Wave Effects on Bathymetric Depressions

Wave refraction and diffraction effects resulting from bathymetric depressions (a dredged borrow pit and a dredged entrance channel to a port) were investigated for a major port expansion project. Due to the complex bathymetry and port configuration, a combination of physical and numerical modeling was developed. A third generation, phase-averaged, spectral wind-wave model was developed for this study using an unstructured mesh. The model was used to transform offshore wave conditions to provide the extreme nearshore wave design conditions of significant wave height, peak period, and mean wave direction (H_m0, T_p, MWD). A time-domain, Boussinesq-type wave model was also established to simulate the propagation of waves traveling from deep water to shallow water involving the primary wave motion as well as the bound and free long waves. This model is capable of reproducing the combined effects of most of the wave phenomena of interest to the harbor engineer, including shoaling, refraction, diffraction and partial reflection of irregular, finite amplitude waves propagating over complex bathymetry (i.e., borrow pits, dredged channel). The results of the study indicate that the surface interactions with the deep-draft and wide navigation channels significantly influence the incident waves and manifest themselves into both amplification and attenuation of the incident wave field. Both of these effects in-turn critically influence inputs to design, planning of port infrastructure, as well as assessment of harbor agitation and vessel downtime. While in recent years port designers have typically relied upon numerical wave models to perform most of the initial studies, in our study a combination of physical and numerical modeling was used. Based on the combined observations from the physical and numerical modeling results, conclusions and recommendations from this work show that both models compared well. Wave heights observed in the physical model were similar to or slightly higher than the numerical predictions.