An Assessment Of Storm Surge Modeling PDF Download

Currently in the United States, more than 100,000 miles of levees exist that are facing substantial risk from storm surge. Data from modeling efforts can provide valuable information to decision makers to act against these risks. Probabilistic models for the performance of coastal structures against storm surge hazard are needed to gain a deep understanding of their failure mechanisms and occurrence probabilities in order to support various critical decisions needed prior to and during these events. Current probabilistic models for flood defense systems are primarily focused on a single mode of failure without consideration of causal relations among failure modes. In current models, failure assessment is treated as a snapshot in time neglecting the time evolution of failure processes. To address these limitations, a well-defined model of levee together with robust reliability analysis techniques are needed. The finite difference platform of FLAC3D is utilized in this study to analyze the failure mechanisms of levees. Subsequently, stochastic finite difference models of levees combined with machine learning techniques enable generation of a novel class of multi-dimensional fragility surfaces that will enhance our understanding of various failure processes. Another limitation of existing models is their inability to account for impacts of failures in flood defense systems on characteristics of storm surge. This issue is addressed through integration of proposed probabilistic levee performance models with hydraulic models to enable them to adapt in response to changing conditions of the flood protection systems. This integration improves forecasting capabilities of the models. Therefore, the goal of this research is to provide a better understanding of the complex behavior of flood defense systems, enable more accurate forecasts of the performance of these systems during storm surge events, and finally devise more effective ways to transfer the produced information to decision makers. To this end, this study advances the state-of-the-art in numerical modeling in geotechnical engineering, uncertainty quantification, and risk analysis, by integrating advanced approaches in reliability and fragility analysis with flood inundation modeling capabilities to improve decision making in the face of flood events. First, a physics-based performance model for geotechnical failure mechanisms in levees is developed. The produced model properly captures coupled behavior of soil instability and backward erosion piping that considers the hydro-mechanical coupling between the material properties. Then, a highly efficient probabilistic calibration framework that is integrated with multivariate Kriging surrogate modeling is developed and applied on the produced performance model to reduce uncertainties in the input variables. After that, the current issues in reliability analysis of geotechnical systems are investigated and a new active surrogate reliability method is developed using an effective learning function that facilitates the active learning process. The developed reliability method is integrated with random field methodologies to evaluate reliability of levee systems with spatial correlation in input properties. The developed probabilistic models are integrated with breach models to provide a set of two-dimensional fragility models that are functions of key factors including initial water level before surge and peak surge elevation. Unlike existing fragility models, this new class of fragility models can accurately represent various undesirable events in levees including initiation of breach, local geotechnical instabilities, and breach development. Then, the proposed fragility models are integrated with flood hazard models and flood inundation modeling to analysis the flood risk in the downstream of the levees. In the developed risk analysis framework, the consequences are modeled in terms of human casualties, damage to the levee systems, and economic loss. The developed models in this study helps decision makers for short-term during-hazard responses such as evacuation, and long-term risk management planning.