Climate Engineering

The articles in the Spring 2017 Issues by David W. Keith, “Toward a Responsible Solar Geoengineering Research Program,” and Jane C. S. Long, “Coordinated Action against Climate Change: A New World Symphony,” provide informative views into several points of current debate on climate engineering research, its management, and governance. The authors agree on several key points. They agree that it is urgent to expand research on climate engineering interventions; that research is needed on both carbon-cycle and solar options; that research must address both scientific and engineering questions; that the agenda should be driven by societal need rather than investigator curiosity; and that research should target interventions that are plausible candidates for actual use, not idealized scenarios. They also agree that research must vigorously pursue two competing aims: to identify and develop interventions that are as effective and safe as possible, and to aggressively scrutinize these to identify potential weaknesses or risks.

Their main disagreement concerns how to organize research on the two types of climate engineering: carbon-cycle and solar methods. Long argues that they should be combined, because the two approaches must be evaluated, compared, and decided jointly, together with mitigation and adaptation, to craft an effective strategic climate response. Keith argues that they should be separated, because of large differences in the bodies of scientific knowledge and technology on which they draw; the nature and distribution (over space and time) of their potential benefits, costs, and risks; and the challenges they pose for policy and governance.

A first step toward clarifying this disagreement is to note that the authors emphasize different elements of policy-making processes. Keith is mainly concerned with designing research programs. His programs are not purely scientific in their motivation and focus, in that they aim to develop and test technologies that can contribute to solving a societal problem. But they are well enough separated from policy decisions, and from the comparative assessment of capabilities, risks, and tradeoffs needed to inform decisions, that their management and funding are best optimized for each type of climate engineering separately. Long is mainly concerned with assessment and decision making. She argues that effective climate policy making must strategically consider and compare all response types, and that assessments, scenarios, and research programs must therefore also be strategically integrated if they are to usefully inform policy decisions. The authors thus agree on the need for integration of carbon-cycle and solar methods in assessment, scenarios, and policy making, but diverge on what this implies for the design, funding, and management of research programs: separate programs for carbon-cycle and solar methods, or combined?

This question turns on whether achieving successful integration in assessment, scenarios, and policy making requires integration in research program management and funding. In my view, such dependency could arise in three ways. First, integrated research would be favored if a coherent and defensible research program mission cannot be defined at the level of one response type, but only at some higher level of aggregation: as Long points out, “make solar geoengineering work” is not a suitable mission statement for a research program. Second, integration would be favored if effective assessment requires strong control over research management decisions, including allocation of resources between carbon-cycle and solar interventions. Finally, integration would be favored if research governance needs are driven less by differences in the opportunity and risk profile of different responses, and more by aggregate public or political views of climate engineering that do not clearly distinguish the two types: in this case, integration might be required as a matter of political risk management.

Dan and Rae Emmett Professor of Environmental Law

Faculty Co-Director, Emmett Center on Climate Change and the Environment

UCLA School of Law

Jane Long makes several important points. Among them is that geoengineering research should not have as its mission the deployment of geoengineering concepts. She cogently argues that “The goal for climate intervention research must be to understand the potential efficacy, advisability, and practicality of various concepts in the context of mitigation and adaptation.” David Keith makes a similar point and provides two guiding principles: that research on solar radiation management should be part of a broader climate research portfolio on mitigation and adaptation action, and that research should be linked to governance and policy work.

We generally think of solar radiation management research in terms of small tests that can define particular parameters, such as the atmospheric residence time, transport, and fate of aerosol scattering particles. As both Long and Keith observe, these tests require thoughtful governance arrangements that may be difficult at present.

Twenty-six years ago there was a large-scale natural experiment in solar radiation management: the eruption of Mount Pinatubo in the Philippines that injected roughly 17 million tons of sulfur dioxide into the middle and lower stratosphere. Sulfate aerosols spread across the Pacific Ocean in a few weeks and around the globe within a year. Spectacular sunsets over the next two years were one indication of the stratosphere residence time of the aerosols. The event produced observed cooling in the Northern Hemisphere of 0.5 degrees to 0.6 degrees Centigrade, equivalent to a reduction in radiative forcing of perhaps 3 watts per square meter. Globally averaged cooling of approximately 0.3 degrees was observed.

Such natural experiments in stratospheric aerosol injection are infrequent. The eruption of Krakatau in 1883 produced a forcing of a little over 3 watts per square meter. There were three eruptions between Krakatau and Pinatubo that produced forcings of 1.5 to 2 watts per square meter and five additional ones of 0.5 to 1 watt per square meter. The average frequency was once every dozen years, although there was a long quiet period from about 1920 to 1963.

It seems both worthwhile and feasible to develop a program to learn from the next such eruption. Much was learned about scientific models from Pinatubo, but as the 2015 National Research Council report Climate Intervention: Reflecting Sunlight to Cool Earth stated, “More work is needed in characterizing these processes in nature (through measurements), and in modeling (through better model treatments and a careful comparison with observed features of aerosols and their precursor gases) before scientists can produce truly accurate models of stratospheric aerosols.” Understanding the chemical reactions, mixing, and particle formation after such an event can help characterize not only solar radiation management but also aerosol-forcing effects on climate. Global observations can help understand the consequences of solar radiation management on precipitation, plant productivity, and carbon uptake, among other effects.

The Climate Intervention report had a short section describing observational requirements for making better use of volcanoes as natural experiments. It points out that “our ability to monitor stratosphere aerosols has deteriorated since [Pinatubo], with the loss of the SAGE II and III satellite-borne instruments.” Both satellite systems and a deployable rapid-response observational task force (that would have other atmospheric science uses to occupy it between eruptions) are suggested.

The creation of an international program to learn from the next Pinatubo could jump-start both needed instrumentation and perhaps governance arrangements in a low-key way that could build trust and indicate whether governance of deliberate solar radiation management experimentation is feasible along the lines that Long and Keith describe.

Co-Director, Carnegie Mellon Electricity Industry Center

Professor, Tepper School of Business and Department of Engineering & Public Policy

Carnegie Mellon University

David Keith issued a strong call for geoengineering research, echoing calls that I and others, including Keith, have made previously. I completely agree with him that mitigation (reducing emissions of greenhouse gases that cause global warming) should be society’s first reaction to the threat of human-caused climate change. I also agree that even if mitigation efforts are ramped up soon, they may be insufficient to prevent some dangerous impacts and society may be tempted to try to actually directly control the climate by producing a stratospheric aerosol cloud or brightening low marine clouds.

It will be a risk-risk decision that society may consider in the future: is the risk of doing nothing in terms of advertent climate control greater than the risk of attempting to cool the planet? To be able to make an informed decision, we need much more information about those risks, and thus we need a research program.

My only disagreement with Keith centers on his overall favor of eventual geoengineering implementation. I think the governance issues will be much more difficult than the examples he gives. Remember, we are talking about governing the climate of the only planet known to support life. Air traffic control or international banking do not have to be perfect, and small mistakes, even if very unfortunate for those affected, will not result in a global catastrophe. And how can we agree on how to set the planetary thermostat, with imperfect compensation for those who end up with a worse climate? How will we ever be able to attribute regional climate changes, either bad or good, to geoengineering, when natural variability is so large?

I support geoengineering research because we need to reduce the unknowns. We may discover large risks that we are unwilling to take, and the research may end up with enhanced cooperation toward rapid mitigation, with the realization that there is no safe “Plan B.”

But what about the “unknown unknowns,” as Donald Rumsfeld put it? Will the world ever be willing to take a chance on a complicated technical endeavor to control Earth’s climate, in the hope that there will be no bad surprises? Will we accept whiter skies and not being able to see the Milky Way as easily as now? Will we trust the militaries of the world to not use this new technology as a weapon? Can we live with more ultraviolet radiation reaching the surface due to ozone depletion caused by stratospheric particles?

Doing our best with the limited resources available, we are now trying to see if we can produce some new combination of materials, locations, and timing of injections of particles into the atmosphere that will produce a better climate for most. So far, we have not been successful. But it is early days, and we owe it to the world to do much more such research, while at the same time advocating for rapid reductions of emissions of greenhouse gases that are causing global warming. It will not cost much, and it is money that will be a wise investment of the governments of the world. We can’t wait.

Distinguished Professor of Climate Science

Department of Environmental Sciences

Rutgers University

David Keith justifies his call for a large-scale international solar geoengineering field research enterprise in environmental justice terms. He argues that in light of the mounting evidence that emissions reductions alone may be insufficient to limit severe climate risks, beneficiaries of a research program to understand the risks and benefits of potentially deploying solar geoengineering technologies to rapidly cool Earth would include “the world’s most vulnerable people, who lack the resources to move or adapt” to rising sea levels and increasing extreme weather. Thus, the multiple “reasons for reluctance” that Keith acknowledges constrain support for solar geoengineering research must be weighed “against the evidence that solar geoengineering could avert harm to some of the world’s most vulnerable people.”

The problem is that such evidence is not established. The benefits and risks of any solar geoengineering program will be unevenly distributed across the world and nations might have widely divergent preferences for whether, when, how, and toward what ends solar geoengineering technologies should be deployed. Who would decide whether solar geoengineering is deployed to support the climate resilience goals of farmers in the Sahel or Bangladeshis if they conflict with, say, maintaining and expanding ice-free ports in the Russian Arctic?

According to Keith, a “responsible” solar geoengineering research program should “have an engineering core,” using atmospheric experiments to investigate detailed plausible operational scenarios for deployment. It would focus on assessing various researcher-determined measures of risk and effectiveness in achieving desired climate outcomes with results informing governance and policy developments.

This is not sufficient. Recent research suggests that in the absence of broader societal input and consent, even small-scale, low-risk field experiments will trigger concerns over the slippery slope to larger-scale, riskier experiments and deployment. Without the meaningful input and support from the climate vulnerable constituencies it is intended to benefit, a solar geoengineering field research program would lack much-needed legitimacy and risk significant opposition. A responsible research program needs to account for how climate-vulnerable nations and communities themselves might view the value of such a program and ensure that they are fully engaged in co-creating research and governance goals and objectives.

Thus, a responsible solar geoengineering research program should include several core elements. As a prerequisite, clear support for solar geoengineering research should be established from an international coalition of nations. This should include nations particularly vulnerable to climate change as well as high-carbon-emitting nations that are fully committed to ambitious emissions reductions. Research priorities should be explicitly codeveloped in collaboration with technical experts, social scientists, and civil society organizations from climate-vulnerable nations. Finally, an international research governance system must be designed with meaningful input from civil society to address concerns about transparency, liability, and justice.

Director of Science and Policy

Union of Concerned Scientists

Dean’s Professor of Sustainability Science and Policy

Northeastern University

David Keith’s article gives rise to an interesting question about the utility of the label “responsible research” in the context of solar geoengineering. One of the central tenets of responsible research is that society, broadly defined, should have a meaningful stake in debating and modulating the direction of scientific research. In the case of research on solar geoengineering, with its inherently global impacts, the development of effective mechanisms for facilitating broad societal discussions about the desirability of this direction for research seems to be hugely important. However, this is not Keith’s focus. The notion of a genuine two-way dialogue around the desirability of research on this topic is absent: society features either as people meekly awaiting the benefits of techno-scientific intervention or as subjects to be enrolled in research projects to improve the effectiveness of the intervention.

Keith’s treatment of the so-called “slippery slope” concern (that research may generate momentum toward deployment through various mechanisms of lock-in) is particularly revealing of his understanding of the proper relationship between science and society, suggesting an expectation that research can ultimately bypass the need for societal debate and discussion. For example, he claims that a slippery slope is not a problem in itself if “research reveals that solar geoengineering works better and with less risk than we think.” But this assumes that research will be able to establish “once and for all” whether benefits outweigh risks. However, this is simply impossible: not only is there much that is likely to be unknowable about such interventions, but there are also many different disciplinary and social perspectives about what would constitute acceptable levels of risk, about the kinds of knowledge that would be necessary to answer such a question, and even about the meaning of risks and benefits themselves. Presuming that scientific research will be able to come up with a single answer and make these disagreements go away is quite simply unrealistic. There will always be multiple, contested answers to the question of whether geoengineering is on balance a good or bad idea—hence, the need for a genuinely responsible approach to research that incorporates a wide range of societal stakeholders in deciding if (not just how) this kind of research should go ahead.

By belittling concerns around the slippery slope as unfounded as long as the science shows us everything is all right, Keith reveals an overblown faith in science and a fairly dismissive attitude to the concerns that other people might bring to this debate. Despite nodding toward a number of other arguments against research, he quickly concludes that these “do not amount to a strong argument,” before promoting his own particular (and questionable) view of the benefits of research. Closing down the space for debate in this way would appear to limit the possibility for a really “responsible” attitude toward any potential research in this area.

Research Fellow, Science Policy Research Unit

University of Sussex

Brighton, United Kingdom

By using the adjective “responsible” in the title of his article, David Keith points to a dilemma: responsibility goes forward and backward. In the case of solar geoengineering, there’s the forward-looking “move by humanity to take deliberate responsibility for managing the climate,” as Keith puts it, which can be viewed most generously as the caretaking or stewardship responsibility for creating conditions in which life can flourish. But there’s also the backward-looking taking of responsibility for past actions that created the situation, the “cleaning up our mess” part, which mingles with accountability and liability. Forward-looking responsibility is entangled with agency; backward-looking responsibility is entangled with causality and blame.

Keith points toward five reasons why people are reluctant to form a solar geoengineering research program: uncertainty, slippery slope, messing with nature, governability, and moral hazard. But there’s also a sixth: the notion that solar geoengineering represents an avoidance of responsibility. As one of the people interviewed as part of my studies of perceptions of solar geoengineering put it, “It’s like transferring the responsibility from myself to somebody else in tackling climate change.” There’s a transference of agency here, as well.

Who can take that backward-looking responsibility? Scientists and researchers can’t do much about this on their own, and the intense debate about “loss and damage” in the climate regime belies the difficulty. There’s no real social process for responsibility-taking on the scale of global climate change. The best that we have is the Common but Differentiated Responsibilities and Respective Capabilities principle included within the United Nations Framework Convention on Climate Change to acknowledge the different capabilities and differing responsibilities of individual countries in addressing climate change. Fossil-fuel companies, the states that subsidized them, and the citizens of rich nations who burned the carbon and benefitted from it all deserve some share of responsibility. But instead of putting a price on carbon, the US government subsidizes it—irresponsibly.

The dilemma is that a research program itself can’t be fully responsible as an independent, self-organized entity. The context is what makes it so. Right now, the context is one of extreme irresponsibility. Research based in the United States will be “responsible” only if the state and corporations are making attempts to curb the harm, recognize past harms, change everything. So what’s a researcher to do? Best guess: listen, be responsive, align with researchers around the world, and support them in taking their research in the directions they want it to go. Recognize and name whenever possible the irresponsibilities and asymmetries, rather than speaking of a common humanity that’s created the mess and now has the responsibility of repair. Prospects of actually governing this technology, like the prospects for governance of climate change, may depend upon such recognition. It’s beyond the common purview of science to take responsibility for more than forward-looking science or its outcomes, but these are extraordinary times.

Department of Development Sociology

Cornell University

David Keith provides a useful provocation for thinking about the intersections of science and society in the context of solar geoengineering research. What does responsibility mean, and for whom? Keith’s notion of responsibility seems to entail more “transparent” research on solar geoengineering to enable responsible decision making. To this end, he lays out some key issues (though certainly not all) raised by the prospect of solar geoengineering research, and he suggests that they are amenable to resolution through the provision of more science. However, a different account of the relationship between science and politics opens up a set of questions that he doesn’t address. The question of the “responsibility” of a decision—or a research program—is not just a matter of scientific facts, but of values, interests, and context. This raises important questions about the relationship between science and policy, the potential distributional implications of innovation, the role of ignorance and uncertainty, and the importance of public engagement.

Keith argues that an international research program on solar geoengineering—one that is linked to, but distinct from, research on carbon dioxide removal approaches (see Jane Long’s counterpoint to this claim for separation)—is urgently needed for societies to effectively manage climate risks, especially for “the world’s most vulnerable people.” But this argument demands further scrutiny. Keith seems to argue that by virtue of his expertise he knows what matters to vulnerable people, and that solar geoengineering research will benefit them. Scientists frequently make these kinds of claims, but as the British researcher Jack Stilgoe has pointed out, the history of technology suggests that many sociotechnical systems tend to exacerbate the gap between rich and poor, rather than close it. If we want to treat this as an empirical question, we might, at the very least, develop mechanisms to ask people who are indeed vulnerable if they want solar geoengineering research to move forward on their behalf.

Keith also argues that uncertainty alone is not a sufficient reason to oppose research, because “the central purpose of research is to reduce uncertainty.” However, this view of uncertainty may miss the mark in at least two ways: it misunderstands opposition to research, and it seriously overestimates the ability of science to resolve controversies about technology and risk.

With regard to the first point, for some opponents of research, ignorance is not only an option, but the right option. There are certainly some areas of innovation that, for better or worse, societies have chosen not to pursue (for example, human cloning). An “ignorance is not an option” rationale for research could have the effect of limiting social choice in problematic ways, and it implies a level of inevitability about innovation that is not obvious. Debates over whether or not to move forward with solar geoengineering research will tend to depend on how people perceive the purposes, values, and risks of research, which is not at all a straightforward proposition answerable by more science.

On the second point, as Arizona State University professor and writer Daniel Sarewitz has argued, persistent debates about genetically modified organisms, nuclear power, and chemical toxicity evince that science often does little to limit controversies—and can sometimes make them worse. Uncertainties in these domains often resist scientific reduction, more science does not always tell us how to act wisely, and partial knowledge can create excess confidence that action is worth taking. Promises that more research in complex areas will reduce uncertainties, and that this will compel political or policy action, should be met with healthy skepticism.

Certainly, many of these concerns extend well beyond the emerging domain of solar geoengineering research, including into climate change science and politics more generally. However, this isn’t a reason to sidestep thorny questions at the heart of science policy. Experience suggests that neither Keith nor any other expert has the political privilege of determining what “responsible” approaches to solar geoengineering might be. Democratic deliberation, not expert monopoly, should lead the way in discussions of the future (or not) of research in solar geoengineering.

Doctoral Candidate

Environmental Science, Policy, and Management

University of California, Berkeley

Cite this Article

“Climate Engineering.” Issues in Science and Technology 33, no. 4 (Summer 2017).

Vol. XXXIII, No. 4, Summer 2017