Cost-Benefit Analysis of Solar Geoengineering – Garth Heutel, Juan Moreno-Cruz, and Soheil Shayegh


Tuesday, Dec. 15, 2015

First time here?  Read our "what is climate engineering" page.

Solar geoengineering (SGE) offers the possibility of offsetting greenhouse-gas-induced temperature increases by reducing incoming solar radiation. Its key advantages are 1) it is fast. Unlike emissions reductions, which can take decades to bear results, SGE can reduce global temperatures nearly instantaneously. 2) it is cheap. SGE can potentially reduce temperatures at costs that are several orders of magnitude lower compared to abatement – this has been called “the incredible economics of geoengineering”. Against this, it has two key disadvantages. 1) it does not directly address the root cause of climate change – greenhouse gas concentrations – but merely masks its effects, and 2) it comes with a host of problematic side effects – it might deplete the ozone layer or over-dry certain regions of the planet if temperatures are restored to baseline levels.

These advantages and disadvantages of SGE ought to be compared to each other using a cost-benefit analysis. While there has been much discussion about the trade-offs inherent in SGE, there has been surprisingly little quantitative, formal modeling of these trade-offs. Our recent working papers do just that. We modify a widely used integrated assessment model of climate policy, the DICE model, to include the possibility of SGE. With two policy tools (abatement and SGE), the regulators can choose the policy portfolio that maximizes net benefits, taking into account both the advantages and disadvantages of SGE.

As with any quantitative climate modeling exercise, there is a great deal of uncertainty involved, and results should be interpreted cautiously. This is especially true for the incorporation of SGE into the model, since the costs, benefits, and risks of SGE are very poorly understood, even relative to the state of knowledge over climate change in general. Because of this, we attempt to be as “conservative” in our model calibration as possible; that is, we assign very high values of damages and risks of SGE, in order to bias our results against SGE deployment.

Still, under these parameters that are unsympathetic to SGE, we nevertheless find that SGE can play a substantial role in managing climate risks. SGE deployment increases gradually over time and ends up being deployed at up to a 50% intensity – that is, it offsets about one-half of the radiative forcing caused by anthropogenic greenhouse gas emissions. As with the time path of SGE, abatement also increases gradually over time and eventually reaches 100% – we stop emitting greenhouse gases entirely. At this point, SGE, after peaking at around 50%, begins to decline.

With SGE, the abatement rate can be a bit lower, and the date by which all emissions cease is delayed by several decades. It follows that the carbon price necessary to achieve the optimal abatement rate is lower than the carbon price in models that omit SGE. Because SGE is a substitute for abatement, less abatement is required.

Importantly, though SGE plays a substantial role in optimal climate policy, abatement plays a vital role as well because SGE is an imperfect substitute for abatement. While it reduces temperatures (and does so more quickly and cheaply than through abatement), it does not reduce greenhouse gas concentrations in either the atmosphere or the oceans. Damages from climate change can be a function of both the temperature change and the carbon concentrations; even if SGE brings temperatures back to preindustrial levels, if carbon concentrations remain high then we may still face risks. This liability of SGE has been discussed before, but to our knowledge it has never been incorporated into an economic analysis of SGE. We show that when SGE is a less perfect substitute for abatement – that is, when damages from carbon concentrations matter more relative to damages from temperature – then SGE is used less intensively and abatement is used more intensively. As with so much else about SGE, the degree to which damages arise from temperature versus carbon concentrations is not well known and our results are, to some degree, sensitive to the magnitude of the non-temperature climate damages.

Importantly, though SGE plays a substantial role in optimal climate policy, abatement plays a vital role as well because SGE is an imperfect substitute for abatement.

Because uncertainty is vital to the analysis of climate change and SGE, we incorporate it into our analysis by solving a stochastic version of the DICE model, in which there is uncertainty over either climate sensitivity – the degree to which increases in carbon concentrations cause increases in temperature – or the risks of SGE. Under either form of uncertainty, the range in optimal SGE paths is much wider than the range in optimal abatement paths. The optimal rate of SGE can range from hovering around 10% intensity to 150% intensity (causing a negative level of radiative forcing to more quickly bring down global temperatures).

Lastly, we consider how SGE may help to ensure against the probability of reaching a climate tipping point (CTP). CTPs are irreversible, uncertain events that can be triggered by climate-change-induced high temperatures and that cause large damages or disruptions to the climate system. Because SGE can quickly and cheaply reduce or maintain temperatures, they are a vital policy option for reducing the chance of reaching a CTP. The possibility of CTPs can increase the optimal use of SGE by up to a factor of two, depending on the type of CTP. We consider a rule where SGE can only be used after a CTP has been reached – under this scenario, the possibility of a CTP entails greater use of abatement to prevent the CTP from being reached.

Enormous uncertainties about the costs, benefits, and risks of solar geoengineering are all too present in this analysis. Nonetheless, our research represents a crucial first step in understanding how SGE can be part of an optimal climate policy portfolio, and how it interacts with other climate policy options.

Garth Heutel 125Garth Heutel, PhD is Assistant Professor in the Department of Economics, Andrew Young School of Policy Studies, Georgia State University and a Faculty Research Fellow at the National Bureau of Economic Research.  He studies energy and environmental policy, behavioral economics, public economics, and the economics of nonprofit organizations.  His research has been published in theJournal of Public Economics, The Economic Journal, American Economic Journal: Economic Policy, Journal of Environmental Economics and Management, Review of Economic Dynamics, and elsewhere.

Juan Moreno-Cruz 125Juan Moreno-Cruz, PhD is an Assistant Professor in the School of Economics. He has a PhD in Economics from the University of Calgary and a B.S. and M.S. in Electrical Engineering from the University of Los Andes in Bogota, Colombia.  Moreno-Cruz’s research focuses on the interaction of energy systems, technological change, and climate policy. Moreno-Cruz has investigated how technologies designed to modify the climate affect the strategic interaction among nations. Currently, he is developing a new set of theoretical and empirical tools to study energy-system transitions in order to inform energy and environmental policy.

Soheil Shayegh

Soheil Shayegh, PhD is a Postdoctoral research scientist in the Department of Global Ecology at the Carnegie Institution for Science.



The Forum for Climate Engineering Assessment does not necessarily endorse the ideas contained in this or any other guest post. Our aim is to provide a space for the expression of a range of perspectives on climate engineering.