Nonstationary Based Framework for Performance Enhancement of Coastal Flood Mitigation Strategies

Abstract

The flood protection topic has been the centerpiece of many studies due to the amplification of coastal vulnerabilities, especially on the east coast of the United States. In this study, increasing the reliability of the urban infrastructure has been addressed by a more accurate understanding of flood hazards and coastal protection strategies. The first step in this study is the hydrologic data analysis. Climate change and human activities have often caused an invalidation of stationarity assumption for simulation and forecasting flood hazards. Also, hydrological events often coincide with two or more natural events. Therefore, for a more accurate analysis, three approaches are introduced to assess the nonstationarity in data. Then, the nonstationary frequency analysis on a univariate distribution function and the joint probability distribution have been applied to define the frequency of flood hazards. First, a trade-off curve is plotted to contain all results of a nonstationary joint probability analysis that can be used to select the design value curve based on decision-makers’ preferences. The second step is the selection of coastal protection strategies. Strategies are selected based on their suitability and the characteristics of the protected area. The final step is to utilize a hydrologic distribution model, with the ability to simulate both rainfall and storm surge for obtaining a flood inundation map in order to test the strategies. Although, this model has certain limitations of not including wave dynamics and setup and wind speed. It succeeds in the integration of hydrologic and systems analyses, which is the main advantage of the methodology. The results show the significant value of a nonstationary analysis for obtaining design values and selecting comprehensive coastal protection strategies. The proposed methodology can be implemented in other coastal geographic settings.