Direct simulation and deterministic prediction of large-scale nonlinear ocean wave-field

Abstract

Despite its coarse approximation of physics, the phase-averaged wave spectrum model has been the only type of tool available for ocean wave prediction in the past 60 years. With the rapid advances in sensing technology, phase-resolved nonlinear wave modeling, and high performance computing capability in recent years, the time has come to start developing a new generation tool for ocean wave prediction using direct phase-resolved simulations. The key issues in developing such a tool are: (i) proper specification of initial/boundary conditions of the nonlinear ocean wave-field; (ii) development of efficient algorithm for simulation of large-scale wave-field evolution on high performance computing platforms; (iii) modeling of nonlinear physics in ocean wave evolution such as wave-wave, wave-current, wave-bottom and wave-wind interactions. The objective of this thesis is to address (i), (ii) and part of (iii). For (i), a multi-level iterative wave reconstruction tool is developed to deter- ministically reconstruct a nonlinear ocean wave-field based on single or multiple wave probe records, using both analytic low-order Stokes solutions and High-Order-Spectral (HOS) nonlinear wave model.

(cont.) With the reconstructed wave-field as the initial conditions, the ocean wave-field can then be simulated and forecasted into the future deterministically with the physics-based phase-resolved wave model. A theoretical framework is developed to provide the validity of the reconstructed wave-field and the predictability of future evolution of the reconstructed wave-field for given wave conditions. The effects of moving probe, ambient current and finite water depth on the predictable region are studied respectively. To demonstrate its efficacy and useful- ness, the wave reconstruction tool is applied to reconstruct the full kinematics of steep two- and three-dimensional irregular waves using both wave-basin measurements and synthetic data. Excellent agreements between the reconstructed nonlinear wave-field and the original specified wave data are obtained. In particular, it is shown that the inclusion of high-order effects in wave reconstruction is of significance, especially for the prediction of the wave kinematics such as velocity and acceleration. For (ii), a highly scalable HOS wave model is developed and applied to study both two- and three-dimensional ocean wave-field evolution for a realistic space and time scale.

(cont.) Effective filtering tools are developed to model the wave breaking process in wave evolution. For (iii), the HOS wave model is enhanced to account for not only nonlinear wave-wave interactions, but also nonlinear wave interaction with variable ambient current. With this tool, the effects of variable ambient current on nonlinear wave-field evolution are investigated. As a final illustration, this tool is applied in practical ship motion control. Based on the deterministically forecasted wave-field provided by this tool, an optimal path is obtained to reduce the RMS heave motion of ship in point-to-point transit.