This has rather important consequences when we look at the dissi- 

 pation. When we first began this study it was thought that frictional 

 effects would erode the edges of the beam — the regions of high shear. 

 However, there is a large flux of energy into these regions, hence 

 the possibility of sustaining motion in the presence of strong dissi- 

 pation. Recall that the shear is large during only portions of a 

 cycle. In the central core, where the shears are minimal, there is little 

 flux of energy. Thus, rather than be eroded at the edges alone, the 

 beam of energy may be attacked by fluid friction more or less uniformly 

 across the beam. 



Progress has been made toward solving the spatial dissipation 

 problem; however, it is not yet sufficiently far along that conclusions 

 may be reached on how much energy of the internal tide system reaches 

 the sea floors. 



EXPERIMENTAL PROGRAM 



In experiments where large amounts of data are collected, pre- 

 analysis, conditioning, and processing of the data can be difficult and 

 time consuming. Data from our 1968 ocean experiment are still in the 

 preanalysis stage, but since the steps to be taken in the processing 

 and analysis are established, it is of interest to describe these 

 operations . 



Our experiment plan was to deploy sensors at two locations seaward 

 of the continental shelf break — the sensor array from each vessel to 

 span the predicted beam of internal waves. The model predicting the 

 beam of wave energy was validated when the observed signals showed 

 phase relationships in agreement with those predicted. The USCGC Ivy 

 was stationed at the inshore position and USC&GSS Ooeanographer at the 

 offshore location. 



The locations of the two sites and the deployment of the sensors are 

 shown in Fig. 4. Temperature /time series were obtained at each of 

 30 depths at the inshore site, and at each of 10 depths at the offshore 

 site. The time series were of about 10 days duration. The sampling 

 interval was one minute in the inshore series , and one second in the 

 offshore series. This made 30 series of about 13,000 temperatures 

 and 10 series of nearly 900,000 temperatures to process. In addition 

 to the temperature /time series at fixed depths, 200 STD casts were 

 taken at intervals throughout the observation period. These casts 

 were taken to provide the temperature/depth information necessary to 

 transform temperature variations in the time series into depth variations. 



The first stage of the data processing consists of rearranging the 

 data and casting them into a format that allows for convenient manipu- 

 lation and examination. The technique of delineating vertical excursions 

 through the use of temperature variations depends on the fact that 

 potential temperature is a conservative property. For this case, the 

 vertical velocity of the fluid is given by 



393 



