Potential compound flooding due to pluvial, fluvial and coastal drivers along Eastern North Atlantic and Mediterranean coasts: sensitivity analysis and spatial patterns

Abstract:

Floods are among the most dangerous and costly natural hazards. In coastal regions, floods can arise through a combination of multiple drivers, including storm surge, waves (oceanographic sources), high river discharge (fluvial source) and direct surface run-off (pluvial source). Compound flooding impacts have typically been estimated by quantifying the dependence between extreme surges with either extreme precipitation or river discharge.

In this study, we perform a regional analysis for the European coastlines of the compound flood potential caused by the main four flooding drivers identified using state-of-the-art databases using homogenous forcing (i.e. ERA5). Evaluation of the dependence between multiple drivers is limited to a bidimensional analysis which imposes a conditional sampling (establishing a dominant driver) to select multivariate extremes. First, we analyze how several factors could affect the quantification of compound events: the approach used to select the extreme subsets (annual maxima, peak over threshold), the time window to identify the concurrent events of the secondary driver, the metrics used to describe the compound events or the definition of the wave contribution to sea level.

Besides differences in the regional characterization of the potential compound flooding introduced by these factors, dependence measures are combined to identify different types of multivariate events regarding the predominance of the climate-related drivers and the types of compound extremes which not necessary implies that both variables are extreme. We distinguish several patterns of compound events along the Eastern North Atlantic and Mediterranean coasts which are categorized based on their potential compound flooding risk.