Journal of Waterway, Port, Coastal, and Ocean Engineering

Through construction experience and research, including model tests, on the Nanjing reach of the lower Yantze River, it was found that different methods must be used to stabilize the discharge ratio of two branches around an island, depending on the length ratio of the two branches. When the two branches have a similar length, stabilizing the upstream bends is effective. When the length of one branch is longer than the other, stabilizing the head of the island will be most effective. A subsurface stone blanket of appropriate thickness has been the most popular type of revetment because of its resistance to flow scouring. A design method to determine the thickness of a subsurface stone blanket, how to launch stone to desired positions, and other innovations for improvement of the stone blanket revetment are discussed.

A new concept of statistical analysis of multibar profiles is presented in the study; vertical residuals about a reference Dean curve are treated as sine functions with random amplitude (envelope), bar wave number and phase. The data set consists of 71 transects, collected from 1964 to 1994 by the Polish Academy of Science, Institute of Hydro-Engineering, Coastal Research Facility at Lubiatowo (CRF), situated some 80 km northwest of Gdańsk. The morphology of the nearshore zone there is typical of the south Baltic Sea and is characterized by a mild slope (m= tan α≈ 1.0) and fine sand sediment D₅₀= 0.22 mm). Because the Baltic Sea is semiisolated from the Atlantic, no tides can be observed, and beach profiles are mostly subject to wave attack and storm surges. The envelope of random sine functions (RSF) describes mean deviation about a Dean reference profile with respect to its location on the profile. Thus, it depicts vertical cross-shore variability, which may be useful for determination of wave breaking zones where the profile variability is especially intensive (envelope peaks). Bar wave numbers characterize bar lengths and distances between bars along the profile. Multibar cross-shore profiles can thus be analysed by means of only two parameters, based on very simple RSF methodology. The random sine function method poses an alternative in multibar beach profile description with respect to empirical orthogonal functions, because it seems to be less ambiguous in many cases. The results of the study contain empirical relationships, which depict bar lengths on a multibar profile, and vertical deviations about the reference profile, which can be regarded as assessment of bar height.

A frequency domain wave transformation model is derived from a set of time-dependent extended Boussinesq equations. In contrast to an earlier study, this model is derived directly from the (η, u) form of the equations. The resulting model maintains the excellent dispersive properties of the original equations and also accurately mimics the shoaling behavior of dispersive linear theory for a wide range of water depths. The model is formulated in terms of a free parameter that can be tuned for optimum shoaling behavior. We tune the parameter in two ways; the first seeks the minimum error for the shoaling parameter while using the optimum value for the dispersion parameter, while the second seeks the minimum combined error for both dispersion and shoaling. These lead to two different sets of free parameters, with the second method leading to more favorable linear shoaling behavior than the first. The effectiveness of the linear terms of the model is demonstrated by using them, with both sets of free parameters, to propagate waves over a shoal; comparisons to both experimental data and a linear mild-slope parabolic model are performed, and agreement is favorable for both sets of parameters. The full model, with nonlinear terms and both sets of parameters, is then used to simulate a laboratory experiment involving harmonic generation and nonlinear wave focusing. Results indicate that the model can reproduce the characteristics of a lowest-order Kadomtsev-Petviashvili (KP) model in shallow water. The model also shows improved agreement with data relative to the KP model and another dispersive frequency-domain model in intermediate water depth. Additionally, there is little difference between the results from either set of parameters used in the model, indicating that for these cases it is not clear which optimization is most ideal.

A new model is developed to study the initial motion of sediment in oscillatory flow over a flat bottom on the basis of available experimental data. It is found that the initial motion of sediment in oscillatory flow is uniquely defined by the simple model of A/d=KA²ω/ν+B, where A= semiexcursion of wave orbital motion near the bed; ω= angular frequency; d= grain diameter; ν= kinematic viscosity; and K and B= dimensionless coefficients determined by the immersed sediment weight. For a given sediment, the onset velocity derived from the model is found to initially increase with the wave period T and then approach a constant. A practical formula is also presented to calculate the onset velocity of sediment in the coastal zone.

A numerical model of wave run-up, overtopping, and regeneration is presented. The model (called OTT) is based on the one-dimensional nonlinear shallow water equations on a sloping bed, including the effects of bed shear stress. These equations are solved using a finite-volume technique incorporating a Roe-type Riemann solver. The main advantage of this approach over previously used finite difference solvers is that no special shoreline-tracking algorithm is required, so that noncontiguous flows can easily be simulated. Hence, this model can be used to simulate the transmission of waves over water surface-piercing obstacles. The numerical scheme and boundary conditions are described, and several existing data sets used to test the ability of the model to simulate wave transformation, run-up, and overtopping. Experiments of random wave (unimodal and bimodal) overtopping, presented here for the first time, indicate that the model performs much better than empirical formulas in predicting average overtopping rates, and that it provides good estimates for the number of overtopping events. Experiments of overtopping of a sea wall by a solitary wave are also presented, including measurements of wave regeneration in lee of the dike. The model does a reasonable job of reproducing the water depths on top of the dike, and performs well in simulating the initial height of the regenerated waves.

Two field tracer studies and some numerical model experiments were used to analyze the residence time of freshwater in Boston's inner harbor, and thereby help evaluate water quality impacts of combined sewer overflows. An exponential filter was used to modify the freshwater fraction approach for variable freshwater inflow and to reanalyze data from Bumpus et al. (1953). Results showed an inverse relationship between residence time and inflow rate, with times ranging from 2 to 10 d. An instantaneous dye study gave a residence time of 3.75 d, consistent with the freshwater measurements for conditions of summertime low flow. A three-dimensional (3D) numerical model applied to a schematized domain was able to reproduce trends observed in both the freshwater and dye studies.

Wave reflection from coastal structures is defined by the magnitude and phase of the reflected wave. Both properties have a profound impact on the wave kinematics and coastal processes in front of the structure. This paper focuses on the phase shift on reflection. Using a large experimental data set, involving normally incident and obliquely incident regular and irregular waves, it is shown that the phase is uniquely determined by a nondimensional number χ₃ defined by structure slope, water depth at the structure toe, wave period, and angle of incidence. A theoretical method for predicting phase shift based on matching the equation for linear long waves on a sloping beach with an equation for flat-bed standing waves seaward of the structure toe yields good estimates of the phase shift at low values of χ₃. A second method based on integration of the shallow water wave number over the slope is considered and not recommended. Example cases are presented that demonstrate the practical importance of the phase shift in determining the kinematics in front of the structure.