dumb idea to not get onto our climate engineering, if you ask me, we're going to need to master it sooner than later... but if anyone deliberately changes the climate the run the risk of being blamed if/when something goes wrong. Do we have the political willpower to actually push forward a global climate management strategy? Not by a long shot...

The climate is an EXTREMELY complicated thing, and we're far from understanding it well enough to be sure of the side-effects of dramatic deliberate changes -- the math doesn't even exist yet do a good job of this stuff. I hope that will be borne in mind if we develop substantial abilities along these lines.

The nice thing about engineering is that you don't have to understand a system perfectly to get it to do what you want. Transistor radios were being mass-produced before anyone really understood how a MOSFET worked.

Really, the only problem is we only have one Earth, so we need to be a bit careful, but I wouldn't call it a dumb idea by any means.

I didn't call it stupid. But there's a difference, abge, between not understanding why something you're doing works, and not understanding what the side-effects are going to be.

True, the former usually implies the latter to *some* degree, but if you start engineering perfect rain in the heartland (say) with great success before knowing whether the fourth-order effects are going to cause hurricanes in the gulf, then you're probably not being responsible.

@semck

I was responding to the OP, not to you.

I agree we shouldn't just start shooting chemicals into the air and hope it goes well. At the same time, small-scale environmental engineering could actually help us learn more about the system. As I said, we need to be responsible, because we only have one Earth to ruin, but that doesn't mean we need to wait until science has unlocked every answer; by then it will be too late to do anything.

Climate engineering is a terrible idea. Assuming it has any effect at all, the consequences of error are too high.

You can tinker with a bunch of transistors and the worst thing that will happen is some dumb ass gets hisself electomacuted.

'we're far from understanding it well enough to be sure of the side-effects of dramatic deliberate changes -- the math doesn't even exist yet do a good job of this stuff. ' i entirely disagree, we have the physics, the math is very well understood, the physics even is well understood - what we mostly lack is sensor data and computational power. We lack understanding and some knowledge of the biosphere - but that is mostly biology and hardware, not math.

And with a few small tweaks we might learn how what the effect of doing thing on the system is. Tinkering to help us understand could save us a huge amount in the long run.

'Climate engineering is a terrible idea. Assuming it has any effect at all, the consequences of error are too high.' - the consequences of not engineering could be too high. So that's not really a useful way of thinking...

@orathaic "the consequences of not engineering could be too high. So that's not really a useful way of thinking..."

Considering the magnitude of effect is *always* a useful way of thinking.

This "climate engineering" idea is a pretty scary side-effect of the whole climate change debate. I 100% buy into the idea that we might be inadvertently affecting the climate now, and we should stop. But to take it a step further from "we understand the science enough to attribute cause to effect" to "we understand the science enough to take deliberate actions to reliable achieve an effect" is a complete over-estimate of how much we actually know. We can interpret the results of the experiment we're accidentally doing now; that doesn't mean we know what we're doing otherwise.

All the scary predictions about climate change are working too well as public relations: "the consequences of not engineering could be too high" - man, back up a step and stop thinking like a cornered animal.

"physics even is well understood"

Are you kidding? You mean like "predictive model" type understanding, like the kind you'd need to have to make a plan?

"i entirely disagree, we have the physics, the math is very well understood, the physics even is well understood - what we mostly lack is sensor data and computational power."

I'm sorry, but you're wrong. The kinds of nonlinear differential equations that describe climate can be approximated to low orders, but they're very poorly understood in the kind of detail you would want in order to do this -- we don't even know where to begin studying in order to analyze them. We do understand a few things very qualitatively, and a lot of what that would tell us is that we shouldn't screw around until we understand things *very well.* The field of dynamical systems is beginning to give additional qualititative insights, yes, but that, too remains in too early a state for us to be comfortable screwing around with things (at least on a large scale).