The Geopolitical Challenges of Geoengineering—and Geoengineering’s Challenge to Geopolitics

In 2006 the atmospheric chemist Paul Crutzen, one of the world’s most respected atmospheric scientists, published “A contribution to resolve a policy dilemma” in the journal Climatic Change. The policy dilemma was this: the burden of respiratory disease associated with particulate emissions from coal-fired power stations was apallingly high, killing perhaps 500,000 people a year. Regulators planned to reduce this significantly by cleaning up power stations. As a result, anthropogenic sulphur emissions, then around 55m tonnes a year, would decline significantly over the coming decades.

As well as damaging people’s lungs, though, the aerosol particles which sulphur emissions produce in the atmosphere also reflect sunlight back into space before it can warm the Earth’s surface. The cooling due to such pollution, Crutzen estimated, was depressing the global temperature by up to one degree celsius, offsetting a significant fraction of the damage being done by anthropogenic greenhouse gas emissions. Remove those sulphur emissions, and the world would heat up more quickly than if you kept them in place. The policy dilemma: actions to reduce harm in the short term by curtailing respiratory disease would increase harm in the medium term by unleashing the full impact of greenhouse gas warming.

Crutzen suggested that the loss in cooling when sulphur emissions into the lower atmosphere were reduced could be made up for by releasing sulphur into the stratosphere. Because sulphate particles last much longer in the rainless, snowless stratosphere than in the weather-riled lower atmosphere, the amount of sulphur needed would be small compared to existing emissions: just a million tonnes or so a year, well within the technical capacity of a small fleet of aircraft.

Crutzen’s paper paved the way for a growing interest in solar geoengineering: the use of technology to reduce the amount of sunlight that the earth absorbs. And in the years since its release, researchers have used computer models to produce hundreds of papers modeling solar geoengineering’s climatic effects, technological implementation, adverse impacts, and compatability with notions of justice and democracy.

This climate model-based research shows that solar geoengineering could never simply cancel out anthropogenic temperature increases, let alone other climate change impacts. If carried out with stratospheric sulphur, for example, it would slow down the ozone layer’s recovery from the damage caused by industry in the second half of the 20th century. It would also have a potentially profound effect on the hydrological cycle. A world in which all greenhouse warming after a certain date was offset by solar geoengineering would experience significantly less precipitation than one in which there was neither greenhouse warming nor geoengineered cooling.

However, all in all, the results of this research have increasingly come to suggest that in a world where greenhouse gas emissions peak some time between now and mid-century and then dwindle—which is to say, our current world—a solar geoengineering program that offset a signicant fraction of the resultant warming could greatly reduce overall harm. The benefits of that moderating effect would vary from region to region, but models suggest that there may well be levels and patterns of intervention where even the regions in which the benefits were found to be lowest and/or the harmful side effects highest would see little or no net harm. “Showstoppers” that cast doubt on the whole idea have, from the point of view of natural science, failed to appear.

A (Political) Science Experiment

It has thus become commonplace within the geoengineering research community to hold the opinion that the most substantial risks posed by solar geoengineering are not biogeophysical, but matters of international relations. These risks can be divided into systemic risks, which are geopolitical in the broad sense, in that they bear on international processes of decision-making, and scenario risks, which are geopolitical in the more specific sense of depending immediately on matters of territory and military capability.

The major systemic risk is the erosion of mitigation efforts. This does not depend on anyone actually doing solar geoengineering, just on people being willing to talk about and research the possibility. The insidious nature of the risk is illustrated by the role in recent climate negotiations of solar geoengineering’s speculative sibling, carbon geoengineering—the active removal of carbon dioxide from the atmosphere.

It was by tacitly invoking carbon geoengineering that the Paris Agreement of 2015 was able to succeed where previous climate diplomacy had failed. Previous climate negotiations had been based, implicitly or explicitly, on the idea that the eventual level of global warming would depend on the total amount of long-lived greenhouse gas emitted into the atmosphere. A given temperature limit would imply a specific total amount of emissions reductions required and the rate at which emissions should be cut. Major emitters, however, were not able, or willing, to promise cuts on the scale needed for the then widely accepted limit of a two degree temperature increase.

Paris squared this circle in two ways. The more obvious was that it set up a “ratcheting mechanism” whereby policies to cut emissions would get more and more ambitious over time. The second and subtler method was to open up the possibility of “negative emissions.” The stable level of long-lived greenhouse gases that it imagined for the second half of the 21st centuy would not be brought about solely by an absence of anthropogenic emissions, but by the advent of technologies or policies that remove already-emitted carbon dioxide from the atmosphere. If positive and negative emissions are in balance, you have the world of “net-zero” emissions that the agreement talks about. If negative emissions come to predominate, you have a world of reducing greenhouse gas levels in which temperature will, after a certain degree of lag, actually come down. Invoking this sort of carbon geoengineering allows one to draw lines on graphs that connect today’s modest emission cuts with stable and acceptable temperatures at the end of the century.

There are various plausible carbon-geoengineering technologies by means of which this might be done. But at present they do not exist at any real scale, and efforts aimed at their creation and widespread acceptance are, on a global scale, nugatory. The Paris Agreement’s squaring of modest emission-reduction pledges with ambitious future-temperature targets thus rests on what is in effect a fantasy (abeit one that could, in principle, be actualized). To make matters worse, the fantasy is fungible. Once you have allowed the idea of negative emissions into your thinking it is always possible, at the margin, to trade a little less emissions-reduction today for a little more negative emission tomorrow. Over time, there is nothing to stop a little from becoming a lot. The promise of geoengineering becomes a license for procrastination.

This is true regardless of whether that promise is made sincerely or cynically, or whether the geoengineering in question is carbon or solar. Indeed, the fact that both solar geoengineering and carbon geoengineering decouple climate outcomes from cumulative emissions is one of the reasons why I choose to see them as similar enough to carry the same name, despite the deep differences in the way they act and the effects they can produce. Treating either solar or carbon geoengineering as a medium-term possibility reduces the incentives for near-term emissions reduction.

This problem has been much discussed in the solar-geoengineering literature, often under the rubric of “moral hazard.” The way in which the negative-emissions discussion has evolved before and after Paris strongly suggests that the risk is real: building the assumption of solar geoengineering into the policy framework really can be used to legitimize and valorize levels of effort not otherwise remotely sufficient to meet stated climate goals.

In demonstrating this, though, Paris may also have shown that the risk no longer matters. We already live in a world in which people do too little about climate change, in part because they are relying on notional future geoengineering to bail them out; it is just that the non-existent safety net is being provided by carbon geoengineering rather than solar geoengineering. It is not obvious that more serious discussion of solar geoengineering would further undermine current action on climate change.  

Invidious Intervention?

Actual solar geoengineering, though, undoubtedly would. This is in part because the biogeophysical effects of solar geoengineering are strongly scenario dependent. Imagine a deployed solar-geoengineering strategy aimed at limiting warming to two degrees while ensuring that no country or region faces unacceptable hydrological distress. The amount of cooling required will depend on the emissions pathway the world follows. If the act of starting the intervention changes the way nations act with respect to emissions, that pathway changes in turn—and thus so does the amount of solar geoengineering required. It is easy to imagine such a strategy soon finding itself in a world where its purported aim—capping temperatures but not causing hydrological distress—is no longer achievable. At this point the interests of nations in different regions, once plausibly aligned, diverge.

This is a special case of a general geopolitical problem: strategic stability. As solar geoengineering research quickened post-Crutzen, it was widely assumed that even if a strategy’s net benefits were large, the workings of the climate system would ensure that they were sufficiently heterogeneous that some regions would end up worse off even if most others benefitted—in other words, that there would be winners and losers. Yet one of the most striking results to come out of solar geoengineering research has been the discovery that, in models, there seem to be strategies in which a moderate amount of solar geoengineering leaves no nation worse off than it is today.

There is a catch. A strategy in which no region loses will also be one in which some regions do less well than they would under some other strategy. Some nations will thus always have an interest in strategy change. This builds an asymmetric instability into the system that will tend to increase the amount of solar geoengineering beyond that originally seen as optimal.

Imagine a world in which a consortium of nations is injecting aerosols into the stratosphere in order to provide an optimized effect acceptable to all. Now imagine that one of those nations decides, in the light of how things play out on the ground, it wants less cooling. Unless it can convince all the other nations to agree, it is stuck; if it stops emitting aerosols unilaterally, the other nations can easily make up the deficit, keeping to the originally specified strategy. Reducing cooling (i.e., increasing warming) will always require consensus.

More cooling, though, does not. If Nation A wants more cooling, it merely needs to inject more aerosols. If the other nations reduce their contributions to try and keep things on an even keel, Nation A can decide to inject even more. If all the other nations stop injecting, Nation A can take on the whole task. This is what Marty Weitzmann and Gernot Wagner have called the “free driver” problem. The amount of solar geoengineering carried out will be a function of the level of intervention preferred by the nation with the greatest desire for cooling and the willingness of other nations to impose a price for that cooling that the first nation finds unacceptable.

A New Geopolitical Phenomenon

Thus solar geoengineering depending on stratospheric aerosols offers a clear possibility of international conflict even among nations that, in principle, are all open to its deployment. This is because solar geoengineering is a global geophysical phenomenon subject to some level of national political influence—a class of phenomenon hard to embrace within current conceptions of geopolitics. The “geo” of geopolitics is, after all, fundamentally that of the geographer’s map; a heterogeneous two-dimensional surface divided up among players whose interactions are encouraged or constrained by their relative sizes and positions. The “geo” of geoengineering is the geo of geophysics and biogeochemistry; an earth system in constant flux. This Earth system is in some ways deeply non-local; carbon dioxide emitted from any given source is like that from every other, for example. In other ways it is defiantly local; the climatic state, and direction of change, differ from place to place.

A geopolitics, and system of international relations, conceived on the basis of nations that hold territories, is a poor fit to the political demands of a unified planetary system defined by the flows of energy and material through unending climatic and biogeochemical cycles. Many of the problems that have dogged the development of climate diplomacy can be seen as direct or indirect consequences of that poor fit. But they are not necessarily perceived as such, because the concepts through which the problems are discussed—the natural climate, the economy—do not feel fundamentally new.

Solar geoengineering, on the other hand, is new: a global geophysical phenomenon subject to some level of national political influence that does not exist in any form today, but could tomorrow. It thus sheds a unique light on the degree to which a geopolitics that grew up in a world of great powers can be reconciled with an Earth system of planet-spanning flows, and on the challenges that such a reconciliation will entail. Solar geoengineering raises problems within geopolitics. It also brings into focus problems with the whole idea of geopolitics.