Numerical modeling of extreme events in the Tagus estuary

Abstract

The risk of submersion of estuarine and coastal margins is increasing. The hazard is growing due to sea level rise, declining natural protection associated to coastal erosion and, in some areas, higher storminess. Simultaneously, vulnerability is rising due to the migration of human activities towards the coast. This trend fuels our need to better quantify the hazards and understand their sources. The inundation extent in coastal areas can be determined with numerical models, and two alternative approaches are followed. The first involves simulating past extreme events (Method A). Provided the model is adequately validated, this approach can complement field surveys, provide insight into the importance of different processes, determine the impact that previous events would have under present conditions, etc. However, this approach does not provide information on the return period of these events. In the second alternative, the simulations are preceded by a statistical analysis of the forcing agents, so each simulation is associated to a known return period (Method B). The drawback of this approach is the simplification of the forcings, which may introduce errors difficult to quantify. This paper describes an ongoing modeling analysis to understand and characterize inundation in the Tagus estuary. We aim at: 1) comparing the two approaches outlined above; 2) understanding the importance of different drivers; 3) determining the inundation extent of past events with present mean sea level and bathymetry. Method A is based on the analysis of newspaper records that indicated three major events: the storm of February 1941, the event of January 1996 and the Xynthia storm in 2010. Method B is based on a statistical analysis of water levels at Cascais.