JULY AUG. SEPT OCT. NOV. DEC. JAN, FE6. MAR. APR. MAY JUNEI- 



ANNUAL 

 MEAN 



ANNUAL 

 MEAN 



Figure 4. — Zonal component (t^ ) left panel, and meridional component (Xy) right panel, of the wind 

 stress at lat. 32° N., long. 167° W. , July 1963 to June 1965, from table B, and climatic values 

 from Scripps Institution of Oceanography (19A8) . 



without a dominant seasonal trend, is consistent 

 with the results obtained from July 1963 to June 

 1965. 



The discrepancy of the mean annual values of 

 Tx in the trade wind zone is in marked contrast 

 to the good agreement in the other examples. 

 Also, the comparisons made by Malkus (1962) 

 lead to the conclusion that the large difference 

 of 0.46 dyne cm."^ in annual means is due to 

 real departures during the 2-year study period 

 from the long-term mean conditions rather than 

 to the computational differences used at SIO and 

 in this paper. 



Interseason and Interyear Comparisons 



It is clear from the preceding section that the 

 month-to-month changes of climatic means do 



not necessarily reflect the changes to be ex- 

 pected during any individual year. The mecha- 

 nisms producing changes in properties of the 

 ocean, such as evaporation or wind stress, may 

 also show much greater season-to-season or 

 year-to-year variations than would be expected 

 from climatic mean values. The changes in the 

 heat of evaporation and the net heat exchange 

 across the sea surface that occurred during 

 July 1963 to June 1965 have been illustrated by 

 Seckel (1970). Similar changes also occurred 

 in the wind stress. 



In the equatorial example (fig. 2), ix shows 

 no clear seasonal trend in either of the 2 study 

 years or in the SIO climatic year. Although the 

 absolute magnitude of the zonal stress compo- 

 nent is small, relative year-to-year changes 



