Nearshore hydrodynamic model comparisons and predictions of bathymetric evolution

Abstract

The surf zone exhibits large energetic signals from wave shoaling and subsequent dissipation due to breaking, forcing circulation. The bathymetry responds to the wave and wave-induced circulation with the growth, transport, and destruction of large scale bathymetric features, such as mega ripples and sand bars. There is an obvious fluctuation of energy, which results in the continual change of the coastline, indicating a need for a predictive tool that includes a feedback between the hydrodynamics and morphology. Modeling of the nearshore environment has been of active interest to the engineering and oceanographic research community for close to a century. Wave and circulation prediction have improved dramatically since the advent of coastal modeling, and has reached a point in the modern day where the solution of the full Navier Stokes equation is in the foreseeable future. However, the computational expense associated with these highly skilled predictions of waves and currents are not always suitable for practical application in the field of coastal engineering. Therefore, in practice, engineers must compromise between computational efficiency and prediction skill of a model. A comparison of various model prediction skills and further discussion of such practical applications will occur in this manuscript. In contrast to the highly evolved wave and current models presently available, bathymetric evolution modeling is still in its infancy and has more recently become a primary focus of the nearshore researching community. The success of sediment transport predictions, and subsequent morphological evolution predictions, among other things, are dependant upon the included physics of the model, quality of input data, and stability of the modeling routines. In this study, hydrodynamic models were compared based upon their skill exhibited when predicting cross-shore profiles of nonlinear wave and wave velocity properties: skewness and asymmetry. These nonlinear characteristics have proven to be challenging to predict in a variety of hydrodynamic climates in combination with variable water depths. Predictions of skewness and asymmetry by three types of models were compared, which ranged dramatically in computational expense and complexity. The more complex models were derived using the physics-based governing equations of fluid systems, where the less complex and extremely computationally efficient models are empirically developed based on physics of the nearshore system. The most complex wave evolution model used in this study was FUNWAVE, which is a fully nonlinear Boussinesq wave model (Wei et al., 1995). The nonlinear, phase-resolving, fully dispersive wave model of Kaihatu (2001), in which the evolution of the waves is governed by the parabolic approximated mild-slope equation, is the second complex wave evolution model. Wave and velocity skewness and asymmetry are thus computed from the predicted time series from each of the two nonlinear wave models. The third type of model was empirical formulations developed to predict wave velocity skewness and asymmetry. Two of these empirical models were used in this study; the Doering and Bowen (1995) formulations, which were derived through bispectral analysis of wave evolution, and the Doering et al. (2000) formula, which was generated using an evolutionary algorithm. These empirically-derived expressions are currently being utilized, but have not been thoroughly testes, hence our comparisons of nonlinear wave characteristics between multiple model predictions and observations. Predictions of skewness and asymmetry were used to gauge the skill of these models. The quantity of wave skewness and asymmetry changes vertically in the water column, due to effects by the bathymetry in the wave boundary layer versus the free streaming flow. The characteristics of the nearshore flow field have been of interest to the engineering community because of the vital role these nonlinear quantities play in predictions of the morphologic evolution. Hence, our investigation of the quality of predictions of nonlinear wave properties by these three types of predictive tools. Another focus of coastal modelers is to ably simulate the hydrodynamics and subsequent morphodynamics of the nearshore using a complete feedback system; i.e. the predicted bathymetry after a pre-defined time step would be used as the initial bathymetry for the following time step and so on. A full feedback model of this sort is afflicted with issues ranging from model validity to stability. Also, there are currently few data sets available which contain all the necessary measurements required to thoroughly test such a proposed set of models. In the second half of this study a model series was implemented, using a feedback system, to simulate the 1DH (1-dimensional-horizontal), cross-shore evolution of sand beaches for a lab and a field experiment. The nonlinear, fully dispersive wave model of Kaihatu (2001) was used to predict the cross-shore evolution of wave field. A Bailard (1981) sediment transport model was used to predict the local cross-shore change of the bathymetry, as a function of time, from which the new cross-shore profile was predicted, and subsequently used as the initial bathymetry for the following time step. The overall goal of this study is to make progress in the direction of building community models, in which compatible wave, circulation, and transport models of differing degrees of complexity can be used in combination for simulating the nearshore environment. It is hopeful that in the near future we will be capable of simulating long term coastal evolution in real time, therefore, providing the ultimate modeling tool for the researching and practicing engineering community.