A study of how to increase precision in the categorization of deep uncertainty, and how to assess risk under a specific level of deep uncertainty

Abstract

Many of the recently published articles that try to resolve challenges related to deep uncertainty have based their understanding of deep uncertainty and the various nuances of uncertainty on Courtney’s uncertainty taxonomy. This thesis provides some reflections on some of the foundational pillars this taxonomy is built upon. In doing so, some major challenges and limitations are uncovered. To overcome these challenges and limitations an alternative uncertainty taxonomy is introduced. This taxonomy is built upon the same template as the one used by Courtney, but it contains a higher level of detail, and an additional level of uncertainty. The new level of uncertainty, which covers the transition from moderate to deep uncertainty is added to make sure that every nuance of uncertainty, ranging from low to deep uncertainty is reflected. Later on, a method that can assess risk under the new level of uncertainty is introduced. This method is an adaptation of a regular risk assessment process combined with a probability bounds analysis (PBA) and a qualitative judgement of the assumptions made in the analysis. The PBA form the quantitative basis of the assessment, while the qualitative judgement of the assumptions is used to justify whether the final result of the PBA can be trusted or not. By applying this method to a hypothetical case, it proves itself to be a good tool for assessing risk in cases where the empirical data is limited.