The transformation of gravity waves by a horizontally sheared current

The thesis describes a unique experiment conducted in the UK Coastal Research Facility (UKCRF) to investigate the transformation of near-linear gravity waves propagating across a narrow horizontally sheared jet-like current in constant depth water. The UKCRF is a novel large-scale wave basin incorporating a multi-directional wave generator and a refined current generation system making it ideally suited to the study of three-dimensional wave-current interactions. Setting the current generation system required particular care to create a jet-like current that remained uni-directional and produced minimal recirculation within the basin. A single wave condition was studied, propagating across the current orthogonally and at oblique incidence, in both a following and opposing sense to the current. The length-scale of the current's shear layers was comparable to the incident wavelength. The experiment is the first attempt to assess the kinematics and dynamics of the interaction of regular waves and currents in three-dimensions at a physically realistic scale. The resulting benchmark data set provides direct quantitative measurements of the spatial variation of the primary flow variables, the velocity vector, u, and the surface elevation, η. Observations showed the following wave to be refracted to a more current-parallel direction with an increased wavelength and reduced wave height as it moved onto the current, while the opposing wave became more current-normal with a shortened wavelength and enhanced wave height. Negligible reflection of the incident wave at the current shear layers was observed. Four analytical models, each making different assumptions about the current properties, are compared with the data set to assess their suitability for predicting wave transformation by narrow jet-like currents. Predictions based on the WKBJ approximation agree well with the trends observed in the data indicating a slowly-varying current behaviour. The classification of a current as rapidly- or slowly-varying based solely on the parameter β= ω 2L/g is therefore inappropriate