Parjanya Joshi

The aim of this dissertation is to comprehensively study various robustness arguments proposed in the literature from Levins to Lloyd as well as the opposition offered to them and pose enquiry into the degree of epistemic virtue that they provide to the model prediction results with respect to climate science and modeling. Another critical issue that this dissertation strives to examine is that of the actual epistemic notion that is operational when scientists and philosophers appeal to robustness. In attempting to explicate this idea, the discussion turns to arguments provided by Schupbach who completely rejects probabilistic independence in favour of explanatory reasoning, Stegenga and Menon who still see some value in probabilistic independence, and Winsberg who takes applies Schupbach’s to climate science, going beyond models to involve multi-modal evidence. After an exhaustive discussion on these arguments, this dissertation attempts to provide a thorough and updated notion of robustness in climate modeling and climate science.