A new approach for the assessment of coastal flooding risk. Application in Rhodes island, Greece
Dimitra I. Malliouri,Chara Kyriakidou,Vyron Moraitis,Dimitrios Vandarakis,Nikolas Martzikos,Fragkiska-Karmela Gad,Georgios-Angelos Hatiris,Vasilios Kapsimalis
DOI: https://doi.org/10.1016/j.apor.2024.104006
IF: 3.761
2024-04-21
Applied Ocean Research
Abstract:A directional, copula-based, Monte Carlo approach is proposed for the quantitative assessment of a new definition of coastal flooding risk, namely a probability-based index due to extreme combinations of wave conditions and sea levels, which could be a key element for coastal zone planning and management. This approach differs from the standard practice, which focuses on the extreme events and studies either their specific return values or their worst-case scenarios. The adopted approach incorporates mean wave direction as a covariate into the probabilistic representation of sea states, and via the use of a variety of probabilistic and statistical tools, represents a total of five stochastic variables, namely the significant wave height, the peak wave period, the mean wave direction, the storm surge, and tide. Two directional, alternative methods have been developed; a trivariate copula- and a bivariate copula-based method, that seemed to produce quite similar results. Four types of copulas have been examined for their goodness-of-fit, i.e., three Archimedean copulas, namely, the Clayton, the Gumbel, and the Frank copulas, as well as the elliptical Gaussian copula. A comparative analysis of the two methods is performed aiming to point out their pros, as well as their accuracy and complexity issues. The application of the proposed method to the natural beaches of the Rhodes Island in Greece, in combination with a recently developed formula for wave run-up, highlighted the most crucial parameters affecting the coastal flooding risk in the study area.
engineering, ocean,oceanography