Does geoengineering have to continue for thousands of years?

Cloud whitening, one geoengineering strategy

Chris Reynolds at Dosbat expresses the basic pro-science critique of geoengineering:

With the seriousness of the situation regards AGW becoming more clear, talk of geo-engineering has increased. I think that rather than try to stem population increase, examine our economic system and expectations, and reduce CO2 emissions, this will be seen as a viable option in the years to come.

However geo-engineering will, I am confident, be used as an excuse to carry on emitting CO2 and avoiding dealing with the fundamental flaw in our civilisation; exponential growth in a finite world. It is dangerous and is a recipe for disaster.

He may be absolutely correct on all points; the only parts I take issue with are the bolded ones (and the finite growth thing, a bit.)

The first point, the "rather than," is one we have to confront on a regular basis with the adapt-nik ("Don't mitigate -- adapt!) subspecies of lukewarmer. To wit: we can (and must) do more than one thing at a time. The resources required to investigate and prepare for the possibility that we may need to geoengineer are miniscule relative to mitigation, adaptation, or even population control.

If we were only going to do one thing, it sure as hell wouldn't be geoengineering. But we're going to have to do more than one thing.

The second point, that geoengineering may be used as an excuse to defer action, is a serious concern. Geoengineering would only work as a temporary bridge to allow intensive mitigation to bear fruit. But would it, actually, become an excuse for inaction?

This is a different, and slightly more upbeat, counterargument compared to the related riposte: "We don't need an excuse for inaction; we're excusing it fine as it is." In that fatalistic outlook, geoengineering becomes like a clean needle program for heroin addicts: we wish we could fix the underlying issue; we can't; we're going for damage control.

I am not such a fatalist; and I do not necessarily think that the availability of geoengineering will make mitigation less attractive. Consider, for example, how the adapt-vs-mitigate debate has unfolded (or failed to unfold) after Hurricane Sandy. Experts looking at the flood surge have suggested we could have prevented a large portion of the roughly $50 billion damages using $10-$15 billion dollars in floodgates. So my question is: Where are the adaptniks screaming for these new defenses?

I was able to find a tepid endorsement from Bjorn Lomborg which radically lowballed the cost of said adaptation.

Much of the risk could be managed by erecting seawalls, building storm doors for the Subway, and simple fixes like porous pavements – all at a cost of around $100 million a year.

If you follow the link, it leads to an article from Popular Science which includes this:

If New York—part of the Northeast megaregion—suffers a direct hit, workers will spend weeks pumping a billion gallons of brackish water out of its subway and train tunnels. The salt will corrode power lines, transformers and thousands of brakes and switches that control the trains. Some subsystems could take a year or more to restore.

To avoid such a scenario, New York state recommends the city invest well over $100 million a year in storm protections. City planners are already experimenting with dozens of low-tech fixes, says Adam Freed, deputy director of the Mayor's Office of Long-Term Planning and Sustainability.

 Note "well over $100 million" not "around $100 million." And while the original source describes these "dozens of low-tech fixes" as merely able to mitigate the nightmare scenario, in Lomborg's retelling they eliminate "much of the risk."

Lomborg also repeats the fallacy that the risks of a damage storm surge have nothing to do with climate change -- even though sea level rise, by definition, makes storm surges more destructive.

And that tepid, dishonest, weasel-worthy endorsement was really about all we heard from the "adapt, don't mitigate" crowd after Sandy. Judith Curry had a guest poster make the pitch for more weather satellites. Lucia Liljegren, not a word. So why not pitch adaptation in this brave new world?

My theory is, actual real-world adaptation makes the problem of global warming too real. It's one thing when "adaption" is an abstract concept describing something we may do in the future. But when it actually comes down to spending tens of billions of dollars on flood defenses, planned retreat from parts of the coastline, restoring wetlands, lowering levees, hardening the power grid, and beefing up the first responder network -- well, if you start spending that kind of money (hundreds of billions for starters, talking about the US alone), people might get to wondering why this global warming stuff is so gosh darned expensive and getting more so. And that might lead them to ask when we are going to stop adding to the bill by spewing billions of tons of CO2 into the air.

Geoengineering might similarly offer the public some clarity on this issue. I assume it will be far more expensive that it currently seems, it will be highly controversial on the world stage, it will have unwanted side effects and limited efficacy. Researching and preparing such a system might have the opposite of the effect Chris expects; it might focus the public's mind on what a god-awful problem this is and how we need to get busy fixing it.

If we research and prepare this tool (not deploying it until/unless we win an international consensus and after warming has crossed a specific threshold or we see evidence of rapid catastrophic feedbacks) the debate which ensues may, as Sandy has, stimulate the public and international debate on mitigation, so as to prevent or minimize the use of such desperate measures.

Another major concern with geoengineering schemes is the impracticality of keeping them running for a long, long time:

Why do I say we would need to keep up SRM for millennia?

Archer & Brovkin's 2006 paper "The Millennial Atmospheric Lifetime of Anthropogenic CO2", PDF, shows that the emissions of CO2 will remain in the atmosphere/ocean system for thousands of years. Their abstract sums up their findings perfectly:

The notion is pervasive in the climate science community and in the public at large that the climate impacts of fossil fuel CO2 release will only persist for a few centuries. This conclusion has no basis in theory or models of the atmosphere/ocean carbon cycle, which we review here. The largest fraction of the CO2 recovery will take place on time scales of centuries, as CO2 invades the ocean, but a significant fraction of the fossil fuel CO2, ranging in published models in the literature from 20–60%, remains airborne for a thousand years or longer. Ultimate recovery takes place on time scales of hundreds of thousands of years, a geologic longevity typically associated in public perceptions with nuclear waste.

So if we take any geo-engineering scheme that doesn't involve massive emissions reductions or active draw-down of CO2, we need to keep it up for at least 1000 years, the more CO2 we emit the longer the recovery of CO2 back to pre-industrial will take.

And if we falter...

That's obviously a legitimate concern. There isn't a single government on the face of the Earth that has maintained its present form of government for even a single millennium. If we are expecting them to maintain a stable geoengineering scheme for thousands of years, that's obviously impractical. And since solar radiation management strategies mostly poop out within a few years of stopping, you confront the possibility of decades or centuries of global warming hammering the world in the space of a few years.

Brrrrrrrrrr. I've scared myself. But perhaps the picture is not so dire. What if we look at geoengineering not as a mono-strategy, but, as I suggest, as one component of a threefold strategy of adaptation, mitigation, and geoengineering?

Let's say we get serious about mitigation and end up with 1,200Gt of CO2 equivalent added to the atmosphere (either it took too long to forge agreement, or the cuts could not be made fast enough, or the carbon feedbacks hit us too hard; we missed the 1,000Gt goal for 2C, but only just.) About 60% of that shows up in the atmosphere; the rest is immediately taken up by the carbon cycle. 720Gt. That's us, permafrost melting, forest dieback, what have you.

Let's say 50% of that remains in the atmosphere 300 years later. That's 360Gt. Meanwhile we are practicing some solar radiation management with cloud whitening and contrails and aerosols injected into the stratosphere. But we also have been doing some carbon sequestration.

Carbon sequestration is hard: suppose we don't get it off the ground for 20 years and it then takes us 30 years to ramp up sequestration to a grand total of 3Gt/year (that's about 9% of current emissions). We then practice that for about 150 years. That would take out 450Gt, but some of that would have been sequestered anyway -- we will only count 2/3 of the 450Gt as actual sequestration -- 300Gt. 

That leaves us with 60Gt above preindustrial -- about 330ppm of CO2 -- and we can probably stop spewing stuff into the sky at that point, the 200-year mark. Two hundred years is a very long time, but it is a lot less than thousands of years. Many governments have been more or less stable for 200 years, including the United States.

The exact figures are subject to debate, but the basic thrust is clear: two ideas that seem impractical on their own (solar resource management and carbon sequestration) get much more reasonable if you intelligently combine them with each other and intensive mitigation. Without mitigation, none of this works: you're continuing to shoot holes in the bottom of the boat as you're bailing it out.