FISHERY BULLETIN: VOL. 70, NO. 3 



is 10.6 cm, or 20 9f of the long-term mean sea 

 level difference of 52 cm. 



The seasonal cycle for Hilo is very much like 

 that at Honolulu. The slight peak in April at 

 Hilo, which does not occur at Honolulu, combined 

 with the April minimum at Avila causes the 

 maximum in the sea level difference to occur 

 1 month earlier for Hilo- Avila than for Honolulu- 

 San Francisco. The normal seasonal cycle at 

 Avila tends to lead that at San Francisco by 1 

 month. A stronger rise occurs at Avila in June- 

 August than at San Francisco, which may be a 

 heating effect because Avila is at a lower latitude. 



DISCUSSION 



The characteristics of the normal seasonal 

 cycles of sea level difference for both pairs of 

 stations are in agreement with each other and 

 in accord with our knowledge from other sources 

 of the oceanography of the northeast Pacific and 

 California Current. The geostrophic equation 

 requires that the speed of the surface current 

 be proportional to the transverse slope of the 

 sea surface. From the sea level difference we 

 can infer that the average current normal to 

 the great circles between the Hawaiian Islands 

 and the California coast is southeastward 

 throughout the year. It is strongest in the 

 upwelling and summer season when the North 

 Pacific atmospheric high pressure cell is best de- 

 veloped. It is weakest during the winter season. 



Using an average latitude between stations 

 and the long-term differences in the geostrophic 

 equation, the average current to the southeast 

 is 2.0 cm/sec for the Honolulu-San Francisco 

 difference and 1.9 cm/sec for the Hilo-Avila dif- 

 ference. The range of the normal seasonal cycle, 

 as noted earlier, is 20 ^/c . 



It must be kept in mind that the sea level dif- 

 ference indicates only the average geostrophic 

 surface current normal to the line between the 

 two stations. For the California Current the 

 contours of 0/500 db dynamic height anomaly 

 in the CalCOFI Atlas No. 4, Geostrophic Flow 

 (Wyllie, 1966) indicate that the speed averaged 

 over a distance of 1,000 km is about 5 to 10 cm/ 



sec, whereas there may be narrow regions about 

 100 km in width with current of 25 cm/sec, or 

 greater. 



COMPARISON OF MONTHLY SEA 

 LEVELS WITH STERIC LEVELS 



Monthly sea level differences are to be used 

 to estimate the slope of the sea surface and draw 

 inferences regarding month-to-month and year- 

 to-year changes in current from the geostrophic 

 equation. Traditionally the slope is estimated 

 indirectly by computing geopotential heights, or 

 steric levels. Therefore, the agreement between 

 monthly sea levels and steric levels at nearby 

 offshore locations was investigated using hydro- 

 graphic station data, employed earlier to estab- 

 lish the long-term mean differences in sea level 

 between pairs of stations. Correlation coeffi- 

 cients between the monthly sea levels, linearly 

 interpolated to the date of each hydrographic 

 station, and the steric levels are given in the last 

 column of Table 2. For San Francisco and Avila, 

 where a large number of hydrographic stations 

 were made within a small area at nearby loca- 

 tions, the correlation coefficients are 0.54 and 

 0.57 respectively. Both coefficients are consider- 

 ably higher than the 1 % level of significance. 



For Honolulu the correlations from four sets 

 of data from different areas are quite consistent, 

 ranging from 0.54 to 0.65. These coefficients 

 are as large as those for San Francisco and Avila, 

 but because of the smaller sample sizes are sta- 

 tistically significant only at the 5% level. 



For Hilo the correlations for three different 

 sets of steric levels are less consistent. A high 

 correlation, r = 0.78, which is significant at the 

 1% level, was obtained with Hilo monthly sea 

 levels only from the set of 10 steric levels from 

 observations in 1-degree squares 04 and 05 im- 

 mediately northeast of Hilo, Dynamic topo- 

 graphy charts in TWZO data reports (Charnell 

 et al., 1967a, b, c, d, e, f ) reveal that the presence 

 of an anticyclonic eddy or ridge northeast of 

 Hilo at TWZO station 14 during 5 of the 16 

 observational periods caused the lower correla- 

 tion of data from that station with Hilo sea lev- 

 els. One suspects some similar local effect as 



626 



