We present a non-hydrostatic multilayer solver adapted for phase-resolved simulation of coastal wave dynamics over complex bathymetry and in the presence of coastal structures. The solver discretises the incompressible Navier–Stokes equations using a vertically Lagrangian layered representation in which the layer interfaces are material surfaces and a non-hydrostatic pressure correction couples the layers. A relatively small number of layers suffices to represent the smooth vertical structure of surface gravity waves, so the formulation resolves dispersive nonlinear wave dynamics efficiently over a wide range of water depths.A generating-absorbing boundary condition (GABC) based on a dispersive Padé approximation of the linear dispersion relation is implemented, enabling simultaneous wave generation and absorption of reflections at the boundary without extended relaxation zones. An embedded-boundary formulation is introduced that enables the representation of vertical coastal structures such as breakwaters and quay walls within the multilayer framework.The model is validated against laboratory experiments covering linear, nonlinear and breaking wave conditions, wave transformation over variable bathymetry, wave loads on a bottom-mounted cylinder, and diffraction through harbour openings, and is benchmarked against a volume-of-fluid reference solver and the SWASH and FUNWAVE wave models. Numerical energy losses remain below one per cent during long-distance propagation, and the energy dissipation of waves breaking on a beach is reproduced without an empirical breaking closure. The generating-absorbing boundary achieves measured reflection coefficients of a few per cent across the dispersive range, and the embedded boundary reproduces the measured wave loads on a vertical cylinder, including the secondary load cycle.
Submerged aquatic vegetation (SAV) can modify coastal hydrodynamics through canopy drag, wave dissipation, and changes in near-bed stress, yet embayment-scale models often neglect SAV or treat it as a single vegetation class. This study investigates the hydrodynamic influence of three dominant mapped SAV classes in Port Phillip Bay (PPB), Australia, including Brown Macroalgae, Mixed Macroalgae, and Seagrass, using a high-resolution coupled SCHISM–WWMIII model, with SCHISM operated in depth-averaged barotropic mode. The model was validated against tide-gauge, wave-buoy, and current observations, and reproduced the main variability in water level, wave height, peak wave period, and currents. Vegetation characteristics were represented using class-specific effective parameters derived from previous studies and regional observations and were treated as representative rather than locally calibrated values. Numerical experiments were then conducted to compare the hydrodynamic responses associated with different mapped vegetation configurations and to assess the effects of vegetation evolution. Results show that SAV modifies bay hydrodynamics in a strongly spatially heterogeneous manner, with the response controlled jointly by vegetation properties, habitat distribution, and background hydrodynamic conditions. Mixed Macroalgae configuration produces the weakest response, whereas Brown Macroalgae and Seagrass configurations generate stronger changes in water level, current speed, and bottom shear stress. The combined vegetation case dampens water-level extremes, slightly reduces mean wave height across much of the bay, and substantially decreases current speed and bottom shear stress in shallow vegetated regions and near the bay entrance. At representative stations, SAV reduces current speed by about 48–84% and bottom shear stress by about 66–96%. Simulations using the 2020 and 2025 vegetation maps further indicate that vegetation evolution alone can reduce current speed and bottom shear stress by up to about 50% in parts of the northern and eastern bay. These percentages represent controlled model-scenario responses relative to the corresponding reference simulations rather than direct estimates of observed changes across PPB. These results demonstrate that vegetation characteristics, habitat configuration, and vegetation evolution can act as important controls on bay-scale near-bed hydrodynamics, and that habitat-informed modelling provides a useful tool for assessing the hydrodynamic consequences of vegetation change and restoration.
