Wr. Crawford, A TECHNIQUE FOR QUALITY-CONTROL AND SELECTION OF TIDAL HARMONIC CONSTITUENTS, The International hydrographic review, 72(2), 1995, pp. 135-150
This paper outlines a procedure to select an optimal set of tidal harm
onic constants for use in the prediction of tidal heights. We begin wi
th 18 sets of harmonic constants, each computed from one of 18 consecu
tive years of hourly tidal heights at a port using a standard tidal an
alysis program. These constituents are input to a spreadsheet in which
are computed average amplitude and phase, and standard deviations of
amplitude over the 18 years. As well, graphs of the amplitude and phas
e deviations for all constituents are presented to allow a user to ide
ntify any trend or cyclical variation in the amplitudes or phases, and
to identify years where the quality of measurement is poor. The stand
ard error of the average amplitude is computed and may be used as an a
ccept/reject criterion for inclusion of tidal constituents in subseque
nt predictions. Although nominally intended for 18 years of data, the
program will operate on a smaller number of years, and can be adapted
to include more years. Many tidal constituents of astronomical origin
have frequencies close to or identical with shallow water constituents
generated by the interaction of astronomical constituents. If the sha
llow water constituent interferes with the astronomical one, it will m
odulate the amplitude and phase of the astronomical constituent, often
with periods near 8 and/or 19 years. This modulation can be identifie
d ill. the time series plots of deviations of amplitudes and phases.