When looking is dangerous: How solar radiation management research can harm

Vapes are only four percent as harmful as cigarettes, so said a famous 2014 study assessing the health risks of e-cigarettes, the first of its kind. This figure was at best an educated guess. By the authors’ own admission, it lacked hard evidence.

Regardless, that number stuck. In 2015, Public Health England, an executive agency of England’s Department of Health and Social Care, used this journal article to authoritatively say that vaping is 95 percent less harmful than cigarettes, encouraging the use of vapes as a transition treatment off of cigarettes. The maxim “vapes are safer than cigarettes” contributed to the meteoric growth of e-cigarette use, particularly among young people.

It’s now clear that e-cigarettes pose many health risks, that the chemicals in them and their power output have changed over time and are poorly regulated, and that using e-cigarettes seldom results in quitting cigarettes.

Those initial researchers didn’t intend to define British health policy, and they certainly didn’t intend to market vapes to children. The article’s authors simply tried to perform an imperfect risk assessment of a new technology. But such is the nature of research. Ideas, even if initial and incomplete, can cascade into repugnant actions. The path ahead contains a similar flashpoint for a far more impactful set of risky technologies intended to address climate change.

To research, or not to research. Solar radiation management involves deliberately reflecting sunlight back into space to cool the Earth. It’s intended to avoid the worst impacts of global warming and buy time for implementing renewable energies and removing greenhouse gases from the atmosphere. The solar radiation management technique of most interest is stratospheric aerosol injection, which involves using fleets of planes or balloons to inject particles in the atmosphere to slightly whiten the sky.

The rationale for researching such technologies is understandable. Responding to climate change requires as large of a toolbox as possible, and in a hypothetical deployment, decision makers could be helped by understanding solar radiation management risks and how to respond to them. Research can inform effective decision making. The operative word is “can.” Robust analysis and information can help decision-making, but certainly not always. Research can also harm.

However, the risks of research are not always communicated by solar radiation management research advocates. A 2023 open letter in favor of research emphasized transparency and independence, but not the risks of misrepresentation or normalization. Prominent researchers who argued that “climate engineering research is essential to a just transition and sustainable future” similarly do not also flag the risks of research.

And given that research is already underway—sometimes field testing without sufficient oversight (and where the researcher in question simply bemoaned leakers “led astray by diabolical temptation” when contacted by the press)—understanding research risks is especially important. Some types of research can be useful. Others can dangerously mislead.

Welcome to model land. Idealized climate models, in particular, can erroneously misrepresent stratospheric aerosol injection as technology that can be centrally “designed” and easily deployed. Much of the science so far emphasises how a well-designed deployment could lessen the risks of modifying the Earth’s atmosphere.

Yet the design of stratospheric aerosol injection is often Panglossian. The models investigate the impacts of injection if implemented in a consistent manner over decades, sometimes flawlessly coordinated to only use a smaller amount of injections or slowly increase injections focused 15 degrees south of the equator.

In these highly optimistic outlooks lie assumptions that there would be perfect international coordination on how much, how quickly, and where to inject what substance, and that these decisions would be consistent over many decades. Given that this assumption contradicts the foundations of international politics and policy, this imagined version of solar radiation management isn’t of core relevance.

More realistic and messy deployments—driven by the rhyme, rhythm, and chaos of geopolitics and electoral cycles—are of more interest. Continuously presenting an idealized fantasy version of solar radiation management can make it seem like a better option than it actually is.

These aren’t unfounded future-focused concerns. There is a strong track record of idealized modeling damaging climate ambitions.

DICE—Dynamic Integrated Climate-Economy, a prominent climate economic model—assumes that industries which are indoors won’t be affected by climate change, and requires the modeler to put forward a subjective percentage of how much less the future is worth compared to the present. These assumptions result in severe underestimation of climate risks, contributing to delays of climate action. Researchers should be mindful to avoid similar mistakes for solar radiation management.

Models are grainy snapshots of the world. Despite their imperfections, however, models often define the corridors of discussion. Research dominated by idealized, best-case versions of solar radiation management would be a poor guide for handling the complex risks at play.

Policy playhouse. Similar simplifications occur in policy research. Policy research can at times overly focus on analysis of effective policy instruments and structures. But coming up with abstracted policy instruments has limited usefulness unless those instruments can be aligned with a political window for adopting them. Time and time again, initially good policy ideas have been warped and refracted by political voices and political interests, until what is eventually implemented is a shell of its potential.

Carbon markets are, for instance, in principle the most efficient tool to reduce emissions. However, without the political will to create stringent rules and safeguards, carbon markets can actually increase emissions. (This happens if carbon offsets which do not reduce or remove emissions are used to extend or expand fossil activity.)

Pledge and review—the strategy of putting forward non-binding targets for collective review—is another example. This process can be an effective design for international cooperation. The World Health Organization used pledge and review in its successful campaign to combat tuberculosis.

But its use in climate policy has simply become another avenue for countries to drag their feet. There was stronger political consensus with tuberculosis responses than with climate change. That’s because responding to that infectious disease didn’t require grappling with vested interests and energy politics. The Paris Agreement’s pledge and review system bypassed such political firewalls by letting countries do whatever they want, locking in inadequate incrementalism.

The best available analysis of carbon markets and pledge and review couldn’t overcome the political forces that corrupted them. If an idea is good in theory, but ultimately poor in practice, it’s a bad idea.

Simply creating abstracted policy instruments and structures can present solar radiation management as more governable than it actually is. The name of the game is not just developing our understanding of what effective governance instruments or structures would be (the focus thus far). Rather, researchers and policymakers must determine effective governance that would work within a system of political compromises and self-interests.

Complex politics can raise another core objection to research. There seem to be insurmountable political barriers (namely, political and policy consistency over a hundred years) that would make effective endgame deployment dead before arrival. Better climate models and discussions of effective policy cannot fix realpolitik. This would seem to make research a non-sequitur. If a desirable endgame is out of reach, what is the purpose of research?

Slipping away. Simplifying the science and governance related to solar radiation management is especially important given concerns that research is a first step on a slippery slope to eventual deployment.

But some are dismissive of these concerns, saying that a slippery slope is just as likely for other technologies, and that the experience of research in this area thus far has been social resistance and controversy. So, they say, where’s the slippery slope?

These rebuttals miss the point. A slippery slope may be just as likely for other technologies, but other technologies don’t directly control Earth’s climate. Furthermore, just because there is resistance now does not preclude changes in the future, particularly if researchers continue to present a best-case, idealized vision of solar radiation management. Opinion, politics, and policy can, and will, change.

Most important, different people inconsistently use the notion of a slippery slope to mean different things.

The slippery slope is not one single nebulous concept. It’s best understood as the idea that initial actions can spark many different non-linear feedback loops. These feedback loops can be technological, social, political, hype-based (akin to the hype and funding of Theranos or Juicero), or incrementally expand field tests to at scale deployment (though this is less likely for stratospheric aerosol injection).

Worsening risks of a slippery slope to a general deployment of solar radiation management is a disconnect between those who research and those who deploy. Decision makers will not be scientists and researchers, but more likely politicians and government officials. While decision makers could be partially guided by science and evidence, they will almost certainly be driven primarily by political interests. If researchers correctly say that solar radiation management shouldn’t be a replacement for emissions reductions, that doesn’t guarantee solar radiation management will not be used as a substitute for emissions reductions. Researchers could spark feedback loops that will ultimately be out of their control.

But while there are many risks in conducting this research, abandoning the study of solar radiation management might not be the route to take either. There genuinely could be the chance that some research helps inform good policy, reducing the risks of dangerous deployment. For instance, the research community for nuclear weapons was critical in arguing for what eventually became the 1972 Anti-Ballistic Missile Treaty.

However, research certainly is not risk free. Researchers do not passively create a bank of knowledge for the world to draw from. They shape ideas and the way that issues are discussed. Presenting idealized best cases, both in science and policy, can leave the world ironically unprepared for the worst that could come.