SECT. 6 J TURBULENCK 813 



a tidal current in Uraga Strait at the entrance to Tokyo Bay. Fluctuations 

 attributable to turbulence were recorded with periods from a few seconds up to 

 several minutes. 



An application of the hot-wire technique, much used in wind-tunnel and 

 atmospheric work, to the study of turbulence in the sea has been described by 

 Patterson (1957). He reported observations on the high-frequency turbulence 

 in the flow behind obstructions in a tidal current. A development of the hot- 

 wire method was used in a study of the high-frequency end of the turbulence 

 spectrum by Grant, Moilliet and Stewart (1959). In place of the hot wire they 

 employed a probe consisting of a platinum film of 4 x 10"^ cm thickness and 

 maximum length 1 mm on the tip of a glass cone. The probe was mounted on 

 the nose of a heavy body streamed from the stern of a ship and observations 

 were made in Discovery Passage, British Columbia, in a depth of 60 m and tidal 

 current velocity of 100 cm/sec, corresponding to a Reynolds number of 4 x 10'^. 

 The response of the instrument extended to very high wave numbers but was 

 limited at the lower end by movements of the towed body. The authors regard 

 the spectrum of the u fluctuations, derived from a 30-min record, as reliable 

 for wave numbers k from 0.02 to 1 cm~i. Within this range the energy is very 

 nearly proportional to A;~5/3^ ag would be expected from the theory of local iso- 

 tropy in the inertial sub-range. Most of the dissipation of energy appears to 

 occur from fluctuations of A;> 1 cm~i with a peak at about k = ^ cm~i. It was 

 estimated that the measured portion of the energy spectrum contributed only 

 about a fifth of the total energy of w^ and that the fluctuations of A; > 0.02 cm~i 

 could not contribute significantly to the Reynolds stress. As far as the condi- 

 tions were comparable, this view is consistent with the results of Bowden and 

 Fairbairn, described above, which apply to the range k< 0.01 cm~i. 



A turbulence meter for measuring fluctuations of velocity and temperature, 

 as well as their mean values, has been developed by Kolesnikov et al. (1958). 

 The horizontal and vertical components of the velocity fluctuations are 

 measured by crossed hot wires, while a single hot wire measures the mean 

 velocity. Thermistors are used for recording the mean temperature and its 

 fluctuations. Observations with this equipment were made below the ice from 

 the drifting station "North Pole 4" in 1956. The intensity of the turbulence was 

 found to be high within 1 m of the lower surface of the drifting ice, but to fall 

 off fairly rapidly at greater distances. Correlation functions and spectra were 

 computed. 



An interesting example of the application of the turbulent flux method on a 

 much larger scale was provided by Ichiye (1957), who computed the momentum 

 transport across the Kuroshio from current observations made by GEK 

 (towed electrodes). Taking the a;-axis parallel to the direction of maximum 

 current in the Kuroshio and the ?/-axis across it, the components of mean flow 

 U, V and turbulent flow, u, v were found from the series of observations at 

 each GEK station. From these, the turbulent transport uv and the transport 

 UV due to the mean flow were computed. The results showed that in most 

 cases the turbulent transport of momentum was large compared with the mean 



