.

Or to put it another way - after creating a CO2 atmosphere on Mars similar but not as dense as Earth's - then you'd have to remove enough carbon from the atmosphere to cover the entire surface of Mars with peat or coal to some meters depth. You can see intuitively that that's not likely to happen in a century or two.

So, in practise most of the time involved would be for this last step which would take millennia almost certainly, to build up those huge deposits of wood or peat over the surface of Mars, and one hundred thousand years may well be about right (on the Earth of course it took millions of years).

Of course nanotechnology could change everything, this is assuming you use biology.

There are many other issues though with terraforming Mars. Mars is different from Earth in so many ways. No continental drift. Far less ice than you'd expect on a snowball Earth - it's cold enough of course so entire surface would be covered in ice if it had anything like as much water as Earth - no magnetic field to protect from solar storms. Closer to the asteroid belt so higher impact rate. Less gravity.

Obviously colder - even an exact duplicate of Earth - if you moved it out to the Mars orbit would become a snowball Earth right away so just for that reason obviously the climate cycles that work on Earth won't work unmodified on Mars.

Also the gravity at a third of Earth's means you need three times the mass of oxygen per square meter to get the same atmospheric partial pressure of oxygen - same for nitrogen and CO2.

One of the big differences for long term terraforming is that it has no continental drift which on Earth is what returns the CO2 to the atmosphere. So it's not at all clear how a terraformed Mars would stay terraformed on the very long time scales - you'd have to set up a way of microbes eating limestone as its formed, in a way that doesn't happen on the Earth - not impossible - but suggests beings with great understanding of biological cycles to be able to set up something like that that works long term without glitches.

At the moment I think pretty clear we have nowhere near the understanding to terraform even a second Earth that's exactly like ours but say in snowball phase. After all look at how much controversy there is over the effects of adding 0.01% of CO2 to the atmosphere - and how difficult it is to decide what to do or to deal with the issues. So how can we think about terraforming another planet yet, and especially one that needs a different solution from Earth?

At any rate we need to be careful not to contaminate Mars with Earth life until we have a chance to study it to see what's there. Then - if we do decide it's okay to colonize (which I'm not sure we will) and want to terraform it -then again we have to think carefully about the effect of introducing Earth life to the planet on our terraforming attempts.

For instance we might well want to start by using cyanobacteria to create oxygen. That's going to be more effective if there are no aerobes to eat up the oxygen and nothing on the planet to eat the cyanobacteria. If you landed humans on Mars first then both those things would exist and could eat the cyanobacteria and consume the oxygen - do both and result might be almost no effect. Also microbes introduced to the planet could change it's atmosphere in unpredictable ways -e.g. methane atmosphere or whatever, just depending on which happens to win out in the evolutionary changes that happen in them as they adapt to Mars.

And - what big mega technological project have humans ever done with modern technology that lasted a century and was done successfully? Never mind a thousand year project? What happens if it is abandoned half way - what happens to Mars?

Also Chris McKay has suggested that if Mars life turns out to be interestingly different from Earth life, then instead of terraforming Mars, we should carefully remove all traces of Earth Life from the planet- in the form of the rovers on the planet that we know have dormant Earth microbes on them - and then "Mars form" it (my name for what he suggests) so that it is best suited for Martian lifeforms.

We can afford to do things like that - nothing says we have to colonize Mars - as arguably free colonies using materials from the asteroids are easier to construct especially if you need or want to have full Earth gravity in a large spacious area surrounded by meters of cosmic radiation shielding - that's likely to be far easier to do with a free spinning habitat in space than to attempt a huge spinning habitat on the surface of a planet with gravity less than Earth.

And there's enough material in the asteroid belt for cosmic radiation shielding for living area 1000 times the surface area of the Earth.

And - that's all far future anyway - in near future then Earth is by far the most terrraformable planet in the solar system - obviously - and would remain so immediately after a giant asteroid impact or global nuclear war or even both combined.

Lots of questions. I think the ideas for terraforming Mars are great intellectual exercises and can help us understand Earth better. They may also be useful for the larger closed system habitats such as the Stanford Torus. Whether we ever use them for a planet - I don't know, but surely need more understanding than we have now - e.g. through more than a few decades of good data and models working fine on Earth - and through study of exoplanets.

See also my Trouble With Terraforming Mars

Also Asteroid Resources Could Create Space Habs For Trillions; Land Area Of A Thousand Earths

and other articles in my column.