A new approach to navigating uncertainty in climate-related hydrologic risk

Projections of future climate at the local scale are highly uncertain. This is partly because future emissions of greenhouse gases are uncertain (and to a degree unknowable), but largely because different climate models simulate quantitatively and qualitatively different changes in weather for a given forcing. The conventional approach to assessing uncertainty has been to use increasingly large ensembles of scenarios derived from multiple models. At the extreme, the UKCP09 climate projections are based on 10,000 scenarios. Additional uncertainty is added by boundary conditions, downscaling methods, choice of impacts models and model parameterization. The propagating uncertainties that result from these decisions has been conceptualized as the "cascade of uncertainty" (Wilby & Dessai 2010 Weather). It can be challenging to apply such big data to real-world risk assessments and adaptation decisions, and attitudes to confronting this "uncertainty monster" (Van der Sluijs 2005 Water Sci. Technol.) are varied. Navigating the cascade of uncertainty is an overwhelming task (Smith et al 2018 J. Extreme Events), and there are increasing calls for new approaches to organize climate risk information in ways that align better with policy needs (e.g. Kennel et al. 2016 Science).
One such approach develops and uses a small number of 'storylines' to characterise risk (Clark et al. 2016 Curr. Clim. Change Rep.). A storyline is a plausible pathway, without any a priori probability attached. Storylines offer the benefit of allowing an end-to-end quantitative analysis and thereby incorporating compound risk, which is difficult to do within a probabilistic framework in the face of the cascade of uncertainty. Storylines are also easy for lay people to understand, and so provide a natural language for communication in the policy arena.
This project develops the storyline approach, focusing on future drought risk in the United Kingdom. The proposed concept is to navigate the cascade of uncertainty to analyse and bound the system components contributing to climate influence on drought risk, and develop storylines that crystallize that risk (see details in Section 1d). The storylines will be developed to characterize the full range of potential changes in climate that are relevant to drought occurrence (such as change in the frequency of successive dry winters, and delays to the start of the winter recharge season).
UK climate projections 2018 (UKCP18) will contain probabilistic projections, as well as ensembles of simulations, at global, regional, and national scale. In this project, using expertise from supervisors at the University of Reading and the Centre for Ecology & Hydrology, the storyline approach will be applied to these climate projections and propagated through hydrological models. Working with Anglian Water, these results will then be applied to water resources and reservoir yield models in order to stress test current water resource management plans, and develop drought risk assessments. This work has the potential for direct application in policy via drought risk management plans in the Anglian region, and can provide widely transferable methods to help better manage climate change impacts on drought risk across the UK.