whose incident direction is almost parallel to the current, linear 

 theory is valid only for waves of stricftly zero amplitude. Smith notes 

 this. Peregrine (1982) considers nonlinear effects, and shows that 

 certain dlffractive effects, which are usually not considered, are 

 important in the case of near-parallel incidence for linear and non- 

 linear waves. 



Another case that may be relevant to this problem is wave-wave 

 interaction, some examples of which might be interpreted as wave-current 

 interaction (see discussion of internal waves in Sec. II, 5). Another 

 isolated example is Taylor's (1962) study of standing waves on a 

 current . 



9. Experimental and Field Observations. 



Most of the experimental work has been in flumes, with or without 

 wind, which constrain waves and current to travel in the same or 

 opposite directions. The most detailed work of this type is by Kemp and 

 Simons (1982). They confirm observations by Van Hoften and Karaki 

 (1976), Bakker and van Doom (1978, 1980), George and Sleath (1979), 

 Brevik and Aas (1980), and Brevik (1980) that waves have a significant 

 effect on the mean flow and hence on the bed friction. This then raises 

 questions about the equivalent flow without waves, and means that it is 

 difficult to assess any measurements which do not Include observations 

 of the mean level and mean flow in a flume. (An exception can be made 

 for waves of very small amplitude where any reaction on the flow is 

 negligible; see also Sec. II, 2.) As discussed previously, Sarpkaya's 

 (1957) experiments which show wave amplification are among the most 

 interesting. 



Experiments which have set out to check the linear wave dispersion 

 equations, such as Shemdin (1972) and Thomas (1981), have shown good 

 agreement with theory when the variation of current velocity in the 

 vertical has been taken into account. Shemdin also included the airflow 

 since he was considering wind waves. However, Plant and Wright (1980) 

 find that inclusion of finite-amplitude effects does not improve 

 comparisons between theory and experiment. Freds^e (1974) finds 

 satisfactory agreement for stationary waves. Observations and 

 experiments of wind -generated waves on currents were briefly discussed 

 in Section II, 3. 



Measurements of water waves in regions of appreciable tidal currents 

 normally show fluctuations which correlate with the tides. Examples of 

 such observations are given by Vincent (1979) and Vincent and Smith 

 (1976). 



One of the problems in correlating measurements in the open sea is 

 that wave energy arrives from different directions with different 

 frequencies, and all components have different time histories of wind 

 and current. However, since there are wave-forecasting numerical models 



45 



