Learning about Climate Change and Implications for Near-term Policy
Mort Webster¹, Lisa Jakobovits², and James Norton³

¹ MIT Joint Program on the Science and Policy of Global Change, Massachusetts Institute of Technology
²Technology and Policy Program, Massachusetts Institute of Technology
³ Department of Statistics and Operations Research, University of North Carolina at Chapel Hill

Climate change is an issue of risk management. The most important causes for concern are not the median projections of future climate change, but the low-probability, high-consequence impacts. Because the policy question is one of sequential decision making under uncertainty, we need not decide today what to do in the future. We need only to decide what to do today, and future decisions can be revised as we learn more.
To answer the question of what we should do today, conditional on what we expect to learn, it is useful to decompose this into two distinct questions. The first question is: how much do you need to learn and by when for today’s policy to differ from what you would do without learning? The second question is: how much and by when can we expect to reduce uncertainty about climate sensitivity by observing temperature change? After addressing each of these in turn, we can synthesize the results to determine whether expected learning about climate should affect today’s policy stringency, and in what way.
In this study, we use a stochastic version of the DICE-99 model (Nordhaus and Boyer, 1999) to explore the effect of different rates of learning on the appropriate level of near-term policy. We show that the effect of learning depends strongly on whether one chooses efficiency as the criterion for policy choice (balancing costs and benefits) or whether one chooses cost-effectiveness (stabilizing at a given temperature change target). Then, we calculate posterior distributions of climate sensitivity from Bayesian updating, based on temperature changes that would be observed for a given true climate sensitivity and assumptions about errors, prior distributions, and the presence of additional uncertainties. In the case of a temperature target, we show that observations of temperature alone are unlikely to reduce the uncertainty enough in the next few decades to justify a different level of emissions reductions than if we did not expect to learn more.

Responsible for this page: Katja Scherbov
Last updated: 12 Feb 2007