Description
Summary:International audience Storm surges are often characterized in terms of magnitude, duration and frequency. Here, we propose a novel statistical method to help characterize the full dynamics of storm surge events. The method, called ECHAR, is based on techniques already successfully applied in astrophysics. Analysis of 20 tide gauges in the North-East Atlantic consistently reveals that storm surge events display two distinctive components, a slow-time background Gaussian structure and a fast-time Laplace structure. Each of these structures can be reduced to its duration and amplitude. For large events, occurring 5 times per winter, the slow-time structure lasts around 16 days, varying from 9 days in the South to 45 days in the North (Baltic Sea), with almost the same amplitude at all the stations (around 0.17 m). The fast-time structure lasts around 1.7 days at all the stations, but its amplitude greatly varies, from 0.1 m in the South to 1.6 m in the North Sea. The wind stress contributes mostly to the fast-time component of the storm surge event, whereas the atmospheric pressure contributes to both components. The proposed ECHAR method, helping to characterize extreme events, can be applied anywhere else in the global ocean, e.g. where tropical storm surges occur.Key PointsA new method ECHAR is proposed to characterize the dynamics of typical storm surge eventsStorm surge events display a slow-time and a fast-time component lasting about 16 days and 1.7 days respectivelyThe wind stress mostly contributes to the fast-time component whereas the atmospheric pressure contributes to bothPlain Language SummaryStorm surges are an increase of the sea level, due to low atmospheric pressure and strong winds during storms. We propose a new method, to characterize storm surge events in the North-East Atlantic. We consider the largest events, that happen only 5 times per winter. A typical storm surge event is a gradual slow increase and then decrease of the water level, over a period of few days to few weeks, from 9 days ...