Abstract. Winds, waves and storm surges can inflict severe damage in coastal areas. In
order to improve preparedness for such events, a better understanding of
storm-induced coastal flooding episodes is necessary. To this end, this paper
highlights the use of atmospheric downscaling techniques in order to improve
wave and storm surge hindcasts. The downscaling techniques used here are
based on existing European Centre for Medium-Range Weather Forecasts
reanalyses (ERA-20C, ERA-40 and ERA-Interim). The results show that the 10 km
resolution data forcing provided by a downscaled atmospheric model gives a
better wave and surge hindcast compared to using data directly from the
reanalysis. Furthermore, the analysis of the most extreme mid-latitude
cyclones indicates that a four-dimensional blending approach improves the
whole process, as it assimilates more small-scale processes in the initial
conditions. Our approach has been successfully applied to ERA-20C (the
20th century reanalysis).