C-ROADS Climate Model
MIT system dynamicist John Sterman, who gave a plenary on climate policy modeling at INFORMS 2009, now has a podcast about that topic on the INFORMS Science of Better site. A good deal of the podcast serves as general background on climate change projections, international accords, and the concept of cap and trade. Sterman and his colleagues developed a climate policy model called C-ROADS, available from the website climateinteractive.org. C-ROADS models the impacts of policies (such as ones discussed at the recent Copenhagen meetings) on climate change. The base case of the model uses the IPCC assumptions and outputs global concentrations of CO2, methane, and more. There is a freeware version of the model for non-specialists (from Senators to middle school students) to play with. The idea is that in using the model, people will attain a better feel for the relationship between policy and impact on climate, not unlike the way flight simulators are used.
Large models like these have numerous inputs and a large degree of uncertainty in both the data and the assumptions being made. This is not unlike the life cycle assessmemt (LCA) models. And so like the LCA case, there is a danger of obtaining vastly different results under the same policies, because of different assumptions (see this post for more on the LCA version of this issue). Sterman emphasizes that C-ROADS is meant to be a “sensitivity tool”, not a precise point forecasting model. He points out that it makes the uncertainties and assumptions visible to the model user as well as adjustable. So a form of sensitivity analysis can be conducted by changing these levels and analyzing the impact.