Climate change attribution of the compound 2013/14 winter storms flooding in Somerset, UK

The sequence of compounding winter storms of 2013/14 in the United Kingdom (UK) caused a range of significant impacts across the country, totalling an economic cost of approximately £1.3 billion. Heavy rain, totalling 545mm over the season, caused widespread flooding, and coastal impacts were exacerbated by high spring tides and strong winds. The Somerset Levels particularly felt the impact of the flooding, accounting for 30% of the total UK area of flooded agricultural land. This event is a case study for the Horizon-Europe COMPASS project (Compound events attribution to climate change: towards an operational service), where the main goal is to produce a flexible and harmonised methodological framework for such compound extremes, with a focus on impact attribution.

Using the flood model SFINCS (Super-Fast Inundation of Coasts), developed by Deltares, we model the total flood extent for a small region of the Somerset levels for the season. We drive this model with both factual and counterfactual (“natural”, with anthropogenic warming removed) simulations of the winter precipitation, using the HadGEM3-A large-ensemble attribution runs. Initial results demonstrate the magnitude of the observed flood extent to be 1.21 times more likely due to climate change, based on return periods. We also found that a flood event under “natural” forcing but with the same return period as the factual event would be slightly less severe in its extent, 113.40km² compared to 114.02km².

We aim to extend this initial work further both by improving the representation of flood defences, and including the sea level rise component. Attribution of the flood extent to sea level rise can be investigated by removing post-industrial sea level rise from the model boundary conditions to create a counterfactual; this opens another avenue for exploring the compound nature of the event. The impact of the proportion of sea level rise specifically attributable to anthropogenic influence will also be explored.