Recent concern over rising quantities of carbon dioxide in the atmosphere and the accompanying "greenhouse effect" has prompted calls for caps on, or even reductions in, the overall output of carbon dioxide from fossil fuels burned to produce energy for the world economy. This project will investigate the effects of modeling investment uncertainty in a global energy-economic interaction with varying limits on carbon emissions in different regions of the world. The project will develop a model that assumes long run economic growth maximizing decisions are taken. The consistency of this objective with current economic decisions will have implications for policies aimed at encouraging new technology investment, exploration for undiscovered resources, and energy saving substitutions. The overall goal will be to determine the effect that investment with uncertain outcomes might play in limiting expected economic costs of carbon emission policies. The model will be built as an extension of Manne and Richels' Global 2100 model that has already considered economic effects of carbon limits without explicit random investment effects. These effects will be incorporated in the proposed model that will take the form of a stochastic nonlinear program. The solution methodology will be a distributed decomposition scheme that runs on a network of powerful computer workstations. New bounding techniques for expectations with limited information will be developed to bound the value of information and the value of solving the complex stochastic model.