Coastal flooding hazard mapping in Normandy: dynamic and static method comparison

International audienceIn France in February 2010, Xynthia storm and its related coastal flooding caused the death of 47 people (41 related to the flooding, 6 related to wind) and nearly 2 billion € of damage. This event highlighted the lack of relevant protection zoning and led to the revision of Coastal Risk Prevention Plans (MEDDE, 2014). In the previous version (MATE-METL, 1997), the flood map of the reference hazard was mainly carried out by static method. This method is controversial because the extreme water level is topographically extrapolated in low areas (Cariolet and Suanez, 2009; Cariolet, 2011; Mercier and Chadenas, 2012; Letortu, 2013) and may deny obstacles to flow such as sand barriers. In addition, in terms of safety, taking into account volumes of overtopping water and flow (acceleration or slowdown) is essential. The new version of the methodological guide recommends the use of dynamic mapping (modeling) when static mapping is unrealistic.Our goal is to identify the most appropriate method (static or dynamic) to map coastal flooding on the Norman sites of Etretat, Pourville-sur-Mer Ver-sur-Mer and Asnelles. The originality of our work is treble: 1) a historical inventory of flooding (1950-2010) enabled to know for each site the reference hazard; 2) the genuine mapping of flooded areas during extreme events was performed, enabling to evaluate whether the simulated methods are close to reality; 3) reliable and accurate topographic data were available (LIDAR RGE Alti®, 2011).To complete the static mapping based on the reference hazard, we bounded flooded areas by vertical threshold of LiDAR data sets (eCognition® and ArcGIS®). As for dynamic mapping, TELEMAC model was used to simulate the progression of overtopping water in these areas sheltered by the sea wall thanks to an extraction of the LIDAR data on a fine mesh (at least 1 m). It enables to observe the channeling of water through roads and non-extensive low areas. The flow of overtopping water at the sea wall, represented as 25 water arrival sources arranged along the defence structures, depends on the breaking wave height (meteorological and marine data observed during the reference hazard) and the morphology of the beach and the sea side sea wall (calculation tool "crossing works", available on the website HR Wallingford).The static mapping has little relevance in the study areas since the seafronts are never affected by the simulated flood, contrary to what was observed in the field. In addition, incoming water volumes are overvalued. The reason is simple: it is not flooding by overflowing but by overtopping, with water volumes pushed by the waves at high tide, which may on the one hand, overtop defense structures, and on the other hand, flood places higher than the sea level at the time by projection and by runoff. The dynamic mapping is more realistic and instructive because it enables:- To visualize the coastal flooding, not only in its result, but also in its temporal evolution;- To highlight the importance of the environmental context of flooding events (in time and location). Indeed, the flood depends on previous overtopping flows responsible for volume of water present on the site (cumulative effects);- To identify the threshold effects of the flood extension;- To indicate the part of the road network which ensures the runoff transfer function, and moves water up to low areas sometimes located far away from the shoreline;- To locate dangerous areas that correspond to the bowls.This exploratory model-approach is promising for risk management but needs improvement such as the inclusion of the buildings that are an important roughness factor