Recent studies have shown that the spectral wind wave model SWAN (Simulating Waves Nearshore) underestimates wave heights and periods in situations of finite depth wave growth. In this study, this inaccuracy is addressed through a rescaling of the Battjes and Janssen (1978) bore‐based model for depth‐induced breaking, considering both sloping bed surf zone situations and finite depth wave growth conditions. It is found that the variation of the model error with the breaker index γBJ in this formulation differs significantly between the two types of conditions. For surf zones, clear optimal values are found for the breaker index. By contrast, under finite depth wave growth conditions, model errors asymptotically decrease with increasing values of the breaker index (weaker dissipation). Under both the surf zone and finite depth wave growth conditions, optimal calibration settings of γBJ were found to correlate with the dimensionless depth kpd (where kp is the spectral peak wave number and d is the water depth) and the local mean wave steepness. Subsequently, a new breaker index, based on the local shallow water nonlinearity, expressed in terms of the biphase of the self‐interactions of the spectral peak, is proposed. Implemented in the bore‐based breaker model of Thornton and Guza (1983), this breaker index accurately predicts the large difference in dissipation magnitudes found between surf zone conditions and finite depth growth situations. Hence, the proposed expression yields a significant improvement in model accuracy over the default Battjes and Janssen (1978) model for finite depth growth situations, while retaining good performance for sloping bed surf zones.
Read full abstract