Performance of Bayesian Learning Applied on the Climate Sensitivity in EconomicClimate Modeling

Abstract

In the last years economic climate modeling has attracted a lot of attention because of its fundamental interest and applications to policy-making. The most famous economic climate model is DICE, made by William Nordhaus. In his model he assumed that the climate sensitivity is constant. Motivated by the results of the IPCC reports, which over the years show big deviations in estimates of the climate sensitivity, this thesis develops a framework to model the economy and the climate with uncertain climate sensitivity. This is done by implementing Bayesian learning, the results in this thesis show that the framework works and that Bayesian learning has a signi?cant impact on the welfare. It is found that Bayesian learning will enhance the performance of a policy by 476%.