Forced trends and internal variability in projections of European windstorms associated with extratropical cyclones

Climate change projections of windstorms associated with extratropical cyclones for Europe are highly uncertain. This is due to differences between models and large internal variability present. Furthermore, year-to-year variations are very high, and the different representations of the driving extratropical cyclones are large, resulting in any forced changes from a warming climate being hard to detect. Windstorms and the associated extratropical cyclones are objectively identified in 20 CMIP6 models, and then Generalized Linear Models and a weighted median estimation are used to extract forced trends for a number of storm impact metrics. Trends are assessed over time, but also as a function of global mean surface temperature changes. Trends in aggregate severity are attributed to changes in storm average severity, frequency, and area impacted, with changes in area being the dominant driver of changes to average storm severity. Using a large ensemble we find that trends between individual members can vary significantly, however the uncertainty due to internal variability is generally 2-3 times lower than model variability. With largest uncertainty coming from model differences, a large proportion of uncertainty in future windstorms is therefore potentially reducible with modelling advances.