Table 2.— Intercorrelation of monthly mean anomalies of sea level with various oceanic and 

 atmospheric variables at Monterey, Calif, (see text). Correlation coefficients enclosed in paren- 

 theses are not significant at 5% level. 







ADJ 





MERID 



ZONAL 



EKM 



SVP 







DYN 





SL 



SL 



PRESS 



WS 



WS 



TSPT 



TSPT 



SAL 



SST 



HT 



SL 



1.00 





















ADJSL 



.96 



1.00 



















PRESS 



-.70 



-.46 



1.00 

















MERID WS 



.43 



.42 



-.28 



1.00 















ZONAL WS 



(-.14) 



-.18 



(-.03) 



-.48 



1.00 













EKM TSPT 



-.42 



-.42 



.26 



-.99 



.59 



1.00 











SVP TSPT 



(.01) 



(.07) 



-.18 



-.33 



.15 



.32 



1.00 









SAL 



-.35 



-.31 



.29 



-.31 



(.07) 



.30 



.20 



1.00 







SST 



.61 



.65 



-.29 



.38 



-.17 



-.37 



(-.05) 



-.37 



1.00 





DYNHT 



.79 



.79 



-.46 



.25 



(-.08) 



-.25 



(.00) 



-.44 



.65 



1.00 



increases in SST are associated with increased sea levels and increased 

 salinities are associated with decreased sea levels. These relationships 

 are consistent with basic considerations of seawater density changes. 



Dynamic height (0/400 db) at the mid-Monterey hydrographic sta- 

 tion was found to be strongly correlated with sea level fluctuations at 

 Monterey. The correlation coefficient of 0/400 db dynamic height was 

 0.79 with Monterey sea level and was the highest of any of the varia- 

 bles tested. The higher correlation of sea level with dynamic height 

 than with SST (r = 0.61) suggests that subsurface fluctuations are 

 important in causing changes of both sea level and dynamic height at 

 Monterey. A possible cause of such subsurface fluctuations is the 

 northward propagating coastally trapped wave mentioned earlier. To 

 examine this, sea level at Talara, Peru, was used as an index of El Nino 

 conditions and was lagged to 10 mo for correlation with sea level at 

 Monterey. The correlation coefficient peaked at r = 0.37 at a lag of 6 

 mo. A wave propagating the approximately 6.300 km between Talara 

 and Monterey in 6 mo would have a phase speed of about 34 km/d. 

 This is somewhat lower than speeds reported by Enfield and Allen 

 f 1980) but not inconsistent with their results. 



Regression Analysis 



Table 3.— Results of multiple regression analysis of sea level at Monterey, Calif., 

 with various oceanic and atmospheric variables for entire year. Davidson Current, 

 and upwelling periods. Data series are sea level (SL) in centimeters, atmospheric 

 pressure (PRESS) in millibars, sea surface temperature (SST) in °C, meridional 

 wind stress (MWS) in dynes/cm 2 , and dynamic height (DYN HT) in centimeters. 



Step Variable 



Explained variance 



Increase in explained variance 



.62 



A. Entire year (Ian. -Dec. 77 mo of data) 



1 DYN HT .62 



2 PRESS .70 .08 



3 SST .74 .04 



4 MWS .76 .02 



Sea level = -0.057 + 0.470 DYN HT - 0.894 PRESS + 1 .208 SST + 4.491 MWS 



B. Da 



/idson Current period (Oct 



-Feb. 



31 mo ol'data) 







1 



DYNHT 





.76 





.76 





2 



MWS 





.82 





.06 







Sea level 



= -0.0653 + 



0.732 DYN HT + 



15.402 MWS 





C. Upwclling period (Apr 



-Aug. 



33 mo ol'data) 







1 



DYNHT 





.36 





.36 





2 



MWS 





.52 





.16 





3 



PRESS 





.60 





.08 





4 



SST 





.66 





.06 





Sea 



evel =-0.108 -0.935 PRESS 4 



0.256 DYNHT + 



4.256 MWS + 



1.271 SST 



We have seen that the monthly anomalies of sea level at Monterey 

 are significantly correlated with dynamic height, atmospheric pres- 

 sure, SST, meridional wind stress, offshore Ekman transport, and 

 surface salinity. To quantify these relationships, a multiple regression 

 analysis was performed using the BMDP2R stepwise multiple 

 regression program (Dixon 1975). Since fluctuations of meridional 

 wind stress and offshore Ekman transport are closely related (r = 

 0.99 in Table 2), use of both variables in a regression would cause 

 instabilities in the computation. Ekman transport was omitted from 

 the regressions and only the meridional wind stress considered since 

 the wind stress is the more fundamental variable. 



The results of the regression analysis for the entire year, presented 

 in Table 3 (Part A), show that dynamic height is the major predictor 

 of sea level, with atmospheric pressure, SST, and meridional wind 

 stress as second, third, and fourth predictors. The remaining varia- 

 bles explained only negligible portions of the variance and their coef- 

 ficients are not included in the table. Together, the four major 

 predictors explain over 76% of the variance of the monthly sea level 

 anomalies with dynamic height alone explaining 62% of the vari- 

 ance. Considering that the sea level was recorded hourly in a consis- 

 tent fashion while dynamic height was computed from observations 

 taken at scattered times by several institutions using different meth- 

 ods, the strength of the relation seems very good. 



The relationship between sea level and dynamic height was further 

 examined in a seasonal sense. There is good agreement in both phase 

 and amplitude of the long term monthly means of dynamic height and 

 adjusted sea level (Fig. 10). The observed seasonal cycle for 

 dynamic height is somewhat more variable than that of sea level, pos- 

 sibly as a result of limited sampling (there were only 12 or 13 stations 

 per month during winter but up to 24 stations per month the rest of the 

 year). The figure shows that both sea level and dynamic height near 

 Monterey are highest in winter and lowest in spring. 



Reid and Mantyla (1976), showed that south of lat. 40°N in the 

 eastern North Pacific Ocean sea levels are typically highest in late 

 summer and early fall and lowest in late winter as a result of annual 

 solar heating. North of lat. 40°N, however, sea levels are highest in 

 winter and lowest in summer; this pattern cannot be explained by the 

 steric response to seasonal heating and cooling. Using Sturges" 

 (1974) data from Neah Bay, Reid and Mantyla further demonstrated 

 that maximum sea levels occur in winter when inshore northward 

 flow is strongest and minimum sea levels occur during summer when 

 flow is southward, thus relating seasonal changes in sea level to geo- 

 strophically balanced flow. Monterey lies at lat. 36°N and has a sea- 

 sonal cycle that is intermediate between these regimes. 



15 



