Geoengineering poses risks that combine natural and social aspects. For example, will stratospheric aerosols destroy ozone? Will the availability or implementation of geoengineering prevent sustained action to mitigate climate forcing? Here we focus on the technical risks, and defer consideration of social risks to the following section.
The biogeochemical risks differ markedly for the two principal classes of geoengineering strategy - albedo modification and CO2 control. For each class, risks may be roughly divided into two types: risk of side effect and risk that the manipulation will fail to achieve its central aim. For albedo modification, the division is between side effects such as ozone depletion, that arise directly from the albedo-modifying technology, and risk of failure associated with the difficulty of predicting the climatic response to changes in albedo. Side effects of CO2 control include loss of biodiversity or loss of aesthetic value that may arise from manipulating ecosystems to capture carbon, and risk of failure is associated with unexpectedly quick re-release of sequestered carbon.
The risks posed by geoengineering are sufficiently novel that, in general, the relevant biological and geophysical science is too uncertain to allow quantitative assessment of risk. Absent quantitative assessment, various avenues remain for robust qualitative risk assessment, for example, if a geoengineering scheme works by imitating a natural process we can compare the magnitude of the engineered effect with the magnitude and variability of the natural process, and then assume that similar perturbations entail similar results (Keith and Dowlatabadi, 1992; Michaelson, 1998). For example, the amount of sulfate released into the stratosphere as part of a geoengineering scheme and the amount released by a large volcanic eruption are similar. We may estimate the magnitude of stratospheric ozone loss by analogy.
In decisions about implementation, judgment about the risks of geoengin-eering would depend on the scalability and reversibility of the project. Can the project be tested at small scale, and can the project be readily reversed if it goes awry? These attributes are vital to enabling management of risk through some form of global-scale adaptive ecological management (Gunderson, Holling et al., 1995; Allenby, 1999). Even crude qualitative estimates of risk can give insight into the relative merits of various geoengineering methods when considered in conjunction with other variables (Keith and Dowlatabadi, 1992).
We have examined the risk of geoengineering in isolation. More relevant to real choices about planetary management is a comparison of the risks and benefits of geoengineering with those of other response strategies. Here we are in unexplored territory as the literature has largely avoided this question. Without attempting such a comparison, we note that it would have to be explicit about the goals; i.e., is geoengineering a substitute for abatement, an addition to abatement, or a fallback strategy? Also, it would have to assess the risks of abatement or adaptation per se.
Was this article helpful?