SECT. 2] 



SMALL-SCALE INTERACTIONS 



55 



steerage way and at a liigher s])eed alternately. In the absence of a "hull effect" 

 the increased ship speed should produce equal increments to the anemometer 

 speeds at all heights ; insofar as this was not the case, hull effect corrections 

 could be calculated. This showed that, even at 5 m forward of the stem of this 

 small vessel, the wind speed was decreased by some 2 or 3% at the lower 

 heights. In view of the small magnitude of the wind gradient, corrections of 



1000 



4 6 



Wind speed , m/sec 



Fig. 3. Wind-profiles over shallow water. (After Roll, 1948, Fig. 5.) 



several per cent are undesirably large and undoubtedly detract from the 

 accuracy obtainable with this method. The work supported Roll's (1949) 

 suspicions of earlier profiles involving measurements on ships. 



A special buoy with a 9-m light-alloy mast has been devised by Brocks (1959) 

 and this appears to be a considerable advance on anything used previously. 

 The general arrangement is shown in Fig. 4. It has guiding fins to maintain 

 orientation with respect to the wind and damping plates to reduce vertical 

 motion. The buoy is towed to the observation area where, by means of cables 

 or radio signals, the measurements are transmitted to recording instruments on 

 board ship. With the ship drifting, the buoy automatically takes station to 

 windward so that undisturbed observations are secured. In waves 3 m high the 

 maximum inclination of the mast was only 10° from the vertical. Some speci- 

 men profiles obtained over the North Sea by Brocks are shown in Fig. 5. 



Where rolling cannot be eliminated entirely, measurement of the root-mean- 

 square rolling amplitude, a, and period, T, permits correction to be made to 

 the recorded wind speed, Ur, given by a cup anemometer using the following 

 formula (Deacon et at., 1956) 



UrjUc = l+[iThal{TUr)]^, 



