This was an ask stoat question, and probably a fairly easy one, so I’ll have a go.
First of all, what is it? AF (ie, Airbo(u)rne Fraction, is the proportion of human emitted CO2 that stays in the atmosphere, the rest being sunk in land or ocean. Now it is important not to confuse the “proportion that stays in the atmosphere” with “the concentration in the atmosphere” otherwise you get silly little skeptics running around thinking that “airbourne fraction is constant” means that CO2 has stopped increasing. Sigh. However, I see that last time I looked at this I was having to slap down the other side who seemed to think that AF was rapidly rising to 100%. Sigh #2.
Having poked around a bit after the recent Knorr paper, I find that skeptical science seems to have done a pretty good job on the story in general, so I don’t think there is any need to explain that side much more (or, you can have mt and Eli). But since I’ve started this post I’d better find something to say. And that is… two things:
First off, there does seem to be some “pressure” to find increasing AF. That it should increase doesn’t seem to be a strong scientific prediction, but it wouldn’t be very surprising given increasing human emissions and possible degredation of sinks. And if it was increasing, that make future CO2 predictions more exciting. Which brings me on to the second point:
As someone commented, AF itself isn’t a very physical variable. The real physical variables are the sources and the sinks. What we can measure fairly easily are sources (fossil fuel use well; deforestation harder) and the concentration. From those you can compute the sinks and the AF. Or you can try to estimate the sinks directly but that is imprecise. But now suppose you need to predict *future* CO2 levels (and if you want to project future climate change, you do need to). You can run your economic models forwards and deduce emissions, but modelling the sinks is hard (you can stuff them all into a coupled GCM with carbon model, and the Hadley folk at least can do that, but I’m not sure how accurate it is thought to be). So it is an awful lot easier just to use the constant scaling factor of AF to deduce future CO2 levels. And so you have this “tuning knob” and what value should it be? Since AF ~ 0.5, I think most people use a half, which seems about fair – you don’t know its future value accurately, but then you’re guessing at the emissions too, so it all washes out togther. But of course, if you knew that AF was going up, you could get to higher CO2 levels earlier. That would be bad, wouldn’t it?
[Update: per L in the comment "Some folks talk about GHGs running away a bit, what with all the forest fires and permafrost melting -- this would also show up as an increased AF, wouldn't it". More easily, it would show up as yet another thing you can measure, the year-by-year change in CO2 levels. Which was what the earlier post, and its graph, was about. Which shows that nothing too wildly exciting is occurring.
ps: we're up to comment ~9,800. Nearly at the majic 10k, for which there will be a Prize! -W]