Population in Greenhouse Gas Emissions Projections: Bridging Scenario-Based and Fully Probabilistic Approaches to Representing Uncertainty

Brian C. O'Neill, International Institute for Applied Systems Analysis (IIASA)

Integrated assessments of climate change begin with an assessment of the possible paths for future emissions of radiatively active gases from human activity. Future emissions will be driven by many factors with complex interlinkages and are therefore subject to deep uncertainty. The primary approach to managing this uncertainty has been the use of alternative scenarios conditional on particular storylines about future development trends. This approach has many benefits but does not provide any indication of the likelihood of alternative outcomes. I propose a new approach that may serve as a bridge between scenarios and fully probabilistic methods: conditional probabilistic projections. As an illustration, I produce a modified version of widely used emissions scenarios by incorporating into them conditional probabilistic population projections. The resulting projections provide a fuller representation of uncertainty and an opportunity to investigate the effects of learning over time about how the long-term outlook for emissions may change.