How much sunlight would need to be blocked to cancel out atmospheric CO2?

Question

Atmospheric CO₂ is trapping too much heat from our star's radiation. How much light would need to be blocked/reflected before reaching Earth to cancel out the excess heat being trapped by atmospheric CO₂?

Answer 1

This may belong to another SE, but based on this, a 2% increase in albedo would roughly offset a doubling of CO2: different models give different numbers and this figure is relatively old in terms of model development & hence quality I suspect. However it will be in the right ballpark.

Note that you can't fully compensate for CO2: you still get ocean acidification for instance, and you also will get globally weird weather, at least in the engineeringly-plausible approach of dumping a lot of stuff into the stratosphere. In fact even the 'big mirrors in space' solutions, which are obvious fantasy technically, still do not compensate: the atmosphere with a high CO2 content is just different to that with a lower one. The best you can do is keep surface temperature under some kind of control.

Surprisingly the stratospheric-aerosol approach is not stupidly expensive: it's expensive but feasible: I think you might need a fleet of 100 planes or something. You do need to keep doing it though: it's not a once-off fix.

The right search terms to look for are 'Solar Geoengineering' or 'Solar Radiation Management'.

This has traditionally been a slightly taboo area for two reasons.

• It leads people to think that no other fix is needed: we can just keep pumping out CO2 and fix it by pumping sulphuric acid into the stratosphere. We can't do that in fact for a bunch of reasons (most of which I don't know: I'm not boing coy).

• It might be weaponizable: if you can mess with the climate over some part of the world where people you don't like live you can perhaps engineer harvest failures & the resulting starvation.

I think the current evidence (based on modelling how stratospheric aerosols behave) is that it's not weaponizable as you can't localise the effects, and I personally think it might be a reasonable approach to deal with the likely 'overshoot' scenario where CO2 emissions get brought under control but there's a very nasty 50-year window before the system responds.