

5«V 









3* 



4# \ 



\\ 







2« 



^ 



1 t 







1* 



125 120 115 



■%r 



4i 90 OS ISO 22S ;tq : 



-" 



135 180 IIS : 



IX 



* 



Macr.tuOe 





40- 





£ 20 





1 





10 





1, , , 





6 6 " 45 " 90 



^ 



ISO 22S 270 315 



Figure 9. — Relative frequency surfaces and frequency histograms for June. Data are shown for the 1-degree squares labelled 1 through 5 in the 

 upper left inset. Contours of relative frequency are drawn at intervals of 2.5%. Mean vector wind stress is indicated by an arrow. 



months. A slight shift in direction and magnitude of the 

 wind stress is indicated between summer and winter. 

 The mean stress in December is directed more toward the 

 south, and the magnitude has decreased. 



The frequency histograms also show these features. 

 During June, the direction histogram is characteris- 

 tically narrow-banded. Over 70% of the observations may 

 be concentrated within three direction intervals. In 

 December, the observations tend to be spread over a 

 wider range of directions. The histograms are broader 

 and the peaks in direction are less well-defined. Peak 

 magnitudes in June are generally one class interval larger 



than peak magnitudes in December. However, the win- 

 ter distributions are characterized by an increase in 

 relative frequency at high values of wind stress mag- 

 nitude. 



A similar pattern of contrasts between summer and 

 winter distributions is apparent in the northern section 

 of the grid. As shown in Figures 11 and 12, the direction 

 histograms are generally well-defined during the sum- 

 mer, although bimodal distributions are evident. Dur- 

 ing winter, observations are nearly uniformly spread 

 among all directions. There is a lack of consistency in the 

 frequency surfaces in Figure 12. Direction histograms for 



15 



