waves. Currents generated by internal waves lead to patterns on the 

 water surface which, in some cases, are directly due to differing 

 steepnesses of short surface waves. Surface waves with a group velocity 

 close to the phase velocity of the internal waves are those most 

 strongly affected. These have been recorded in the Georgia Strait by 

 Gargett and Hughes (1972). Particularly strong effects have been 

 observed in the Andaman Sea by Osborne and Burch (1980) among others. 

 The most comprehensive field observations are those of Hughes and Grant 

 (1978) who generated internal waves with a ship and then measured 

 surface and internal wave fields. Hughes (1978) follows up these 

 observations with a theoretical analysis in which the general features 

 are reasonably well predicted by simplified analysis in which the 

 current magnitude is assumed small. 



There have been different ways of analyzing the interaction of 

 surface and internal waves (e.g., Watson, West, and Cohen, 1976; Rizk 

 and Ko, 1978; Basovich, 1979; Hashizume, 1980). This interaction has 

 also been studied experimentally with some success (Lewis, Lake, and Ko, 

 1974; Joyce, 1974). 



A major effect of the internal wave-current field is to trap 

 sufficiently short surface waves between a pair of caustics each side of 

 the internal wave's crest. The tendency to trap waves and to steepen 

 slightly longer waves leads to clearly visible surface traces of the 

 internal waves. These surface waves are too short to be of any 

 engineering significance except in the interpretation of remote-sensing 

 observations . 



The many ways in which the surface-internal wave interaction problem 

 has been studied indicate that it could be a fruitful field for further 

 wave-current interaction research. If all theories and experiments on 

 the topic are set in a framework of both wave-current interaction and 

 current-generating properties of waves, a distinct improvement in 

 understanding is likely. 



In practical applications, refraction by varying depth will occur at 

 the same time as refraction by currents. Tayfun, Dalrymple, and Yang 

 (1976) and Jonsson and Wang (1980) have investigated the effect of a 

 shoaling bottom with a shearing current parallel to the bottom contours, 

 another example in which solutions depend on a single coordinate. As 

 might be expected, the effects of current and bed gradients on wave 

 refraction can either reinforce or oppose each other. 



Another example which is sufficiently simple for analysis is the 



case of stationary waves on a current. Peregrine and Smith (1975) 



investigate a number of examples including variations of velocity with 

 depth as well as horizontally. 



39 



