58 DEACON AND WEBB [CHAP. 3 



greater speeds, the surface is covered with ripples and these are expected to 

 increase the drag coefficient above that for a smooth surface. 



Roll's (1948) wind-profiles for the Neuwerk shoals given in Fig. 3 yield drag 

 coefficients which show little variation with wind speed and average about 

 25% more than the smooth-surface values. However, as the site was an un- 

 representative one, only the values for his two lowest wind-speed groups are 

 shown in Fig. 6 ; for these the effect of the shoal water would probably be 

 unimportant. 



The neutral profiles obtained by Takahashi (1958) in Kagoshima Bay give 

 drag coefficients for light winds averaging about 0.0012, in quite good agreement 

 with Roll's low wind-speed values. 



The observations of N. K. Johnson (Deacon, 1953) were made on a destroyer 

 in the English Channel. Anemometers were carried on the foremast (at a height 

 of 21.3 m) and, at lower levels down to 3.3 m, rigged 6 m forward of the stem 

 of the ship. With such an arrangement, any effect of the ship's hull on the air 

 flow should have been such as to give somewhat greater vertical gradients of 

 wind speed than the true values, so the mean drag coefficient from Johnson's 

 eight profiles, which is plotted in Fig. 6, is more likely to be somewhat larger 

 than otherwise. 



Deacon, Sheppard and Webb (1956) obtained wind-profiles over the height 

 range 2 to 13 m using a small diesel -schooner and were able to determine 

 corrections for the effect of the ship on the air flow. The drag coefficients for 

 31 near-neutral runs, mainly of 30-min duration, after grouping for wind 

 velocity, give the means shown in Fig. 6. The depth of water was 20 m or more 

 and the fetch of wind over water mainly between 20 and 40 km. 



The latest series of profile observations are those of Brocks, who used the 

 special buoy-mast arrangement of Fig. 4. The authors are indebted to Dr. 

 Brocks for the neutral drag coefficient values (shown in Fig. 6) which he kindly 

 made available prior to publication. As the result of two extensive trials in the 

 North Sea and Baltic, he finds surprisingly little variation in the drag coefficient 

 for neutral conditions over the range 2-14 m/sec.i 



1 Note added in proof. More recent wind-profile observations by Brocks (1961), Deacon 

 (1961) and Fleagle. Deardorff and Badgley (1958) also consistently indicate no significant 

 variation of neutral drag coefficient values with wind speed. For largely land-locked waters 

 the results are : 



Author Site Water Fetch, Wind range, c\q x 10^ 



depth, m km m/sec 



Brocks (1961) Kiel Bight 



Deacon (1961) Port Phillip 



Fleagle et al. (1958) East Sound 



The numbers of ()bser^'ations used in each case are given in brackets below the wind 

 range. iamthmed (ipposite. 



