Can We Test Geoengineering Without Potential Consequences?

A very recent piece in Science, Robock, et al., A Test for Geoengineering?, 327 Science 530-31 (2010) provides an excellent critique of why one of the chief claims of proponents of solar radiation management geoengineering (most prominently, injecting sulfur dioxide or other particles into the stratosphere to reflect incoming solar radiation), that we can start experimentation on small scale and assess the results, may be chimerical. The authors of the article argue that there are two reasons why it might not be useful to begin with small rates of insertion of particles to assess the potential negative impacts of stratospheric injection. First, if one wished to produce an aerosol cloud of sufficient thickness to cool the Earth’s surface, one would need to engage in regular injections. However continuous emissions of sulfur particles or gasses would cause existing particles to grow, reducing albedo effects. As a consequence, even more injections would be necessary, meaning that such effects couldn’t be assessed except at full scale deployment. Additionally, “the signal of small injections would be indistinguishable from the noise of weather and climate variations.” The only way to distinguish the signal from noise would be to inject very substantial amounts of particles for a protracted period of time.

The article also demonstrates why programs of this nature, while touted by proponents as reversible, might have a life of their own. First, the cessation of geoengineering could result in a huge temperature spike, meaning that we might not be able to turn back unless we simultaneously engage in aggressive emissions reductions. Moreover, “the geoengineering infrastructure” would lobby to maintain its program, backed by those with an interest in the program, including those who might stand to profit or have obtained jobs.

Finally, the article argues that one of the consistent concerns expressed about shortwave radiation engineering is potential regional effects, e.g. disruption of monsoon circulation in tropical regions. Unfortunately, the more localized that these impacts might be, the longer experiments would have to be run to assess such potential impacts. Even a 10-year experiment, the researchers conclude, would not necessarily detect local adverse responses.

Geoengineering is an excellent topic to cover in a climate change course, or a more general environmental course. It facilitates discussion of critical issues such as risk assessment, the proper role of the precautionary principle (could it be, for example, that geoengineering is arguably actually MANDATED if there is a risk of catastrophic climatic impacts under business as usual scenarios; see the language of Rio Principle 15, for example), the potential for moral hazards in climate decision making, and issues of equity and how to ensure that governance structures effectively incorporate such concerns. The Robock piece would provide a good jumping off point for such a classroom discussion.

Be Sociable, Share!