Well it really does, and it has other nasty habits you would know about if you ever got a face full. Eli has had that experience too, it ain't no walk in the park, and having written serious SOPs for students working with the stuff (we are oh so DOE Tiger Team at Rabett Labs when we ain't sniffing gases out of cylinders) he was always a bit cynical about proposals for blasting zillions of tons of SO2 into the upper trop, besides the question of how you hope to get it up there.
The bunny was going to post something oh so smart on this, among other things pointing to Ken Caldeira's papers on the subject, another reason why Levitt and Dubner seemed a bit full of it when reporting on Caldeira's opinions, however, James Wimberley says it better
(a) this and all such schemes on the relevant scale are seriously dangerous, since we don’t understand the possible side-effects, including changes in regional weather patterns, and anyway leave untouched ocean acidification and other non-greenhouse effects of increased CO2,
(b) the geoengineering options must be studied in depth as an emergency Plan B in case humanity doesn’t cut emissions enough, or the climate turns out to be more sensitive than we thought to those already made. . . .
That's pretty much Caldeira's position.
It’s possible, however, that these climate experts don’t know much about global governance, and I know a little, having spent my working life in intergovernmental cooperation, so here goes on that side of the problem.
1. Because large-scale geoengineering is dangerous, it will only become a live option when emission control efforts have clearly failed and things have reached a crisis: hundreds of thousands dying every year in droughts, hurricanes, coastal floods and so on. The polar bears will already have gone. Whoever does it will need cast-iron political cover against the unforeseen consequences – including the risk of killing millions more.
2. For the same reasons, the measures cannot be national or regional in scale. They will be inherently global in their effects, even if carried out by or in a single country. The political cover accordingly has to be global.
3. There’s only a little room for experiment – primarily to test engineering feasibility and cost (say of Venetian blinds in space.) There’s so much noise in the climate system that the effects of small-scale pilot projects won’t be properly measurable. It will have to be live or nothing.
4. The knowledge required to manage an emergency global geoengineering scheme is very considerable, and very rapid and expensive action will be essential when things go wrong, as they probably will. Accordingly the scheme cannot be run democratically with any hope of success, only technocratically. Thought experiment: you have a project running on ocean fertilisation with iron in the Pacific. Evidence has come up that this is pumping up the El Niño cycle, with droughts and fires in Australia and the collapse of Peruvian fisheries. Do you suspend or not?
That's only the beginning. The bottom line is that geoengineering requires fleets of black helicopters to get done. The requirement for something that will not amuse the guys at the Breakthrough Institute and their CEI/Heartland type funders. (OK, that's a WAGNER, but Eli is a smart bunny). Stuff like that on a global scale requires a global Ghengis Khan to pull the strings.