Table 2. — Dates for which daily averages represent lewer than the total four possible 

 synoptic samplings. The times (Pacific Standard Time ) of the missing samplings are in- 

 dicated to the right of each date. 



tion is that although the 6-hourly upwelling indices may be 

 satisfactory indicators of temporal variations at individual 

 locations, they g^ve a distorted picture of spatial variations. 



DISCUSSION 



A conspicuous feature of the charts in the appendix is the 

 apparent variability of the upwelling process. A substantial 

 component of this variability is the seasonal cycle, the major 



leaiures oi wnich appear to be repeated from year to year. 

 Clearly evident on a much shorter time scale, characteristi- 

 cally of one to several days, are individual "events" which 

 seem randomly superimposed on the seasonal variation. 

 Between these extremes of scale lie features which appear 

 as clusters of events. These are the subject of further study 

 and will be reported elsewhere (e.g., Bakun 1974) as will the 

 diurnal periodicities which exist in the 6-hourly data but do 

 not appear in the daily means presented here. 



