Panel Discussion

Should We Study Geoengineering?
A Science Magazine Panel Discussion

Last week, an elite group of climate researchers and policy experts met in Cambridge, Massachusetts, to discuss geoengineering, the deliberate tinkering with Earth's climate to reverse climate change. Science covered the event, and its implications, on ScienceNOW and in the magazine .

Four panelists who attended the conference joined Science this week to discuss scientific, political, and moral issues surrounding geoengineering research. They included:

Jim Anderson, a Harvard climate researcher

Dan Schrag, a Harvard geochemist and co-organizer of last week's workshop

Penny Chisholm, a biological oceanographer at MIT

David Victor, a Stanford political scientist

Science: This conference that just finished up at Harvard talked about actually altering the planet. How did we get to this point?

Jim Anderson: Well, the motive force for this concern, of course, starts with fully grasping the degree to which the demand for energy is increasing.

We looked at this additional 20 terawatts [of energy] required between now and 2050, which translates into two, two-and-a-half new coal-fired power plants per day for the next 4 decades. This is becoming directly evident from the information that Dan presented to us that we have increased from 6.5 to 8.4 gigatons of carbon entering the atmosphere between 2003 and 2006.

And then coupled to that is the climate system and the rate at which we observe it to be changing. Arctic summer ice has dropped by 50% just since 1980 in the latest data, and this arctic summer ice is now forecast to be essentially gone by somewhere around 2020. And this removal of arctic sea ice drives these powerful feedbacks of ocean heat input, downwelling, and the release of permafrost.

Dan Schrag: I mean, to put it simply, there are many things in the climate system that we don't understand, and most of them lead to exacerbating current projections of the impacts of future warming. And we don't seem to be doing a very good job of controlling emissions, and therefore people have suggested trying to offset the greenhouse gases with reflecting shortwave radiation, and this conference was trying to explore that.

Part of the intention of the conference was really just to broaden the discussions and actually bring a broader spectrum of climate scientists and scholars into this discussion, because it's been a narrow subset of the community that's looked at this.

David Victor: Politically, it is proving incredibly difficult to control emissions. Emissions control is one of the hardest problems to address because it requires a very large number of countries to cooperate. Those countries have often diverging interests, and at the end of the day it's very difficult with the known tools of international law to constrain a country that doesn't want to do its part to control emissions. So, in addition to all the climatic factors that have been identified politically, this is proving to be even more difficult than people had anticipated originally.

Science: What did each of you take away from the conference, and did it surprise you?

Penny Chisholm: I took away two things. The first thing I took away was a heightened sense of alarm. I'm not a climate scientist, so some of the data that was presented brought into focus how just in the last few years things have gotten seriously worse, apparently. And the other thing I took away from it is how incredibly complex [geoengineering] is, much more complex than I appreciated before this discussion with all the various perspectives.

Anderson: The issue is really not global warming, and to state it that way is to misconstrue the primary issue. The real issue is global irreversibility, the loss of control of the system. And that to me was brought home very clearly, and it pervaded the entire discussion.

Schrag: For one thing, it was clear to me that the claims that have been made of the effectiveness of geoengineering--that is, climate engineering--were overdone. Our understanding of what putting reflective material into space or into the atmosphere would actually do to the climate is really crude. To a certain extent, we will never know exactly what it does because the climate system is complex and difficult to understand and probably impossible to predict in detail. But we're not even started yet in terms of understanding what things would do.

In response to what Penny just said, the complexities of the climate system--and all the uncertainties about the climate system that make you very nervous about trying to engineer it, try[ing] to control it--make me doubly nervous about what greenhouse gases are doing to it. And so I really think we're in an awful place where we have to explore something that makes many of us feel uncomfortable, and yet the alternative may be worse.

Victor: I agree completely with everything that Dan has just said. I'm a political scientist, and one thing I take away from this conference is that this option is not implausible. I had always just kind of relegated [geoengineering] aside and thought it was interesting to talk about and think about, but you really wouldn't do this at scale.

But it's not implausible. There are going to be unpleasant side effects and unknown-unknowns. There are going to be so many interactions that it's probably unlikely we're going to have a decent sense of how the climate system is going to respond without testing some of these options at scale. And even the test phase is going to raise a lot of complicated and difficult questions about how to regulate this internationally, how to compensate countries and people who are harmed. Those are really difficult conversations.

Science: David, do you think discussing geoengineering will undermine efforts to cut greenhouse gases?

Victor: I think it's an academic question, because the geoengineering option is becoming available autonomously. One effect is what is sometimes called the moral hazard effect or the laziness effect--that talking about this will make people complacent. I will say that the Harvard meeting showed that there are a lot of really nasty side effects [to geoengineering] and even more that people don't know about. So I'm not sure that talking about this option will automatically cause people to become lazy on controlling emissions. If anything, it will possibly make folks aware of how sensitive the climate system is and how crucial it is that we deal with the problem at the source.

Schrag: I think that's exactly right, David. I think that's a critical point. I think it's fair to say that some of the people who have been advocates for climate engineering in our community and in the broader scientific community have made claims that we can control it perfectly--that we can actually produce a climate that's exactly what we want it to be. I think that flies in the face of all our understanding of the climate system. In fact, what was very clear to me after this workshop is that [geoengineering] is not something that will ever compensate either perfectly or even imperfectly for greenhouse gas emissions. It may be a desperate measure, but it doesn't really take the pressure off for reducing greenhouse gases, because living for centuries or millennia in a climate that has elevated greenhouse gases and we're trying to compensate and making mistakes all the time--that's not a world that anybody wants.

Victor: I agree completely, and let me just add that, in addition to all those factors, this has to be viewed as part of a rational overall strategy for thinking about the problem. Because if the climate proves to be a lot more sensitive--the data on the sea ice I found terrifying, and the evidence keeps coming in--then we will need a quick-response option that you might deploy for a while. It would be truly outrageous if we didn't think about that option and probably even do some more work on it, and test some elements of it, so we're ready.

Schrag: We talked about psychological aspects at the meeting--the problem that people are very uncomfortable about errors of commission rather than errors of omission. And so the intentional engineering of the climate makes people very uncomfortable even if the unintentional changing of the climate by greenhouse gas emissions might even be worse.

At the same time, there's another psychological element that we didn't discuss much, which is when people get scared they demand action, and we could see this after 9/11, where a huge, vast majority of the American public supported going into Afghanistan. The question is if people ever really got scared about the climate--really scared, like post-9/11 scared--and they demanded action of their leaders, I worry that an aggressive plan to reduce emissions may not be considered decisive enough. And this is going to be on the table.

Anderson: I think that's right. In my mind, after transitioning from putting geoengineering in the category of an unacceptable approach, just a very few years ago, I think geoengineering will bring a focus to this debate that no other discussion can engender. And I think it's a focus of our scientific knowledge--what we do and do not know. It's a focus on the very intimate link between the energy issue and the climate issue. It brings focus on the political structure, and in particular [on] which political leaders bring the blueprint to the table that would actually lead to the negotiation of international treaties and so on.

When you look at all of them together, that's when you realize that the rates of all these changes are out of control. Regaining control of the system--the global system, the energy structure, the political system--this is the challenge. And I think that's why geoengineering carries with it the requirement to look at these things together.

Science: Penny, do you think we should be studying geoengineering?

Chisholm: I think we should discuss it, talk about it, and perhaps do research on it. One thing I would disagree with in what Jim just said is "gaining control of the system." I think, as a biologist, understanding the complexity of the biosphere and its feedbacks on the climate, I think, it might lead us down the wrong path to think that we can actually control it.

Anderson: I'm not talking about complete control. I'm talking about garnering control over a system such as the arctic glacial system, which of course is just a small part of the connected, global mechanism, but if you look at the numbers for arctic sea ice and what that means for Greenland, that's where the issue comes up. For example, it took 5 X 1021 joules [of energy] to melt the arctic ice cap, and in order to regain control, you need to decide how that thermal energy is going to be extracted from the system to put you back on a stable course.

Schrag: Penny's point is critical. The idea that we can control this is wrong. It may be that a world with climate engineering is better than a world without geoengineering because we're losing control of climate change--but that shouldn't make us think that we actually know what we're doing.

Anderson: No, if it's construed that way, I completely agree with both of you. It's just that you do have to look at the risks and we have to look at alternatives. If there are components for which we have lost control, like arctic sea ice, we have to look to that component as well.

Science: I just want to ask David: How might you regulate geoengineering? Can you regulate this on a global scale, and how might that work?

Victor: I think eventually you could regulate this on a global scale. Right now, it's not clear how to build a regulatory system, and it's also not clear how you'd begin those negotiations for two reasons.

First, we don't know what to do. This is still a research program, and it would be unfortunate if the regulatory discussions took a bunch of options off the table just because we didn't know how to think about them right now. There have been treaties in the past that have done that inadvertently, and that steers the direction that countries are going to be willing to spend limited resources on.

The other problem is that this is the exact opposite of the emissions-control problem. The emissions-control problem is all about how to get a large number of countries to comply with obligations to control their emissions. This problem is the inverse. It's feasible--or at least imaginable--that individual countries, and possibly even individuals, could do this, and that makes it especially difficult to constrain behavior under international law, because most legally binding treaties require governments to consent to be bound. If a government finds this inconvenient, they can leave. So in my view, the way to proceed with this problem is to treat it as a research program and to make it an explicitly international research program and to be sure that you involve the countries that are possible or likely geoengineers of the future in that program, so that you start to build up some norms about responsible research, responsible deployment, about sharing information, about peer review, and things like that.

My guess is that over time if we start to see some of these options as being more feasible than others, then you're going to need a regulatory system. Possibly with some kind of international control mechanism, [countries] maybe even [could] do this as an international deployment activity, but that I think is still a ways down the road.

Science: What kind of science questions do we need to answer if we're going to study geoengineering?

Anderson: There is a huge disconnect between knowing what we need to know and what we know. In fact, the research program is really quite inappropriate to the scale and severity of the problem. We need a far more complete observation of the global glacial system, a high-accuracy time series of how ice is being lost from Greenland and the Antarctic, far more sophisticated forecasts of the dynamics of these glacial systems and what impact that has on projected longevity, for example, of Greenland [ice]. We need very high accuracy observations of the global climate system as a coupled structure. That is, time series of temperature, water vapor, clouds, and how they're responding to the carbon dioxide forcing. It's very easy to calculate the forcing in the system, but the way in which the global climate system responds--and the way that's captured quantitatively in climate forecasts that are tested and trusted--is a huge gap.

Schrag: We have a problem, Jim, if we're going to use the Earth as an experiment--which we're already doing by adding greenhouse gases. If we're going to do an experiment by testing injection of reflective material, say, sulfur, into the stratosphere, we don't have a control. And so if something happens, it's almost impossible, given the complexity of the system, to attribute it either to the CO2 or the sulfur.

Science: What do you mean by "something happens"?

Schrag: Well, say we did the experiment and we started in March by injecting sulfur, and in June or July we see a series of giant storms that we've never seen before. Are those related to one or the other?

Chisholm: In that sense, I don't think it's right to call it an experiment, because it implies that it's scientific. And it's really not.

Anderson: Dan, I'm looking at something that's quite different than that. Look at the way weather systems drive perturbations in the overall climate system. ... There's a really clear mathematical formulation [that] tells you how that system responds to those perturbations. ... Nature provides us with very large perturbations that we need to study very carefully, including the glacial ice system, including the ocean steps, that's what I meant.

Schrag: I agree with you, Jim, I just think that those are the sort of observations we need today for the climate system as well. Not just for geoengineering or climate engineering, but we need them for climate change and we don't have them. So we're kind of in the dark today, already.

Not only do I want to minimize the negative impacts [of geoengineering], to the extent that we can understand them, but I also want to make sure that it doesn't ever fail. That aspect of it is real important. So to me, a huge fraction of the research effort in this area needs to go into what could go wrong. I'd love if the smartest people in the climate science community could spend a little bit of time thinking about all the possible ways such a system could fail, because that we really need to know about.

Anderson: I think we all can agree on that. I think the question for us was the foundational research program that would move us in the direction of understanding all the couplings in the system. ... The truth is that the federal structure is not moving rapidly to place that fundamental information in our hands.

Science: I want to just finish with one last question: Do you think your role as a scientist who studies this is changing? How did this conference affect you personally?

Victor: I find myself looking at and thinking about a lot more unpleasant options than I used to look at.

Anderson: It really does come back to dealing with this issue as early as we possibly can, first of all so we can understand it in a much better way. But also to articulate the issues and feedbacks and loss of control and so on. So yes, I think it has significantly changed our role.

Schrag: My only feeling is that the system is a lot more complicated than I imagined. I knew it was complicated. But I'm surprised at how little we understand it.

Chisholm: I think the issue is that people will look to us for guidance and answers. But the answers aren't there. This is what I struggle with. So we're trying to think about it, talk about it scientifically, but in many way we're going to have to do this blind, because it's just so extraordinarily difficult to understand the unknown unknowns, which are enormous. That's what really worries me about this. And I think a lot of us came away from that meeting much more scared than we were when we went into the meeting. Frightened about climate change, and frightened about what humans might get desperate enough to do about it.