So what are the biggest problems with Geo-engineering? The biggest objection is the unforeseen effects that may occur from our tinkering with the environment. If we put up lots of dust in the atmosphere to block out the suns rays, what might that do to various ecosystems around the world? Humans have a notoriously horrific track record on being able to predict the consequences of their actions on the environment.
Ideally, we would be able to conduct experiments to help determine the effects of some of our actions. The Mt. Pinotubo eruption in 1991 has provided some pretty concrete data on the effects of large sulfur dioxide injections into the upper atmosphere. Some of the results were a 1 degree cooling of global temepratures for a year afterwards, rainfall changes around the world, and a weakening of the ozone layer. Of course its the first effect that most people are now currently interested in, but the second effect, of affecting rainfall amounts could have enormous effects to certain parts of the world as well, essentially causing droughts in one part of the world, and flooding in others.
But if we look at the trade-offs that we are dealing with, global warming will also cause similar changes to microclimates around the world. Arn't we really trading one kind of flooding and drought for another? Both are man-made problems. Currently, the first world nations do little to assist others for the negative impacts of global warming on other nations, and legally challenges have mostly failed. However, with active geo-engineering, it would be much more plausible that legal actions would succeed.
On top of that, assuming that we become rather good at geo-engineering, or that we decide to engage in different kinds of geo-engineering, we might have impacts on local microclimates, rather than around the world, or we might become very good at manipulating the weather at other places. Recent CIA reports released indicate that the US military tried and was relatively successful at increasing rainfall during the Vietnam war, which slowed down supply lines to the Viet-cong by muddying up roads. If we begin manipulating weather in a global manner, one could possibly foresee that we happen to cause droughts in the middle east, and increase rainfall in perhaps the breadbasket, or in central California. We're probably far away from being able to manipulate weather on that level, however.
The philosophical problem is that these arguments rest on the precautionary principle. We ought not do X because of the possible outcomes of X. If we don't really know the outcomes of X and we have reasons to believe that ABC might be the consequences of X, then as a matter of prudence we shouldn't do X (assuming ABC is undesirable). It sounds like a pretty reasonable principle, but ultimately it rests on an appeal to ignorance. If I don't do X, the potential consequences are DEF, in this case global warming and all the potential consequences of it. The precautionary principle cuts both ways, and gives us no reason to act one way or the other. But if I don't do X, then I must do not X. Anything I place as alternative options for X, say reduce carbon emissions and let things work out naturally, also suffer from the precautionary principle problem.
So if the precautionary principle doesn't give us a solid reason not to engage in geo-engineering, and from the reasons in my last post we should consider engaging in geo-engineering, the arguments seem to favor engaging in geo-engineering. First world nations would simply have to accept the (legal) responsibility of the fall out of geo-engineering, whereas they have been able to escape the (legal) responsibility of global warming for the most part.