Will Wilkinson quotes me arguing that geoengineering should be our last hope, for use after other plans, including emission cuts, have failed, and he writes:

I’ve thought the matter through, but I still don’t understand this ordering of priorities.  I understand the strategic political motivation to make all potential technological fixes to global warming seem like wacky, hare-brained, mad-scientist schemes to block the sun, but the more I think that through, the less it looks like responsible politics.

Just suppose that some form of climate engineering could (1) do as much or more to slow or halt warming than could regulatory approaches (2) at a much lower cost while (3) posing no special problem of international coordination. Perhaps Avent has already made the case that some technology (or combination of technologies) meeting this description is less likely to emerge in the coming decades than an effective scheme of international carbon emission controls. If he has, I’ve missed it. However, if the success of a primarily technological approach is no less probable than the success of a primarily global political-regulatory approach, it would be egregiously irresponsible to discourage public support of efforts to discover such technology. If the probabilities turn out to favor engineering over politics, then emissions cuts, not engineering, should be considered last ditch.

I think that finding a workable geoengineering solution which meets Will’s criteria will be harder than he thinks. In his comments section, he mentions super-carbon-eating trees as one technology he has in mind. Of course, increasing the number of regular trees, particularly in equitorial regions, would also be extremely helpful to the cause of slowing warming, but that has proven to be quite hard to do — so hard that it’s the subject of international negotiations, and we’re all struggling to find ways to generate the right incentives for preservation of rainforest and so on.

But let’s say we develop super-carbon-eating trees. It will still take a lot of trees to offset the emissions being produced by humanity. Those trees will compete for land area with other potential uses, including agriculture, and so there will need to be some incentive mechanism in place to ensure that enough land is devoted to carbon eating. But that will displace other uses, increasing the price of their output, which may cause some domestic political headaches. And no one country will want to have all the carbon eating trees (because they’d rather be using the land for agriculture or whatever) and so negotiations will have to take place to determine who will plant what where according to which formula.

Meanwhile, the population of earth will continue to grow, and energy use will continue to rise, and carbon emissions will continue to rise. At some point (and perhaps from the very beginning) human emissions will outpace the ability of the trees to absorb carbon. We may then want to adopt incentives to get people to reduce their emissions anyway, simply to make the trees effective, but then we’re back at the international negotiation table, figuring out how to get that done.

And the trees might fail. Let’s say it takes science a decade to develop a pretty good carbon-eating tree. Then let’s say deployment of the trees, to the point that human emissions are cut in half, takes another two decades (the things have to grow, you understand). That’s thirty years, during which carbon has been accumulating in the atmosphere, moving us toward concentrations that may well set off feedback loops the trees can’t hope to control. And if the trees don’t work at that point, then we have entirely missed the opportunity to reduce the threat of warming by pursuing slow reductions over those three decades. We can’t get that time back. And an emissions reduction program adopted thirty years from now would have to be incredibly strict and expensive to have any hope of working.

In other words, geoengineering deployment will share many of the problems involved in obtaining agreements to reduce emissions, will probably not work well unless we’re also reducing emissions, and may fail, in which case we’d really, really like to have been cutting emissions to avoid disaster. There’s really no scenario in which it isn’t a good idea to go ahead and try to cut emissions while investigating the opportunities for geoengineering.