Addressing ‘deep’ uncertainty over long-term climate in major infrastructure projects: four innovations of the Thames Estuary 2100 Project

摘要

Climate change brings new challenges for traditional environmental risk management, particularly for major infrastructure projects, where the decisions made today can have long-term implications. A major challenge is that projections of future climate are deeply uncertain. If this uncertainty is not managed appropriately, long-lived infrastructure may need to be replaced or expensively retrofitted before the end of the design lifetime. The Thames Estuary 2100 Project (TE2100) was one of the first major infrastructure projects to explicitly recognise and address the issue of the deep uncertainty in climate projections throughout the planning process. In this paper, we discuss the innovations and techniques that were adopted in this case that enabled it to cope with this uncertainty. We classify the overall approach as ‘dynamic robustness’, which aims to build flexible strategies that can be changed over time as more is learnt or as conditions change. The project combined four key innovations: (1) the ‘decision-centric’ process; (2) the combination of numerical models and expert judgement to develop narrative sea level rise scenarios; (3) the adoption of an ‘Adaptation Pathways’ approach to identify the timing and sequencing of possible ‘pathways’ of adaptation measures over time under different scenarios; and (4) the development of a monitoring framework that triggers defined decision points. A secondary focus of this paper is an exploration of how climate information was used in TE2100. We suggest that the techniques employed in TE2100 imply different needs from climate science that may cause them to redefine their research priorities related to adaptation; namely a shift in emphasis away from probabilistic modelling, toward greater investment in observations and monitoring; improved understanding of historical climate variability; and improved understanding of physical Earth system processes and their representation in models to enhance ‘best guess’ models and to better bound future projections using narrative scenarios.