SECT. 5] LONG-TERM VARIATIONS IN SEA-LEVEL 591 



known of these are probably those of Doodson (1928) and Groves (1955). 

 Integration of a tide chart by planimeter is also used by some authorities. In 

 order to avoid the labour involved in the preceding methods, "mean tide 

 level" is sometimes computed instead of mean sea-level, being defined as the 

 average of the observed high and low waters. For all tidal waters, however, 

 mean tide level must be considered a poor substitute for mean sea-level in that 

 the short-period tidal contributions are not adequately eliminated. All the fore- 

 going methods provide daily mean values of level, and Rossiter (1958) has given 

 a resume of their efficiency in reducing contributions from the shorter-period 

 tides. An account is also given of a labour-saving filter which operates on 

 heights at 3-hourly intervals with an efficiency equal to the averaging of 24 

 hourly heights. 



Monthly and annual values derived from the daily means are generally 

 obtained by simple averaging, and the results for a world-wide network of 

 stations are published at regular intervals by the Permanent Service for Mean 

 Sea Level in the Publications Scientifiques series of the International Association 

 of Physical Oceanography. 



Table I 



Maximum Contribution from Certain Tidal Constituents, 

 Expressed as Percentages of Their Amplitudes, to the 

 Means for (a) a 30-day month and (6) a 365-day year 



Mg N2 Ki Oi M4 Me Msf 



(a) 0.055 0.209 0.267 0.401 0.058 0.060 1.55 



(b) 0.035 0.005 0.000 0.072 0.023 0.008 1.27 



M2 : principal krnar semidiurnal tide, period 12.42 hours. 

 N2 : lunar elliptic semidiurnal tide, period 12.66 hours. 

 Ki : lunar declinational diurnal tide, period 23.93 hours. 

 Oi : lunar declinational diurnal tide, period 25.82 hours. 

 M4 : lunar quarter-diurnal tide, period 6.21 hours. 

 Mg : lunar sixth-diurnal tide, period 4.14 hours. 

 Msf: fortnightly tide, largely arising from shallow water 

 theory, period 14.8 days. 



It has been customary to determine monthly and annual means to 0.001 ft or 

 1 mm according to the units employed, but this tends to give a false impression 

 of accuracy. Table I illustrates the maximum contributions ^ of certain tidal 

 constituents, expressed as a percentage of their amplitudes, to the mean of a 

 30-day month and a 365-day year, using a simple average of 24-hourly heights 

 to the day. It will be seen that for a moderate amplitude (5 ft) of M2, the 

 principal lunar semidiurnal constituent, there will be maximum contributions 



1 These contributions will be associated with a cosine term, the argument of which will, 

 in general, take up all values between and 2tt. 



