SAUR: SEA LEVEL DIFFERENCES 



b=Ol7cm/yeor 



5(5=^0023 cm/yeoi 



SAN FRANCISCO 

 b=020 cm/year 

 St,= 4002l cm/year 



1950 I960 1970 



HILO 



b=0 57 cm/yeor 



S[,= tOOecm/yeor 



AVILfl 



b=024 cm/yeor 



Stj=-0 10 cm/yeor 



1940 1950 1960 



1970 1940 

 YEAR 



1950 



I960 



Figure 2. — Changes in annual mean sea level at four 

 tide stations and linear trends determined by least 

 squares regression. Slope of regression, b, and standard 

 error of regression, s^, are shown. 



fore, an adjustment for this "inverted barome- 

 ter" effect is made by correcting the sea levels 

 to a "normal" atmospheric pressure which com- 

 pensates for both the normal seasonal cycle and 

 the monthly anomaly of atmospheric pressure 

 at the station. For our purpose a correction of 

 1 cm in sea level for 1 mb change in atmospheric 

 pressure is sufficiently accurate. Adjustments 

 of less than 1 cm for monthly deviations of mean 

 atmospheric pressure over the world oceans re- 

 ported by Pattullo et al. (1955) have not been 

 made to the individual station data reported 

 here, because they are small and would have no 

 effect on sea level differences. 



Monthly mean atmospheric pressure reduced 

 to sea level from weather observations of the 

 National Weather Service at Honolulu, San 

 Francisco, Hilo, and Santa Maria (less than 20 

 nautical miles from Avila) were obtained mainly 

 from published sources. The World Weather 

 Record series (Clayton, 1944a, 1944b; Clayton 

 and Clayton, 1947; U.S. Weather Bureau, 1959, 

 1965, 1968) contain climatological data, includ- 



ing monthly atmospheric pressure, through 1960. 

 Monthly mean pressures for first order weather 

 stations are included in the monthly issues of 

 Climatological Data, National Summary, U.S. 

 Environmental Data Service, published since 

 January 1950. Some manuscript records ob- 

 tained directly from the Weather Service offices 

 were helpful in standardizing all pressure data 

 to sea level. 



NORMAL SEASONAL CYCLES OF SEA 

 LEVEL AND SEA LEVEL DIFFERENCE 



The normal monthly sea levels are obtained 

 by averaging the monthly sea levels" for a given 

 month from all years of record. Because the 

 zero level for the scale on the tide gauge is ar- 

 bitrary at each station the normal monthly sea 

 levels are shown in Table 1 in terms of departure 

 from the long-term mean for all months and 

 years. The standard deviations for given months 

 in Table 1 are measures of the year-to-year vari- 

 ability of monthly sea level for the given month. 

 From these one can compute that the 95% con- 

 fidence limits on the monthly normals for Hono- 

 lulu and San Francisco are 0.8 cm or less, and 

 1.4 cm or less for Hilo and Avila. 



We are interested in the real differences in 

 normal monthly sea level between stations as a 

 base against which to measure the variability. 

 For continental stations it is possible to deter- 

 mine this difference of long-term mean sea level 

 between two stations by reference to the geodetic 

 leveling network,' but no such reference leveling 

 exists between the continent and Hawaiian 

 Islands. As described in the next section, the 

 long-term sea level difference has been estimated 

 using oceanographic data. 



COMPUTATION OF LONG-TERM MEAN 

 SEA LEVEL DIFFERENCE 



The long-term mean sea level differences be- 

 tween Honolulu and San Francisco and between 



" Monthly sea level with trend removed and adjusted 

 to normal atmospheric pressure is hereafter implied. 



* From oceanographic data, however, Sturges (1966) 

 suggests there may be some systematic north-south lev- 

 eling error. 



623 



