Assessing the Sensitivity of a Coastal Flood Model for the Development of a Coastal Flood Awareness System at European Scale.
Author(s): Marine Le Gal; Tomas Fernandez Momblant; Paolo Ciavola
Linked Author(s): Marine Le Gal, Paolo Ciavola
Keywords: Coastal flood; Sensitivity analysis; Lisflood-fp
Abstract: The European Copernicus Coastal Flood Awareness System (ECFAS) project aims to deliver a proof of concept of a coastal flood alert system for an improvement of the Copernicus Emergency Monitoring Service. A flood map catalogue for European coast derived from numerical models will be created. In order to generate such a catalogue at pan-European scale, it is necessary to optimize the development of the flood models to find a balance between robustness and feasibility according to the project’s scope. These decisions concern the model’s parameters such as the grid resolution, the numerical solver and the formulation of the wave contribution used for the boundary condition with the available data. To support these choices, it is important to identify which are the most influential parameters on the performance of the models and to quantify their impact on the models’ output. In the ECFAS project, 17 test-cases were selected for which LISFLOOD-FP models are being developed covering sites across Europe and different storm events. It gives a panoply of local morphologies and boundary conditions derived from the storm characteristics, i.e. duration of the storm and total water level. In addition, a few selected physical and numerical model’s parameters were tested and varied giving more than 45 models per test-case. These parameters are the grid resolution and the numerical solver, the ground friction, and the wave set-up formulation used to estimate the total water level necessary as a boundary condition. A sensitivity analysis was performed using a one-driver-at-a-time method on the maximum flooded areas and water volumes. The differences between models when changing a single parameter were computed and then averaged for each parameter. The results of this sensitivity analysis allow us to identify patterns through the test-cases correlated to geographical region and provide vital information for the calibration the flood models across Europe. In the present analysis, the floodplain solver was the parameter that most strongly influenced on the Atlantic coasts while it was the wave set-up formulation on the Mediterranean coasts.
How to cite: Marine Le Gal, M.; T. Fernandez Montblanc; P. Ciavola (2022)
Assessing the Sensitivity of a Coastal Flood Model for the Development of a Coastal Flood Awareness System at European Scale. Proceedings of the 39th IAHR World Congress (Granada, 2022), Doi: https://doi.org/10.3850/IAHR-39WC252171192022986