DIRECT CURRENT MEASUREMENTS 



Since the early 1930's the International Ice 

 Patrol has been conducting oceanographic sur- 

 veys in the area of the Grand Banks with the 

 objective of computing surface currents for ice- 

 berg drift determination. The method used to 

 compute these currents has relied on the baro- 

 clinic geostrophic assiimptions. Chief among 

 these assumptions is that a level of no motion 

 exists and that the current is a balance of the 

 Coriolis force and the baroclinic pressure field. 

 International Ice Patrol has used the 1,000 deci- 

 bar (meter) surface as the reference surface 

 since 1932. 



In an effort to evaluate the validity of the 

 assumption that there was no motion at 1,000 

 decibars, Ice Patrol initiated a direct current 

 measuring program to determine the current ve- 

 locities under the Labrador Current. In 1973 

 two moored current meters, using subsurface 

 floats were deployed near the bottom under the 

 Labrador Current. Geodyne Model 850 current 

 meters were positioned approximately 40 meters 

 above the bottom (see Table 2). The meters 

 were set to record in the interval mode with a 

 burst of 15 samples every 15 minutes. 



The first meter was set at a depth of 879 meters 

 at 44°13'N and 48°52'W on April 13th and re- 

 covered on May 14th of 1973. The second meter 

 was set on 18 May at 951 meters depth at 45°33'N 

 and 48°16'"W and was recovered 30 June 1973. 



The data records were complete for both meters 

 and all components appeared to have functioned 

 properly. 



A simple average was taken of the measure- 

 ments in each 15 minute interval. The averaged 

 raw data show velocities as high as 25 cm S'' for 

 both meters. These data were then smoothed 

 using a numerical filter to effectively remove fre- 

 quencies higher than 0.5 cycles/hour, thus reduc- 

 ing noise in the data. The current measurements 

 were separated into two components; one tangen- 

 tial to the slope of the Grand Banks and the 

 other normal to the slope. Positive values indi- 



cated northward flow for the tangential compo- 

 nent and eastward flow for the normal component. 



The smoothed data were further treated using 

 a numerical filter to remove frequencies higher 

 than 0.8 cycles/day or periods shorter than 30 

 hours. This removed the diurnal and semidiurnal 

 tidal currents. A difference between the smoothed 

 data and the filtered data was calculated, which 

 contained pei'iods of approximately 2 to 30 hours. 



These filtered data for the first current meter 

 mooring at 44°13'N and 48°52'W show a surpris- 

 ing degree of variability. The current ranged 

 from 14 cm s"' southward along the shelf to a 

 quite unexpected current reversal of about 5 days 

 duration, with a northward flow of up to 2 cm s"^ 

 The mean current for this period was 3.4 cm s"^ 

 toward the south. The component normal to the 

 shelf was virtually zero. The periodic residuals 

 presented a rather complex picture. The stand- 

 ard deviations from the mean for the normal and 

 tangential components were ±2.2 cm S'' and 

 ±2.8 cm s"^ respectively. The only identifiable 

 periodic component was the lunar semidiurnal 

 tide. Harmonic analysis indicated that there was 

 a 17% probability for the normal component that 

 the coefficients for this frequency could have been 

 generated from random data and a 32% prob- 

 ability for the tangential. This rather high 

 probability appears to stem from the fact that 

 any baroclinic tides present are intermittent and 

 not "phase-locked" to the forcing uf the equi- 

 librium tide. 



For the other current meter at 45°33'N and 

 48°16'W the aperiodic current varied from 15 

 cm s"^ southwestward along the banks to a 1 

 cm s"^ northward flow that persisted for approxi- 

 mately one day. The mean current was 5.3 cm S'^ 

 toward tlie southwest. Again the normal current 

 component was small in comparison with the 

 tangential component. 



The periodic residuals in this case had a stand- 

 ard deviation of ±3.0 cm s"^ and ±2.9 cm s"^ for 

 the normal and tangential components respec- 



