Internal solitary waves (ISW) have been detected on the interface of a stratified water column in the ocean. It is believed that ISW could affect oil drilling operations, nutrient pumping, and acoustic signal obstruction. In the ocean, the thickness of a pycnocline is finite which differs with the theoretical assumption as being a thin layer. This thesis reports the effect of an ISW propagation in various pycnocline thicknesses. Laboratory experiments were conducted in an internal wave flume (0.5¡Ñ0.7¡Ñ12m) at the National Sun Yat-sen University, Kaohsiung, Taiwan. ISW in depression or elevation type were generaled using a stratified two-layer fresh/brine water system with a total depth of 50 cm in the flume. Upon creating an ISW propagating on a flat bed or over a triangular obstacle later, several physical parameters of the ISW (i.e. wave amplitude, phase speed, characteristic wave length, and wave energy) were measured or calculated for different thicknesses of the pycnocline. The major controlling factors in the experiments included the depth ratio of the upper to lower layer H1/H2, interface displacement £b0 between the wave generating chamber and the main flume, and the thickness of the pycnocline. The thickness of the pycnocline was estimated from the result of density profile in the vertical direction in the flume, experiments under the same H1/H2 and £b was terminated when the pycnocline thickness became large enough. As the thickness of the pycnoline increased, the values of all the physical parameters (including wave amplitude, phase speed, and wave energy) under consideration decreased. Their reduction rates were more significant in the case of small interface displacement (£b0=10cm) than that with large £b0=15cm. On the other hand, the changes in the physical variables associated with a depression ISW were more significant than those in an elevation ISW.

Laboratory experiments were conducted to investigate the propagation of internal solitary waves on a uniform slope in a two-layered free surface fluid system. The laboratory facilities employed in this study is the first in Taiwan, which include a stainless steel wave flume (dimensions: 12 meters long with cross-section 0.5 m wide and 0.7m deep) and experimental apparatus for generating and measuring internal waves. The flume incorporates a movable vertical gate at one end for generating internal solitary waves, and a uniform slope (either £c = 30o, 50o, 60o, 90o, 120o or 130o) at the other end. The upper layer had fresh water with density £l1 (999kg/m3), to a depth H1; the lower layer was saline brine density £l2 (1030 kg/m3), which was slowly filled into the flume to a depth of H2 by gravity through several openings at the bottom of the flume, Boussinesq parameter . A mini pump was used to remove a small quantity of fresh water from one side of the vertical gate to another side. By creating a prescribed difference £bo in the interface levels on either side of the gate beforehand, internal solitary wave was generated by the mechanism of overturning the brine and fresh water behind the movable gate. Five ultrasonic probes at equidistant distance recorded the interface fluctuations, one density probe measured the change of density at the interface, while two electrical capacitance gauges for the free surface displacements likely to occur. Digital cameras were also used to record the motions of internal wave in the flume and on the slope for further analysis. Laboratory test on internal solitary wave were arranged from one of the combinations using different layer thickness ratios H1/H2, interface differences £bo, density ratios £l1/£l2, and bottom slopes £c. In addition to internal solitary wave reflection from a uniform slope, laboratory investigations included internal wave propagation on a rigid impermeable bottom and evolution on a uniform slope. Keeping the total water depth in the flume at H = 40cm, an increase in the depth parameter |H2-H1|/H produced large internal wave amplitude, reduced phase velocity, and enhanced soliton feature. From the experimental result analyzed, it suggests that the Korteweg-de Vries (KdV) theory fits solitary waves of small amplitude, and the modified KdV is suitable for large amplituded waves. Considering wave motion in an inviscid fluid, the dissipation of internal solitary waves propagating in a flume may occur through bottom friction and wave breaking. Subjected to bottom friction alone, the amplitude of most internal solitary waves in the experiments decayed approximate by 10% over a journey of 6 meters. Two types of wave breaking mechanism were found to produce strong mixing and local vortex in the fluid, causing significant energy losses. For internal solitary waves of large amplitudes, reflection coefficient for wave amplitude or energy decreased, as amplitude or energy increased. Under this condition, however, the reflection coefficient due to bottom friction may be assumed as constant. Using the experimental results obtained, empirical equation is now proposed to account for wave dissipation due to for non-breaking internal waves. The equation indicates that decrease in reflection coefficient as wave amplitude or energy increases may be expressed using a second order polynomial. Overall, experimental results suggest that good agreement can be found between experimental data and the empirical equation so derived. Upon assuming the wave reflection coefficient is solely dependent on the incoming wave amplitude or energy, prediction for reflection coefficient can be calculated in a straight forward manner. Either large-scale, high-frequency internal wave motion or internal solitary waves have been observed in natural lakes. The observed rapid decay of internal wave energy after severe breaking events seemed to be mostly due to dissipation on various sloping boundaries in a lake. From the basic laboratory experiments on internal wave reflection from various single slopes, the results many benefit provide researchers to promote further research on practical applications related to limnology.

Laboratory work were conducted to investigate the behaviors of an internal solitary wave (ISW) in a two-layer free surface fluid system in a wave flume (12m¡Ñ0.5m¡Ñ0.7m) at the National Sun Yat-sen University, Kaohsiung, Taiwan. A series of fundamental experiments on wave generation, propagation and interaction with uniform slopes and topographic features were carried out in the flume with stratified two-layer fresh/brine water. Factors governing the experiments included the thickness ratio of the upper and lower layers H1/H2, interface difference

Internal solitary waves (ISWs) are commonly observed in the ocean, and they play important roles in many ways, such as transport of mass and various nutrients through propagation. The fluids considered in this thesis are assumed to be incompressible, inviscid, non-diffusive and to be weakly affected by the Earth's rotation. Comparisons of the evolution of an initial solitary wave predicted by a fully nonlinear model, IGW, and two weakly-nonlinear wave equations, the Ostrovsky equation and a new alternative Ostrovsky equation, are done. Resolution tests have been run for each of the models to confirm that the current choices of the spatial and time steps are appropriate. Then we have run three numerical simulations with varying initial wave amplitudes. The rigid-lid approximation has been used for all of the models. Stratification, flat bottom and water depth stay the same for all three simulations. In the simulation analysis, we use the results from the IGW as the standard. Both of the two weakly nonlinear models give fairly good predictions regarding the leading wave amplitudes, shapes of the wave train and the propagation speeds. However, the weakly nonlinear models over-predict the propagation speed of the leading solitary wave and that the alternative Ostrovsky equation gives the worst prediction. The difference between the two weakly nonlinear models decreases as the initial wave amplitude decreases.