Researcher(s)
- Ryan Canatella, Mathematics, University of Delaware
Faculty Mentor(s)
- Philippe Guyenne, Mathematical Sciences, University of Delaware
Abstract
Numerical models are used to forecast ocean waves for applications such as shipping, fishing, and recreation. However, current operational models trade accuracy for computational efficiency and rely on many simplifying assumptions. A new model in development solves the full Euler equations for ocean surface waves and has the potential to improve forecasting accuracy in a variety of scenarios. It requires thorough testing against an operational model, which needs to be done carefully to ensure that the simulation setup is identical and the same quantities are being compared between the models. This research project investigated an existing forecasting model (NOAA’s WAVEWATCH III v6.07) in preparation for detailed testing against the new model. WAVEWATCH documentation was reviewed and a detailed parametric study was performed by varying propagation and source term time step length, simulation duration, spatial domain size and grid spacing, initial conditions, boundary conditions, and the nonlinear term calculation method. The initial conditions tested included Gaussian wave packets with various peak frequencies and spreads, JONSWAP spectra, and linear varying wave heights. Boundary conditions included absorbing points and both time-independent and time-dependent active boundary points. The JONSWAP initial conditions were found to be suitable for comparative testing. Periodic boundary conditions are ideal for testing the new model but are not supported by WAVEWATCH for Cartesian grids, so an implementation will need to be written. Future work includes adding support for periodic boundary conditions in WAVEWATCH and performing the tests to compare against the new model.