Current and future risks of unprecedented UK droughts

The UK has experienced recurring periods of hydrological droughts in the past and their frequency and severity are predicted to increase with climate change. However, quantifying the risks of extreme droughts is challenging given the short observational record, the multivariate nature of droughts and large internal variability of the climate system. We use EC-Earth time-slice large ensembles, which consist of 2000 years of data each for present day, 2°C and 3°C conditions, to drive the GR6J hydrological model at UK river catchments to obtain a large set of plausible droughts. Applying the UNSEEN (UNprecedented Simulation of Extreme Events using ENsembles) approach show an increasing chance of unprecedented dry summers with future warming and highlight the chance of an unprecedented drought with characteristics exceeding that of past severe droughts.

This study also aims to bridge the probabilistic UNSEEN approach with “bottom-up” storyline approaches. Physical climate storylines of preconditioned compound drought events are created by searching within the large ensemble for events resembling specific conditions that have led to past severe droughts and are relevant for water resources planning. This includes conditions such as 1) dry autumns followed by dry winters, 2) consecutive dry winters (both of which are relevant for slow-responding catchments), and 3) dry springs followed by dry summers (relevant for fast-responding catchments). The storylines can be used to understand the conditions leading to unprecedented droughts and the impacts of future droughts triggered by the same conditions. Unprecedented drought sequences and synthetic experiments conditioned on these storylines can be used to stress-test hydrological systems and inform decision-making.