The first phase of the project investigated the heart of the problem: understanding how extreme sea levels and storm surges are connected across space and time. Using advanced statistical methods like K-Means clustering and copula analysis, we mapped the spatial footprints of these events on a global scale. We discovered recurring patterns, particularly in regions like the Gulf of Mexico, where seasonal and decadal changes in spatial dependencies were pronounced. By identifying the weather patterns—such as atmospheric pressure and wind dynamics—driving these changes, the team gained deeper insights into the physical processes behind coastal flooding.
We identified the years when the flooding risk is higher due to the interaction between the storm surges and astronomical tides using statistical models and sea level observations from tide gauges. The results allowed us to also project in the future when the next peak in flooding risk will occur and where along the global coasts.
In addition to the spatial clusters, we assess the temporal clustering of storm surges. This approach revealed that some coastal areas experience clusters of storm surges in quick succession, leaving little time for recovery between events and significantly increasing flood risk.
With a clearer understanding of spatial dependencies, the team shifted focus to developing a new framework for coastal flood risk assessment. We created a synthetic dataset of spatially dependent ESWLs, which provided a more realistic basis for modeling flood risks. This dataset was complemented by the development of MatFlood, an efficient and user-friendly flood mapping tool. MatFlood allowed the team to simulate flood depth and extent, accounting for the spatial variability of coastal water levels.
The team then applied this framework to calculate present-day flood risk metrics, such as annual damage losses. By comparing these metrics with conventional approaches—which often assume complete independence or dependence between events—they demonstrated that neglecting spatial dependencies can lead to significant underestimations of flood risk. This work was further enriched by a study on compound flooding drivers along the northeastern US coasts, which explored how coastal flooding interacts with other hazards like river discharge and precipitation. The findings underscored the importance of considering multiple hazards in flood risk assessments.
The final phase of the project looked to the future, evaluating how coastal flood risks might evolve under different climate change scenarios. Using the MatFlood model, I simulated flood maps for future mean sea level conditions. A key case study focused on the Baltic coast of Germany, where collaboration with Kiel University enabled the integration of spatial dependencies into economic damage assessments.
Within the project, we delivered: 9 peer-reviewed papers, 4 conference contributions, a website containing flooding maps under future mean sea level conditions and extreme events in the Gulf of Mexico, a model to efficiently simulate coastal flooding taking into account the spatial varying coastal sea level, and a dataset containing spatial dependent storm surge events that can be used in flooding risk analysis.
These advancements provide valuable insights for policymakers, insurers, and coastal communities, helping them better prepare for and mitigate the impacts of coastal flooding in a changing climate. The project’s outcomes contribute to global efforts to enhance coastal resilience and reduce the societal and economic impacts of extreme sea level events.
No website has been developed for the project yet.
