546 



Internal Waves 



Their superposition shall give for the various depths of observations the 

 observed values of amplitude and phase of the vertical displacement. This 

 enables us to determine the amplitude and the phase of each of the four 

 waves. It has to be emphasized that this is only possible if the number of 

 internal waves under consideration is smaller than the number of depths of 

 observarion. Figure 227 (left) shows the vertical distribution of r\ for the 

 four waves, as well as the related phases. On the right the figure shows the 

 combination of these four waves and a comparison with the observations. 

 A similar, good agreement will be obtained in the case of waves with a diurnal 

 period. But this agreement signifies only that the observations can be re- 

 produced sufficiently exactly by superposing a small number of internal waves. 

 But that will be always possible if one takes as many internal waves as possible 

 in successive order. A better proof of the theory is a comparison between 

 the theoretically computed and observed currents. Table 95 presents such 

 a comparison for the diurnal tide wave. The semi-diurnal wave could not 



Table 95. Currents of diurnal tidal period at "Snellius" Station 253a 

 (According to observations and computations based on vertical displacements) 



be used to test the theory in this respect. For there are indications that the 

 semi-diurnal wave is not a simple progressive one, therefore will not entirely 

 comply with the assumptions. The table shows quite a good agreement, if it 

 is kept in mind that the observations include also the normal tide current. 

 The greatest difference will be found in a depth of 350 m, where according 

 to the theory, the diurnal wave should disappear almost completely. It is 

 probable that in this depth the observations represent the undistorted normal 

 tide current. It is remarkable, that such complicated periodical variations 

 of the thermohaline structure can be reproduced by a relatively small number 

 of internal waves. But for carrying out such an analysis it is necessary to 

 repeat the serial observations in short intervals and to make simultaneous 

 and continuous current measurements over quite a long period. Of course, 

 other periods than the well-known tidal periods may be present and can con- 

 fuse the oscillation considerably. During the current measurements, which 

 were made by Ekman and Helland-Hansen (1931), on the Armauer Hansen 



