2. Methodology
2.1. Wave dynamics model selection
The SWAN (Delft University of Technology) and MIKE 21 SW (Danish Hydraulic Institute) third-generation wave models were considered for use in this study. Comparison of SWAN with MIKE 21 SW shows that both models give very similar results. However, as MIKE21 SW is designed for wider application and not specifically for application as a 1D model, it is more computationally expensive than SWAN. Similarly, SWAN 1D provides wave setup as an output based on an explicit solution; therefore there is no need to run a hydrodynamic model to calculate wave setup, again reducing computation time. Consequently, the SWAN 1D model was selected due to the extensive number of simulations required to cover the range of bathymetric conditions. The SWAN model has been used for prediction of waves over coral reefs in a range of locations (Vitousek et al., 2007;Storlazzi et al., 2011), and extensively tested for wave propagation in a wide variety of coastal environments (Ris et al., 1999). The inputs required for the model are bathymetry, water level, surface roughness and wind and wave conditions at the offshore model boundary.