Innovative absolute probability approach for coastal vulnerability assessment due to sea level rise: Application in Tagus Estuary, Portugal

In coastal urban areas highly susceptible to flooding, whether from sea level rise (SLR) or storms, it is crucial to assess the vulnerability and risks posed by extreme and frequent floods. Reliable estimates of extreme natural events' return periods rely on historical data or probabilistic models, requiring extensive and robust data. From climate-scenario-based or semi-empirical models, SLR projections are represented by a central estimate or the full domain cumulative density function (CDF), entailing uncertainties. Each scenario or projection having its own CDF results in relative probability solutions rather than absolute probability. This article presents an innovative approach for determining absolute flood probability for the medium-to-long-term (2090) and long-term (2120), encompassing extreme and frequent floods over the next 100 years, by combining the individual CDF of SLR, tide, and storm surge. This approach allowed the production of a new vulnerability mapping, with high spatial resolution, by adding an image processing algorithm to eliminate all flood areas isolated from tidal surface flow at the maximum total hydrostatic water level (THWL). Geographic Information Systems (GIS) can replicate this methodology. Still, as a test zone, it is specifically applied as a demonstration to the Tagus Estuary, Portugal. In 2090, the highest flood level estimation on the Tagus Estuary margins is 4.06 m, corresponding to a total vulnerable area of 92 km². However, by 2120, it is expected that maximum flood height will reach 4.87 m, expanding the area to 102 km². After applying the image processing algorithm to clean isolated flood areas, a decrease of 20 % in the medium-to-long-term and 17 % in the long-term of the vulnerable area was obtained, relative to hazard assessment. The present study intends to obtain the absolute coastal vulnerability for further coastal risk assessment, based on a non-scenario-dependent flooding probability.