The problem in a nutshell, is that we have discovered this new thing, and it might be a problem. We have to decide what, if anything, we want to do about it (partly because it's the kind of problem that's more difficult or even impossible to address later on, and partly because, as the Rush lyric goes, "if you choose not to decide, you still have made a choice". If we decide not to try and fix it, we should at least be doing that consciously, rather than by default.)
This is a hard problem, because we have to make it without a lot of information that we'd like. What I'm going to try and do here is to present a line of logic that, I think, suggests a sensible course of action given those contraints, and then
Let's start with what we know for sure:
Unarguably, the greenhouse effect exists, and CO2 is a greenhouse gas. Everyone agrees on these things, they're pretty basic physics.
Also, CO2 concentrations have increased by about 30% since the mid-19th century, and are continuing to increase. We know this with certainty because we can measure it directly.
Now the stuff people argue about:
We don't know precisely how much atmospheric CO2 contributes to the greenhouse effect. It's a non-zero amount, but people disagree on just how much that is. Still, it's pretty clear that it's not negligible.
We also don't know for sure whether the CO2 increases are directly due to human activity. It's the most obvious and straightforward explanation, but as some folks mentioned in the previous post, there are a whole bunch of carbon sources and sinks, the dynamics of which are poorly understood, and the overall atmospheric carbon flux is large compared to human emissions.
But consider this: since the beginning of the Industrial Revolution, we've been burning fossil fuels as an energy source. You can estimate how much, and it's an awful lot: the IPCC figures it's 290 gigatons of carbon. That's more than a quarter of a trillion tons. (We've also burned a lot of limestone to make cement, which puts off CO2 as well.) Of course, the natural fluxes are even bigger.
Now, does it seem likely that generating a lot of CO2 would increase the concentration of it in the atmosphere? Sure. How about the odds that it would have no effect, because there are other systems that compensate for it? Sure, that seems plausible, too. How about it generating a net decrease? Well, while it might be possible, we don't have any mechanisms that we expect to have that effect, and it would mean that there's both a negative feedback system over-compensating for human emissions and an unrelated system that's increasing CO2 faster than this hypothetical system is decreasing it. So that seems pretty unlikely.
So, despite all this uncertainty, we can conclude that the effect of human activity on atmospheric CO2 is that it's not unlikely that it's increasing it, and at best is having no effect. (And increasing CO2 will increase the greenhouse effect.)
Okay. Now, what about climate change? We know that climate is plenty variable on its own, and that some climate change is occurring independently of any effects of human activity. So whatever effects human activity have on climate are going to be added on top of that natural variability.
Are things getting hotter? Yes. Although this is a debated point, if you look at polar regions and alpine glaciers, I think it's pretty clear that a lot of things are melting that haven't melted in a very long time. I have a colleague who spends a lot of time with native Alaskans, and she says it's abundantly clear to them what's going on.
Fine. Does it matter? Again, this is a subject that's fraught with unknowns, but it's likely to be undesirable. There are some recent studies that project long-term losses will offset any short-term gains due to climate shift. Heating also gives you sea-level rise, which is bad. Any kind of major environmental change is disruptive simply because our infrastructure is adapted to things as they are, and changing it will take a lot of effort. And finally, climate change exacerbates ecological degradation, because it adds stress to already taxed systems.
What about the costs of doing something about it? There are a lot of arguments that various measures like the Kyoto Protocol would ruin the economy. I call bullshit. Economics is vastly more uncertain than climate science -- economists didn't even believe in positive feedback until the '80s. I'm not arguing in favor of the Kyoto Protocol, I'm just saying that "it'll be bad for the economy" is an indefensible argument against it. Major changes in how business is done can cripple an economy, but they can also lead to innovation, growth, and prosperity. There's no clear reason to believe that implementing greenhouse-gas (GHG) emissions policies would be either good or bad for the economy, so in this analysis we should consider it neutral.
Okay, so let's add it all up.
We know that CO2 causes greenhouse effects, CO2 is increasing, and warming is almost certainly occurring right now. Human activity is probably contributing to CO2 increase, and at best is having no effect. Warming will, in all likelihood, be inconvenient and undesirable, and whatever effects human activity has will be added on top of the already occurring natural climate variability, accelerating it. If we do something to mitigate GHG emissions, it will decrease the human contribution to these effects. We have no idea whether the long-term economic impacts of such mitigation would be positive, negative, or neutral, so we should call it a wash.
So in the best-case scenario, our activity isn't affecting the climate anyway, and mitigating emissions will have no effect. In the worst case, our activity is having a big effect on climate, and mitigating emissions will significantly decrease the resulting negative impact. The truth is probably somewhere in the middle. We can expect some up-front costs (any policy has costs), but the net secondary impacts for all policies (including "do nothing") are highly uncertain.
Therefore: we ought to do something, because even though we've got all this uncertainty in the various elements of the equation, it still ends up on the positive side.
Now, once we get into the details of what, in particular, we should be doing, we can try to attach numeric probabilities to these various quantities and compare policy A versus policy B versus the opportunity cost of doing either instead of policy C. And it's possible that in that debate, we might decide that the benefits we get from any climate change mitigation policy pale in comparison to other options like solving world hunger or what have you, but the important thing is that a necessary prerequisite for any policy discussion is agreeing that climate change mitigation is a problem that needs to be considered.
And I think that's where the public discussion is currently getting stuck. Public opinion is coming to the conclusion that it's an issue that needs to be addressed, but it's been slowed by interests who have fought against recognition of the problem at all. And we can't really have productive discussions until we reach that starting ground of agreeing to evaluate it on an equal level with all our other problems...
EDIT: I forgot about a couple other lines of evidence that strongly support the idea that increased CO2 is due to burning fossil fuels: (1) there's a geographic correlation (northern hemisphere vs southern hemisphere, taking into account atmospheric mixing times) between CO2 levels and industrialization, and (2) the isotope ratio of atmospheric carbon is consistent with the extra CO2 coming from fossil fuels. So actually, when you take that along with the correlation in timing and amount of CO2 increase, what I said above is wrong -- we ARE fairly sure where the extra CO2 is coming from, and it's from human activity.