An understanding of natural coastal hazards is essential for improving resilience of coastal communities. Effective management of coastal hazards relies on prediction tools and mitigation structures that have been designed to handle the dynamic coastal system. This dissertation is a collection of manuscripts that cover a diverse range of fundamental topics on coastal hazards. The three manuscripts are separated into three general coastal hazard topics: (1) propagation, (2) transformation, and (3) impact.
For propagation, the concept for an efficient alternative approach is developed to estimate uncertainties in nonlinear shallow water forecasts. The alternative methodology delivers spatio-temporal means and variances for predictions of coastal hazards. A novel procedure is presented to efficiently and accurately estimate sample variances from a reference solution produced by a nonlinear forecast model by exploiting Green's functions to compute linear perturbations. This method avoids the cost of repeated nonlinear model runs by using pre-computed numerical Green's functions to store the response between source points and receiver points. Numerical Green's functions are computed by solving the first order perturbations to the nonlinear shallow water equations. A linear perturbation equation solver is developed and model responses are stored as Green's functions. Stored Green's functions allow for efficient calculation of ensembles of results, providing variance estimation for an arbitrary number of receiver points. Note that this procedure can produce variances with space and time dependencies, which are critical to ascertaining the uncertainties and sensitivities in problems based on hyperbolic equations. We complement the Green function variance estimation strategy with spatio-temporal Kriging in order to further extend the results to any location on the computational domain, whether on the computational lattice or outside of it. We demonstrate the Green's function approach on a synthetic shallow water flow problem. The efficiency and accuracy of the strategy is assessed theoretically and practically.
For transformation, interactions between a solitary wave and a submerged rectangular obstacle are investigated experimentally. Free surface measurements are taken using wave gauges and the laser induced fluorescence technique to study both global and local behavior of the wave. Experiments are designed to study the behavior of the transmitted wave. For narrow enough obstacles, the transmitted wave are found to be unaffected in waveform and amplitude evolution. Small to medium obstacles, relative to the wavelength, are found to result in a train of dispersive waves trailing the transmitted wave. In some cases, the wave amplitude is found to recover, despite being perturbed, to the state of an undisturbed solitary wave at a certain distance beyond the obstacle. Wide obstacles are found to result in wave breaking which may occur over the obstacle or downstream of the obstacle. The type of wave response is found to depend on the length scale of the submerged obstacle and the nonlinearity of the solitary wave. Wave responses from eighteen different test conditions are summarized in a parameter space involving the length scale of the submerged obstacle and the nonlinearity of the wave.
For impact, impulsive forces on a box-shaped structure due to bore impact are investigated experimentally. Experiments are conducted in the laboratory by generating a Gaussian-shaped wave onto a step before impacting structures of three different cross-shore to long-shore length ratios. The relationship between measurements of the splash-up flow on the structure's vertical wall and the force measurements is investigated. The pressures are predicted based on the Euler model. Maximum impact forces are found to occur during the run-down phase after the maximum splash-up value is reached. The uniform vertical velocity profile assumption used in the Euler model yields a better estimate for the impact force than the linear and hydrostatic assumptions. In order to verify the pressure distribution during the splash-up flow, we utilize the pressure data collected from a separate experiment in a large-scale laboratory flume. The combined evidence suggests that the flow near the splash-up tip behaves like a solid body projectile.
Each subtopic in this dissertation is independently described as a separate manuscript. All three research subtopics combine to provide an integrated and diverse investigation into improving mitigation measures for coastal hazards.