Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: Bulges

The planets DO have an influence on the ocean tides
when their frequencies are a match for an ocean basin -- It is true that
NOAA and the like ignore this. Of course long before this question is
considered, they really ignore the whole issue and do a Fourier transform on
the real raw water height data for their predictions.

I doubt that. I think the component frequencies associated with the
planets' tidal forces are indistinguishable from the solar term, namely
one cycle per 24 hours. Besides that, I think that they are too small to
be important, or even detectable.

Tide tables are not generated by doing fourier transforms on tidal data.
They are generated by doing harmonic analysis, quite a different process.
In that scheme the allowed component frequencies are specified (they are
well known) and the height of the water as a function of time is
synthesized as a linear combination of terms having those frequencies.
The amplitudes and phases of the individual components are then adjusted
to give a best fit to data by some criterion, say least squares deviation
minimization.

Harmonic analysis used to be carried out on machines, analog computers
whose wheels were geared together and fitted with cams of adjustable
radius and phase. The fitting, I think, was done by a man in the loop, by
art rather than by computation. I saw one of these wonderful Spirographs
in a museum somewhere, all spotfaced and knurled brass, and looking quite
spectacular. Those 19th century computer manufacturers had class.

Now one might suspect that a fourier transform would yield these
characteristic frequencies without having to specify them *ab initio*,
and that might well be true on a planet where tides are driven only by
gravitational forces. Tides on Earth, however, are strongly dependent on
atmospheric pressure and winds. The noise generated by these influences
is too great to make any fourier transform method pracical. Tide tables
do not exhibit astronomical accuracy, of course.

Leigh