Author(s): Maurice Mc Cabe; Peter Stansby
Linked Author(s): Peter Stansby
Keywords: No Keywords
Abstract: Spectral wave energy models such as SWAN are widely used to make predictions of waves in deep and shallow water, incorporating the effects of wave interactions and transformations such as refraction, diffraction and breaking. To calculate nearshore processes such as wave runup on a wave-by-wave basis, phase-resolving shallow-water and Boussinesq-type models are more useful. However, these models are more computationally expensive and cannot be used in deep water. Therefore, a convenient solution is to use the output from a spectral energy analysis to create a wave input for a shallow-water/Boussinesq-type analysis in the nearshore. However, it is unclear how and where best to couple the models. To investigate this, SWAN and shallow-water-and-Boussinesq (SWAB) model runs were compared with results from a range of physical model tests. For each test, a 1-D SWAN model was run to give spectra at various locations in the nearshore. These spectra were used to generate random wave inputs for a 1-D nonlinear shallow water (NLSW) model. SWAB model and coupled SWAN-NLSW results for wave heights, setup and runup were compared to the physical model data. Results show that the SWAN model produces wave heights that match well with the physical data in the surf zone, despite the proportions of breaking waves not being so well predicted. The SWAB model tends to give inaccurate wave heights after breaking, with poor wave height distributions near the shore, but fairly accurate runup predictions. The coupled SWAN-NLSW model generally gives better results than the SWAB model near the shore.
Year: 2010