54 



For diurnal components, the length of the component hour, r, is 

 15° /a, for semidiurnal components 30° /a, and so on. The values of 

 the augmenting factor are then: 



Diurnal components, 94~~- — 700^/ = 1 .00286 



2x 



Semidiurnal components, th — ■ — tfo =1.01152 



^ ' 24 sm 15 



M3, 24sin22°30- = ^-°^'^'^ 



^'- 24stn30- ='°^^^° 



Ma, 24sl%5° -=^-"°^" 



^- 2^0^=1-20920 



99. A review of the process by which the amplitude and initial 

 phase of each component are found (par. 94), shows that the appli- 

 cation of the augmenting factor to the hourly component heights 

 (above mean sea level) will increase the amplitude in the same ratio, 

 but will not affect the initial phase. The augmenting factor is there- 

 fore applied directly to the computed amplitude. The application 

 of this correction to the amplitude of the M2 component at Sitka, 

 for example, gives a corrected value of 3.391X1.01152 = 3.430 feet. 

 Evidently, no augmenting factor should be applied to the S com- 

 ponents. The more complicated factors for computations from 

 secondary stencils are given in the Manual of Harmonic Analysis 

 of Tides. 



100. Elimination. — The hourly component heights derived from 

 the process of averaging that has been described will contain the 

 residuals of components other than that sought. After a first deter- 

 mination has been made of the amplitudes and initial phases of the 

 several components, corrections may be computed from them to 

 eliminate from each the effect of the other components. The process 

 is explained in the Manual of Tides, but is not of sufficient general 

 interest to be here included. 



101. Long period components. — The components listed in para- 

 graphs 75 and 76 include 2 having a fortnightly, 1 a monthly, 1 a 

 semiannual, and 1 an annual period. The first 3 of these are too small 

 to be of much importance, but periodic meteorological causes may 

 produce substantial annual and semiannual variations in the sea 

 level. Since a long period component does not change appreciably 

 during a calendar day, the daily averages of the observed tidal heights, 

 instead of the hourly heights, may be used for its determination, or 

 the daily sums may as well be used, the final result being divided by" 



