Statistical analysis of the extreme values of dynamic sea level by spatial interpolation for a beach of the Mediterranean French coast
Abstract
The knowledge of sea levels along the coastline is of the utmost importance for characterization of flood risks in coastal areas, with a particular interest devoted to extreme values because they may induce the most dramatic consequences.The Cerema is a new French public body in support of national and local authorities in the field of sustainable development. At the request of French authorities, the Cerema has studied with a statistical approach the extreme values of the dynamic sea levels for one beach of the Mediterranean coast in Southern France.The beach is located at Saint-Aygulf, between Toulon and Nice. This site is critical because it's a tourist place with some buildings closed to the sea.The dynamic sea level studied includes a predictive part, the tidal level with about 40 cm of tidal range, and a non predictive part due to meteorological effect (difference of atmospheric pressures, wind effect) and breaking wave (wave run-up).There is no data of sea level measurement or numerical simulation at Saint-Aygulf. The development of a model to compute numerical simulations is out of the scope of this study.The closest tide gauges are located at more 50 km from Saint-Aygulf, in Toulon and Nice, with more than 15 years of observations.The originality of this work is to transform data from Toulon and Nice, so that the estimations of the dynamic sea level with these transformed data are representative of Saint-Aygulf. The final result is a weighted mean of both estimations (weight inversely proportional of the distance).The wave run-up is computed with the Stockdon et al. [2006] formula. Wave data come from ANEMOC2 data base (hindcast simulations from Cerema and EDF R&D). The dependence between offshore sea-states and tide gauge measurement is modelled by a Gumbel copula. Data transformation of Toulon and Nice takes into account the specific conditions of wave climate at Saint-Aygulf (exposure and energy loss during the propagation).As specified by the theory, the dependence model (Gumbel copula) is defined if and only if values of wave and still water level (tide-gauge record) are simultaneously extreme. It's why we introduce a low and a high limit in order to border the final result.At the end we discuss about the limits of the method and we give some ideas to improve it.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2016
- Bibcode:
- 2016AGUFMNH31A1896K
- Keywords:
-
- 1641 Sea level change;
- GLOBAL CHANGEDE: 4313 Extreme events;
- NATURAL HAZARDSDE: 4328 Risk;
- NATURAL HAZARDSDE: 4546 Nearshore processes;
- OCEANOGRAPHY: PHYSICAL