may be large. For example, from August to 

 November 1963 the average Xx was -0.28 dyne 

 cm."^ and for the same months of 1964 it was 

 -0.56 dyne cm."^ . 



Bjerknes (1969) drew attention to the large 

 differences in central Pacific equatorial tem- 

 peratures during autumn 1963, winter and au- 

 tumn 1964, and winter 1965 and discussed the 

 significance of the equatorial temperature anom- 

 alies in relation to the atmospheric circulation. 

 The temperature at lat. 2° N., long. 157° W. for 

 December was 27.1° C. in 1963 and 25.3° C. in 

 1964 (Seckel, 1970). How much of the tem- 

 perature difference can be attributed to the dif- 

 ference in equatorial divergence as a result of 

 the changes in wind stress has not been deter- 

 mined. The lower temperature in December 

 1964, however, is consistent with the increase 

 in magnitude of Tx during August to November 

 1964. 



A consistent seasonal pattern of change in Xy 

 is apparent from figure 2 for both study years 

 and for the SIO climatic year and reflects the 

 seasonal shift in the latitude of the intertropi- 

 cal wind convergence. Again, year-to-year 

 changes were relatively large. From July 1963 

 to February 1964 the average Xy was 0.05 dyne 

 cm."^ , and for the same months of the following 

 year it was 0.22 dyne cm."^ . An increase in the 

 meridional stress resulting from a northward 

 shift of the intertropical convergence should be 

 accompanied by a decrease in the vapor pres- 

 sure of the air. For lat. 2° N., long. 157° W. 

 the average vapor pressure from July 1963 to 

 February 1964 was 28.6 mb. and for the same 

 months 1 year later it was 26.3 mb. Thus, de- 

 spite the small magnitudes of Xy , the year-to- 

 year change was consistent with the change of 

 the independently measured vapor pressure. 



The large changes in the zonal wind stress 

 component in the trade wind zone example, 

 combined with a phase shift of about 1 year, 

 have been mentioned in the previous section. 

 A dominant seasonal change during the 2 study 

 years or the SIO climatic year is not apparent 

 from figure 3. Year-to-year changes, however, 

 are pronounced. The mean value of Xx for Au- 

 gust to December was -0.95 in 1963 and -1.71 

 dynes cm.'^ in 1964. The mean value for Janu- 

 ary to March was -1.76 in 1964 and -0.90 dynes 

 cm."^ in 1965. The results during winter and 

 spring 1964 reflect high trade wind velocities 

 and correspond with high autumn 1963, winter 

 1964 equatorial temperatures as described by 

 Bjerknes (1969). 



To investigate whether the variations in Xx 

 during the 2 study years were anomalous,! used 

 transport computations for the North Pacific 

 Ocean of the Pacific Oceanographic Group, Na- 

 naimo, Canada. The zonal and meridional com- 

 ponents of Ekman transport were calculated 

 from monthly mean sea-level pressure charts. 

 From the components of Ekman transport the 

 components of wind stress can be calculated 

 (wind stresses were not given in the tabulations). 

 Because of the limits of the sea-level pressure 

 charts for the Northern Hemisphere, the results 

 of the transport computations do not extend south- 

 ward to the area of the trade wind zone example 

 considered here. The zonal component of wind 

 stress for each month of 1956 and 1957 was 

 therefore computed at lat. 20° N., long. 150° W. 

 (Fofonoff, 1960a, 1960b^) and presented in fig- 

 ure 5. Fofonoff discussed the limitations of the 

 method to compute Ekman transports. The 

 computations are also based on data primarily 

 from an area north of the maximum trades so 

 that a seasonal trend in the wind stresses is 

 apparent in figure 5. For the intended compar- 

 ison, however, note that the absolute magnitude 

 of the wind stress was high in February and 

 March in 1956 but low 1 year later. In Novem- 

 ber and December the magnitude was low in 

 1956. but high in the following year. Although 

 the regularity of high and low Xx values in alter- 

 nating years is not known, the results of the 2 

 study years, 1963 to 1965, are not anomalous in 

 light of Fofonotf's transport computations. 



A seasonal trend in Xy is not apparent from 

 figure 3, except that during both study years low 

 magnitudes during December and January were 

 followed by high magnitudes during February. 

 These relatively large changes are not reflected 

 by the SIO climatic values but are qualitatively 

 consistent with monthly mean sea-level pres- 

 sure distributions.^ 



Fofonoff, N. P. 1960a. Transport computa- 

 tions for the North Pacific Ocean, 1956. Fish. 

 Res. Bd. Can., Manuscript Rep. Ser. (Oceanogr. 

 Limnol-. ) 78, 9 pp. 



1960b. Transport computations for the North 

 Pacific Ocean, 1957. Fish. Res. Bd. Can., Manu- 

 script Rep. Ser. (Oceanogr. Limnol.) 79, 9 pp. 



Northern Hemisphere charts of mean sea-level 

 atmospheric pressure. Extended Forecast Division, 

 National Meteorological Center, Environmental 

 Science Services Administration. 



