Roundtables
ShareHas the time come for geoengineering?
Scientists have long studied and debated the promises and perils of deliberately influencing Earth's weather and climate systems. But today, faced with ever more pessimistic predictions about the pace of global warming and the irrevocable damage it could do to the planet, some are talking seriously about implementing theoretical geoengineering schemes such as blocking the sun as an emergency response. In "20 Reasons Why Geoengineering May Be a Bad Idea" (May/June 2008 Bulletin), Alan Robock raises a host of scientific, social, and ethical issues posed by geoengineering. Below, Robock and his four fellow discussants debate how to weigh geoengineering’s potential benefits against its negative consequences.
All geoengineering techniques raise some common (and complicated) questions: Should legitimate research activities continue? Should experimental as well as theoretical research take place? Who decides whether an experiment or project can go forward? Are people concerned about geoengineering because they fear that the research might be harmful, or because they're worried that the knowledge gained might be dangerous? Are science and business mutually exclusive activities?
Over the past month, a variety of bodies have gathered to discuss geoengineering techniques in an effort to better understand them--and perhaps better control their research and/or practice as well:
- The Scientific Group of the International Maritime Organization's London Convention, which regulates ocean dumping, met in May to provide technical guidance to the convention's parties and included ocean iron fertilization on its agenda.
- The Convention on Biological Diversity, a framework agreement concerned with actions that may affect biodiversity, considered ocean iron fertilization during its recent Conference of the Parties.
- A workshop on geoengineering sponsored by the Council on Foreign Relations took up the foreign-relations implications of geoengineering.
- The American Enterprise Institute also sponsored a conference on geoengineering.
In addition to these meetings, last week, the academies of science for the Group of 8 Plus Five countries released a joint statement calling for additional geoengineering research. Secretary of Foreign Affairs of the U.S. National Academy of Science Michael Clegg interpreted their statement to include "approaches to soaking up carbon dioxide," specifically "the so-called fertilization of the oceans with iron."
Of this recent activity, we believe that the London Convention's proceedings provide a good model of how discussions between governments, scientists, and nongovernmental organizations may evolve. When the Scientific Group met in Ecuador, they formed an ad hoc working group on ocean iron fertilization to provide technical expertise in support of decision-making. The working group called on several oceanographers, including some that had participated in ocean iron fertilization experiments, for assistance in understanding technical issues.
Several external scientific groups also developed statements to inform these deliberations, including the Scientific Committee on Ocean Research; its U.N.-commissioned Joint Group of Experts on Scientific Aspects of Marine Pollution; and the International Oceanographic Commission (IOC). The academic research community also addressed questions of interest to the delegates in a Science magazine policy forum.
Although private companies and individuals cannot be parties to these agreements and cannot directly participate in meetings, parties to the London Convention provide opportunities for private concerns to inform its members through side sessions. Our company, Climos, made technical presentations during side sessions at the London Convention meeting last fall as well as at the most recent Scientific Group meeting.
The delegates reviewed a variety of scientific questions--ranging from whether large-scale experiments are justified scientifically (the consensus of the position papers from scientists was that they were) to whether ocean iron fertilization was harmful to the marine environment. (The consensus of the position papers was that there's insufficient scientific evidence to determine whether ocean-fertilization activities would pose any significant risks of harm to the marine environment.)
The Scientific Group will release its report on these discussions before the fall meeting of the parties in London this October, which will consider policy statements based on the input from the Scientific Group. The London Convention's legal consultants will also provide information on the legal basis for considering whether ocean iron fertilization is "dumping" under the technical definition of this activity within the London Convention.
On the other hand, the Convention on Biological Diversity adopted, with little deliberation or input from the scientific community and no input from knowledgeable private-sector stakeholders, a decision that expresses concerns about ocean iron fertilization and requests that governments ensure that activities don't take place "until there is an adequate scientific basis on which to justify such activities." But the IOC's ad hoc working group on ocean iron fertilization, of which Ken Caldeira is chair, recently released a response to that statement, saying that it "places unnecessary and undue restriction on legitimate scientific activities." The IOC will meet in Paris next week and will review the progress of the London Convention towards a scientific and policy framework for ocean iron fertilization.
We believe that the deliberative, science-based proceedings of the London Convention may serve as a useful model by which other international groups might consider proposals for adding aerosols to the stratosphere and other geoengineering activities.
Alan Robock suggests that I must respond to the "totality" of his argument--that it's not enough to pick off his 20 theses one-by-one, as if the "totality" of his position is somehow more than the sum of its parts.
For example, he repeats the canard that “geoengineering won't stop ocean acidification." The list of things that climate engineering won't do is endless. Reducing non-carbon greenhouse gases and black carbon soot also won't stop ocean acidification, but that shouldn't stop us from reducing them. Carbon emissions and climate engineering are two different things. Let's fault carbon emissions for what they do and fault climate engineering for what it may do.
In my earlier post, I wrote: "The Mount Pinatubo eruption lofted more than enough aerosols into the atmosphere to compensate for a doubling of atmospheric carbon dioxide, yet ozone concentrations fell by only 3 percent. And it's believed that this small reduction was caused by chlorine from human-made chlorofluorocarbons, which are now banned by the Montreal Protocol. So while the threat to the ozone layer is worth studying in greater detail, it's expected to diminish with time." Alan says that I'm wrong. But where exactly? Do we expect climate engineering at this level to destroy on average more than 3 percent of the ozone? Did chlorine not play an important role? Does the threat of ozone destruction increase with time?
Alan also asks us to agree "that if geoengineering is ever implemented, it must be geoengineering plus mitigation, and never geoengineering instead of mitigation." But what if we fail to enact effective mitigation? And what if there was substantial death and suffering as a consequence, and we thought the suffering could be alleviated, at least in part, by climate engineering? Shouldn't we then pursue it without mitigation?
While I'm in emotional agreement with Alan's statement, the rational side of me doesn't think it makes much sense. Some things sound nice, but may not be particularly reasonable when fully considered intellectually. Isn't the point that we wouldn't want to deploy climate engineering systems unless the benefits clearly outweigh the risks, where those risks include effects on both the physical climate system and sociopolitical systems?
Obviously, we'd like to avoid creating a world in which climate engineering is a necessity. Unfortunately, we may be heading down this path. Therefore, we have no choice but to carefully explore climate engineering's potential benefits and risks.
These benefits and risks can be safely explored using computer models and laboratory experiments. (If a decision is made that we need to proceed further, at some point these experiments would have to move outdoors.) These experiments should evaluate both intended and unintended environmental consequences of climate engineering--but we also need to investigate how we might go about constructing such systems, as it's the concrete proposals that will drive the environmental research.
Any real scientific research program will steer clear of value judgments and focus instead on the physical science and technology of climate engineering. Scientists have values, but science is about facts.
Thanks to my colleagues for their thoughtful responses. Geoengineering may indeed prove necessary, temporarily, if the benefits of geoengineering outweigh the negative consequences, and we all agree that much more research is needed to understand the costs, benefits, and potential harm of different scenarios. We also agree that if geoengineering is ever implemented, it must be geoengineering plus mitigation, and never geoengineering instead of mitigation.
Still, I'm concerned that the promise of geoengineering will delay implementation of mitigation. As I say in my essay, reversing global warming is a political problem. In their attempts to maximize profits, the fossil fuel industry has fought the science of global warming so they could continue using the atmosphere as a sewer without charge. The industry exerts tremendous influence in the White House and Congress, and uses the same techniques as the tobacco industry, which recruited scientists to argue that smoking was safe and to confuse the public, thus delaying smoking restrictions and causing countless deaths and suffering.
We are nearing the end of a similar battle with regard to global warming. The latest report from the Nobel Prize–winning Intergovernmental Panel on Climate Change (IPCC) does not present new conclusions; it merely synthesizes and strengthens science we have known for a long time. Yet, think tanks such as the Heartland Institute and the Cato Institute that are funded by fossil fuel interests continue to try to confuse the public. As they begin to embrace geoengineering in an attempt to continue business as usual (the American Enterprise Institute is holding a conference on geoengineering next week), we scientists have to be very careful not to facilitate their efforts.
True, there is little evidence yet of a concerted national or international effort to provide the needed regulations (a gradually increasing carbon tax and a prohibition of new coal plants that lack carbon-capture-and-storage technology) and the requisite research support for new energy technology (i.e., carbon sequestration, improved solar and wind power, and energy efficiency). But this doesn't mean that the regulations and support won't materialize soon. Europe is already leading the way in energy efficiency and regulation, and the United States will soon have a president determined to lead the world in this direction.
Ken Caldeira says that none of my arguments against employing geoengineering is nonnegotiable, but it's the totality of them that needs to be considered. My point about ocean acidification is simply to emphasize that geoengineering won't address all of the consequences of our rising emissions problem. He is wrong about the effects of ozone depletion from geoengineering. (See "The Sensitivity of Polar Ozone Depletion to Proposed Geoengineering Schemes.") And just because Edward Teller suggested that we could engineer our way out of enhanced ultraviolet light, does not make it so. Teller had some other ideas that were pretty bad, to say the least.
I agree with Tom Wigley that we should use "realistic" scenarios and state-of-the-art coupled atmosphere-ocean general circulation models to study the effects of geoengineering. Our recent such study in the Journal of Geophysical Research, shows, for example, that even Arctic-only proposals for atmospheric aerosol seeding would have huge consequences for the African and Asian summer monsoons. The low-intensity geoengineering scenario Tom suggests is only one such scenario, and a new geoengineering research program should take the lead of the IPCC and agree on a set of several common experiments to evaluate.
In contrast to the impressions of Dan Whaley and Margaret Leinen, I am strongly in favor of research on geoengineering and argue for it in my recent Science article, "Whither Geoengineering?" What the international community needs more urgently, though, is research into new energy saving technologies, such as cars that get 200 miles per gallon, public transportation, domestic wind generators that work at low wind speed, and better batteries. And we need to require that all new building roofs that face the equator have built-in solar panels, that all new buildings are built to standards set by the U.S. Green Building Council, and that taxes on huge SUVs make them prohibitively expensive to operate.
A well-funded research program could lessen uncertainties associated with any temporary geoengineering project. But the promise of geoengineering should not delay actions required now to address the root causes of global warming.
My comments are restricted to the type of geoengineering that employs injecting aerosols or aerosol precursors into the stratosphere. This is the primary focus of Alan Robock's article, wherein about half of his arguments against employing this type of geoengineering are related to possible detrimental environmental side effects. Our knowledge of these side effects is still rudimentary. He therefore endorses a "moderate investment in theoretical geoengineering research." His emphasis on "theoretical" accords with his view that we shouldn't undertake even small-scale stratospheric experiments until we know that "we could avoid . . . all of the potential consequences [presumably adverse] that we know about."
Of course, complete avoidance may be impossible--more sensibly, this should be considered a relative risk problem, balancing the possible negative effects of geoengineering against the possibly larger negative effects if we don't pursue geoengineering.
As others have pointed out, no serious scientists suggest deploying any geoengineering strategy now or in the immediate future--at least until we know more about the possible consequences. Nor has anyone suggested we employ geoengineering as a sole climate management strategy, circumventing emissions mitigation. Mitigation is essential, not least because a failure to mitigate would allow carbon dioxide to continue to accumulate in the atmosphere, resulting in further ocean acidification. The primary focus, per my recent Science article, "A Combined Mitigation/Geoengineering Approach to Climate Stabilization," should be on employing geoengineering as a means to gain time to develop and implement cost-effective, carbon-neutral energy technologies that will move us away from our current overwhelming dependence on fossil fuels.
To assess the relative risks of geoengineering to the climate system and stratospheric ozone, we need coupled atmosphere-ocean general circulation models or atmospheric chemistry studies that employ realistic geoenginneering scenarios. No such studies have been published to date. Not only must the geoengineering scenario be realistic (i.e., modeled as part of a combined geoengineering/mitigation strategy), but the results of such a study must be compared with a realistic no-geoengineering scenario.
Possible scenarios for comparison given in my paper are stabilization of atmospheric carbon dioxide at 450 parts per million (ppm) by 2100 through mitigation only, and, as a complementary combined mitigation/geoengineering scenario, an overshoot concentration pathway where atmospheric carbon dioxide reaches 530 ppm before falling back to 450 ppm, coupled with low-intensity geoengineering. These scenarios are projected to have the same effect on global temperature rise--stabilization at about 2 degrees Celsius.
The low-intensity geoengineering case is defined as injecting sulfur dioxide into the stratosphere beginning in 2010, ramping up linearly to a peak of 5 teragrams of sulfur per year between 2040 and 2060, and then declining back to zero by 2090, for a total injection of 130 teragrams of sulfur over 80 years. The consequences of this scenario in terms of sulfur deposition at the Earth's surface (i.e., what is commonly referred to as "acid rain") are likely to be minimal: Globally, 130 teragrams of sulfur is only about two years' worth of current emissions from fossil fuel burning. Optimizing aerosol size and location would require an even smaller injection rate into the stratosphere. Even medium- and high-intensity geoengineering scenarios would lead to relatively little additional surface sulfur deposition when compared to sulfur dioxide emissions scenarios from fossil fuel combustion projected by the Intergovernmental Panel on Climate Change. Likewise, the effect of stratospheric aerosol loading on cirrus clouds is likely to be minor, and the slow ramp-up would give people plenty of time to adapt to suggested changes in the blueness of the sky.
Of course, as Alan points out, regional differences in surface sulfur deposition are possible--the effects from geoengineering are likely to be larger in more pristine areas--and the patterns of deposition resulting from injecting aerosols into the stratosphere are likely to be different from those due to fossil-fuel burning. Further work is required to elucidate these flux patterns, but it's highly unlikely that they would present a significant problem.
Indeed, I cannot see any scientific reason why the low-intensity geoengineering case could cause detrimental effects of a magnitude that would outweigh the positive effects of geoengineering, the key benefit being significantly more time to develop and deploy the carbon-neutral technologies needed for climate stabilization. Alan has a rosy view of our ability to develop these technologies in a timely fashion, which is at odds with many economists and energy experts. The hurdles are immense, not just politically, but also in terms of technology. (See "Dangerous Assumptions" and "Sustainable Developments: Keys to Climate Protection.") Given the uncertainty of meeting the technological challenges, we should treat geoengineering as a real possibility and meet it head on at a funding level that will bring results and reduce uncertainties quickly.
Although Alan Robock's "20 Reasons Why Geoengineering May Be a Bad Idea" raises legitimate questions, it seems to argue against implementation rather than against studying the underlying science. Few people are actively advocating for immediate, full-scale implementation of geoengineering techniques as a means of addressing climate change. But many people are suggesting that we learn more about the efficacy of such techniques--including Alan, who was recently awarded a National Science Foundation grant to study the effectiveness and possible consequences of injecting aerosol particles into the stratosphere to reduce incoming solar radiation.
In terms of geoengineering concerns, it's helpful to group them into three categories:
Efficacy. Clearly, any geoengineering technique first needs to achieve the intended effect for a reasonable cost--whether the goal is buying time for more sustainable solutions by reducing incoming solar radiation or addressing the root cause of warming by removing carbon dioxide from the atmosphere. It's critical that scientists be allowed to study efficacy through experimentation and modeling without being stigmatized by the assumption that their work will cause a rush to full implementation.
Impact. The environmental impacts of the technique must either be minimal or acceptable relative to the benefits of action and the consequences of inaction. Martin Bunzl makes this point clearly in "An Ethical Assessment of Geoengineering," an accompanying essay on p. 18 of Alan's article. In the case of ocean iron fertilization, 12 small, open-ocean experiments have already been conducted by oceanographers to improve understanding of both efficacy and impacts. (See "Mesoscale Iron Enrichment Experiments 1993-2005: Synthesis and Future Directions".) From the start, it was clear that these experiments weren't a danger to the environment and that their effects wouldn't last long. Scientists should be encouraged to study impacts through experimentation and modeling as long as it can be reasonably presumed that their impacts are short-lived.
Implementation. If a technique is both effective and sensitive to the environment, the following implementation questions become important: Who implements it? Who regulates it? And how do we incorporate these activities into existing regulatory and legal frameworks and treaties? These questions are difficult but not intractable, as many carefully negotiated international agreements already demonstrate, including the International Maritime Organization's London Convention on ocean dumping (signed by 80 countries in 1972, including most of the developed world) and the U.N. Law of the Sea treaty (signed and ratified by most countries except the United States).
More broadly, the provocative title of Alan's article and the quick treatment of individual concerns obscure the complexity behind these subject areas--as Martin and Ken Caldeira have addressed. We expected a summary of research results suggesting a priori that geoengineering is a bad idea, but didn't find one. Also, we found it distracting that many of Alan's concerns (i.e., ozone depletion, acid deposition, effects on cirrus clouds and plants) are specific to one technique--aerosol seeding--but offered as reasons why geoengineering in general is a bad idea. Another of Alan's examples presumes that "humans [adopt] geoengineering as a solution to global warming, with no restriction on continued carbon emissions." No one is suggesting that geoengineering replace emissions reduction.
Most surprising is Alan's conclusion that global warming is a not a difficult technical--but rather purely a political--problem, and therefore, geoengineering isn't required to solve it. We disagree. The road ahead is paved with difficult technical challenges in addition to the considerable political ones. Many new, clean technologies that promise incremental improvements in efficiency also require substantial scientific achievements--such as genetic modification of organisms to make novel substances (i.e., enzymes that process various feedstocks for cellulosic ethanol) or revolutionary advances in materials and process sciences (i.e., new thin-film technologies for solar power). Emission reductions don't simply follow from mandates; we must innovate alternatives to fossil fuels.
That we need to contemplate geoengineering to buy us time for that innovation is unfortunate. That we have the scientific, technical, and human potential to do so responsibly is not.
Winston Churchill once famously said, "Democracy is the worst form of government except all the others that have been tried." Climate engineering may indeed be a bad idea, but so far, better ideas to mitigate global warming show little traction.
We can all agree that eliminating carbon emissions is the right thing to do and that everyone in the world should set aside narrow short-tem self interest and instead work together to provide a better environment for future generations. And when Alan Robock provides 20 reasons why geoengineering may be a bad idea, we can all agree--it is a bad idea. However, what we appear to be achieving in the meantime may be much worse. For all the recent talk about reducing carbon dioxide emissions, the concentration of atmospheric carbon dioxide is growing more rapidly than supposedly pessimistic scenarios predicted even a few years ago.
Preliminary climate model simulations show that in a high-carbon-dioxide world, Earth's climate would be more similar to that of several centuries ago with climate engineering than without it. It won't work perfectly, but imperfection isn't an argument against improvement. The question is whether, in the face of rising greenhouse gas concentrations, climate engineering will improve environmental conditions or merely make things worse. This is an open research question that needs to be vigorously pursued, but an examination of a few of the criticisms on Alan's list demonstrates that so far, none are nonnegotiable:
Continued ocean acidification. Emissions reduction and climate engineering are two levers of action that can be employed jointly or separately. Ocean acidification is a consequence of excess atmospheric carbon dioxide getting dissolved into the ocean, not climate engineering. Climate engineering cannot reverse every adverse consequence of carbon dioxide emissions, but no thoughtful person ever claimed it would.
Ozone depletion. The Mount Pinatubo eruption lofted more than enough aerosols into the atmosphere to compensate for a doubling of atmospheric carbon dioxide, yet ozone concentrations fell by only 3 percent. And it's believed that this small reduction was caused by chlorine from human-made chlorofluorocarbons, which are now banned by the Montreal Protocol. So while the threat to the ozone layer is worth studying in greater detail, it's expected to diminish with time. Furthermore, schemes have been proposed that might preferentially scatter ultraviolet radiation, compensating for any minor reduction in the protection that the ozone provides us from ultraviolet light.
Effects on plants. Alan is correct that we need to study possible effects of climate engineering on plant growth. After the Mount Pinatubo eruption, vegetation everywhere grew more vigorously, taking up more carbon from the atmosphere. This is because diffuse sunlight is able to reach down to enhance photosynthesis in the lower leaves of forest trees, which are normally shaded by the upper canopy in direct sunlight. In general, plant growth responds almost linearly to changes in the amount of sunlight--a 2-percent reduction in sunlight might be expected to produce 2 percent less photosynthesis. But people growing crops in greenhouses often elevate the carbon dioxide level to fertilize their plants, and this effect is typically larger than 2 percent. Therefore, it's possible that a high-carbon-dioxide world with slightly reduced but more scattered sunlight would have higher crop yields than today's world. In computer simulations, vegetation grew more vigorously in an engineered high-carbon-dioxide world than it did in the natural low-carbon-dioxide world. Of course, we can expect these changes to affect natural ecosystems in unforeseen ways, and so should certainly be the subject of intense study.
More acid deposition. The amount of sulfur used in a climate engineering system would be a small percentage of today's emissions from power plants. So if current sulfur emissions regulations were tightened by a few percent when such a system was deployed, there would be no increase in overall sulfur emissions. Furthermore, there's nothing magical about sulfur--other compounds such as silica or calcium carbonate could be used to scatter incoming sunlight, although perhaps at somewhat greater economic cost.
I'll leave Alan's other points for future missives, but the take-home message is, preliminary climate model simulations indicate that climate engineering may mitigate some but not all of the effects of rising greenhouse gas concentrations. While we might prefer near-universal cooperation in carbon dioxide emissions reduction, it's clearly time to plan what we will do if those emissions reductions don't come quick enough or are not deep enough to prevent a climate crisis. The question isn't whether we need to plan for such an eventuality, but what form that planning should take.