On Risk Analysis Of Extreme Sea Levels In Falsterbo Peninsula

Abstract: Coastal protection is vital for protecting infrastructure, coastal environments and human lives against flooding. Building ecient coastal protection requires a good understanding of maximum sea levels which might occur indifferent time periods in the future. Extreme value theory provides a mathematical framework for such analyses. On November 13, 1872 the biggest recorded sea level surge devastated the Danish, German and Swedish Baltic Sea coast. This master thesis focuses on estimating the return period of 1872 storm using one-dimensional extreme value analysis based on historical data from the measure stations near Falsterbo Peninsula. A multivariate extreme value approach is applied to include covariates such as wind speed and wave height to further improve the understanding of which factors affect extreme sea levels. Fit diagnostics show that the block maxima model based on observations from Klagshamn measure station provides the best t for the data; hence it has been used to estimate the return period of the 1872 storm. Wind speed and wave height from nearby station were used to improve the accuracy of the analysis in a multivariate framework.