s 

 o 



in -10 



ANNUAL 

 MEAN 



Figure 5. — Zonal component of wind stress, lat. 

 20° N. , long. 150° W. , January 1956 to Decem- 

 ber 1957, derived from meridional Ekman trans- 

 ports computed by Fofonoff. 



In the example from the northern zone (fig. 4), 

 the pronounced seasonal changes in Tx are con- 

 sistent with the shifts in latitude of the North 

 Pacific pressure ridge. Most striking are the 

 extreme values of December 1963 and May 1964. 

 The average of the t^ for December 1963 to Jan- 

 uary 1964 is 2.00 dynes cm."^ and for December 

 1964 to January 1965 is 0.68 dyne cm."^ . The 

 large differences are again consistent with the 

 statements by Bjerknes (1969) relating high 

 equatorial temperatures, autumn 1963 and win- 

 ter 1964, with the strong westerlies in midlati- 

 tudes. 



The average value of t^ for March to May 



1964 is -0.79 dyne cm."^ and for March to May 



1965 is 0.06 dyne cm.""^. Qualitatively, the large 

 year-to-year differences during both the autumn 

 and spring seasons are consistent with monthly 

 mean sea-level pressure distributions. 



The large fluctuations of Ty (fig. 4) also re- 

 flect changes in the North Pacific high-pressure 

 distribution associated with frequent changes in 

 the direction of the meridional component of the 

 wind. Seasonally, from late autumn to early 

 spring, the prevailing wind has a southerly 

 component. During the study period the aver- 

 age Ty for January to February 1964 was 0.84 

 dyne cm."^ and for the same month in 1965, 0.14 

 dyne cm."^ . With the vapor pressure of the air 

 increasing southward it is reasonable to expect 

 higher vapor pressures of the air during winds 



with a stronger component from the south. The 

 average vapor pressure for the same 2 months 

 was 16.2 mb. in 1964 and 13.0 mb. in 1965 

 (Seckel, 1970). This example, like that in the 

 equatorial zone, confirms the consistency of 

 the wind data with the independently measured 

 vapor pressure. 



CONCLUSION 



The foregoing evaluations have shown that 

 despite the limitations of the marine surface 

 meteorological observations, meaningful mea- 

 sures of the month-to-month and year-to-year 

 changes in the wind stress have been obtained 

 for a large portion of the North Pacific trade 

 wind region. The results presented in table B, 

 therefore, satisfy the needs of the TWZO in- 

 vestigation. 



Beyond this application, the need for monthly 

 measures of the wind stress will continue and 

 increase as oceanographers begin to forecast 

 changes in the distribution of ocean properties 

 and meteorologists develop circulation models 

 to begin global weather forecasting. The exper- 

 iences gained in the calculation of large-scale 

 sea-air interactions from marine surface me- 

 teorological observations may therefore be of 

 help in future work. The manner in which the 

 distribution and quality of marine meteorologi- 

 cal data can be improved and processing can be 

 facilitated has already been suggested by Seckel 

 (1970). Additional suggestions concerning wind 

 observations follow. 



Marine surface meteorological observations 

 will continue to be the mainstay of data for the 

 computation of large-scale, sea-air interactions. 

 Improvement in the quality of data is therefore 

 of concern and can be achieved by proper place- 

 ment of instruments and by proper techniques 

 of observation. The quality of wind data can 

 also be improved by avoiding erroneous vector 

 addition of the "apparent" wind and the speed 

 and direction of the ship. At present the appar- 

 ent wind (observed anemometer speed and di- 

 rection) is not reported in the ship's weather 

 observations form (U.S. Department of Com- 

 merce, ESSA form 72-1). If this observation is 

 reported on the weather form, the wind speed 

 and direction can be calculated by computer. 



In the region under consideration in this pa- 

 per, the number of wind observations between 

 the Equator and lat. 15°N. is inadequate. Better 

 wind information in this latitude band is needed 

 by the oceanographer to compute the curl of the 



