computed for frequencies with periods up to 

 41/2 days. 



The observed currents (37 and 81m. levels) 

 indicate a distinct pattern (iigs. 29a, 29c, and 

 30a) of periodic, clockwise motion of about 54 

 hours duration, 22 to 23 July and 28 to 30 July, 

 followed by a tendency toward backing of about 

 four days duration, 24 to 28 July and 2 to 5 

 August. Only two sequences of this pattern were 

 observed because of limited record length. 



The circulation observed at the 37 and 81m. 

 levels did not extend to the 124 m. level (fig. 

 31a). The observed current at the 124m. level 

 was much more persistent in its rapid clockwise 

 rotation with only one instance of slow backing, 

 24 to 28 July. Veering at the 124m. level oc- 

 curred from 22 to 23 July and from 29 July to 

 3 August, some eight days in all. Progressive 

 vector diagrams (figs. 32, 33, and 34) also very 

 clearly showed the veering for all the current 

 meters. The rotational periods during the veer- 

 ing intervals for all depths were approximately 

 semidiurnal. 



The resultant movement indicated by all the 

 progressive vector diagrams (37m. level, 6.0 

 cm. /sec toward 308° ; 81m. level, 6.4 cm./sec to- 

 ward 308° ; 124m. level, 3.0 cm./sec toward 

 290°) was northwestward. This was a surprising 

 result to the authors because the currents over 

 the Labrador continental shelf are commonly 

 believed to set southeastward, excepting some 

 relatively narrow filaments of northwestward 

 flowing water. 



Geostrophic velocities computed using concur- 

 rent station data from stations along standard 

 section Al, immediately after the arrays were 

 set and immediately prior to their retrieval, va- 

 ried from 0.0 to 4.0 cm./sec toward 320°. Thus 

 there was good agreement between the geostro- 

 phic velocity computed from the concurrent 

 oceanographic data and the average velocity 

 computed using the 37 and 81m. current meter 

 data. 



Notice that the magnitude of the geostrophic 

 velocity was consistently less than the observed 

 current speed. Also, the direction of the geo- 

 strophic velocity differed from the observed cur- 

 rent direction (figs. 29a, 29c, 30a, and 31a) by 

 45° frequently and 180° occasionally. These 

 differences between the geostrophic velocity and 

 the instantaneous observed current severely 



vitiate any current predictions obtained from 

 the relative density field in this area. 



All estimates of power density (figs. 35, 36, 

 and 37) show energy concentrated near the im- 

 portant semidiurnal tidal periods (12.42 hours 

 principal lunar component and 12.00 hours prin- 

 cipal solar component). This is not an unex- 

 pected result because of the 12 hour periodi- 

 cities apparent in the observed currents at 37 

 and 81m. (figs. 29b, 29d, and 30b). The move- 

 ment of the water at 124m. (fig. 31b) was more 

 tidal in nature than that at shallower depths. 

 The reason for this is not clear, but apparently 

 the processes occurring near the ocean's surface 

 are not in phase with the tidal currents. 



No confidence limits were computed for any 

 of these spectral estimates. Since they are only 

 estimates, the lines representing the spectral 

 amplitudes should be considered as the axes of 

 envelopes that would contain the true spectra. 

 The confidence limits (i.e., the width of the en- 

 velopes) are a function of frequency and will 

 be large at low frequencies. 



The spectrum of currents measured at 37 m. 

 at 54-29N, 54-30W (fig. 35a) has a broadened 

 peak from 14.22 to 12.19 hours. This broadening 

 is quite interesting because 



(1) The inertial period at 54N latitude is ap- 

 proximately 14.6 hours. 



(2) Currents measured at 37 m. at 54-30N, 

 54-32W (fig. 35b) approximately 2 miles away, 

 did not show this spectral broadening. 



It is relatively simple to explain heuristically 

 how the winds generate an inertial current. 

 First, consider a wind vector that is inertially 

 rotating. This wind would then generate an in- 

 ertial current that would continuously increase 

 in amplitude. This increase occurs because the 

 wind would always be adding momentum to the 

 water in the current direction. Notice that a 

 wind which blew in a fixed direction for one in- 

 ertial period would not generate an inertial 

 current because it would add and remove mo- 

 mentum in equal amounts during the inertial pe- 

 riod. "The features in the wind field that are 

 most efficient in changing the amplitude of in- 

 ertial currents are 



(1) A strong wind blowing in one direction 

 for a few hours up to half an inertial period. 



(2) A strong wind combined with a fairly 

 sudden shift in direction." (Pollard and Millard, 

 1970). 



8 



