Fróðskaparrit - 01.01.1975, Side 88

96
Sea Level Fluctuations in Tórshavn, Preliminary Results
cm/mb the regression analysis yielded a value of —0.73 cm/mb.
A possible explanation for the discrepancy emerges from table
V, which lists results of regression analysis performed for each
month of 1973 separately. In ali but one of the cases the baro-
meter factor is significantly different from the theoretical one
at the 5 °/o level and this one case is for the month with the
minimal standard deviation of pressure. There is a tendency
for the discrepancy to grow with the pressure variation as fig.
10 demonstrates and this is interpreted as follows. With large
variation in pressure — hence large st. dev. — a comparatively
larger amoúnt of the variation may be too rapid for the sea
level to be able to respond, resulting in a lower barometer factor
than theoretical.
A closer investigation into this hypothesis is in progress and
will together with a more detailed description of the low
frequency spectrum be presented elsewhere.
Conclusions
Although no actual comparison has been made with other
methods, it has been demonstrated, that the Maximum Entropy
Method is well suited to perform quantitative anlysis of the
astronomic tides.
The tabulation of harmonic constants permits prediction of
the astronomic tide of Tórshavn quite accurately. The pro-
minence of higher harmonics in the tide indicate strong non-
linearities presumably due to frictional effects in the currents.
The nonharmonic part of sea level fluctuation seems to be
mostly due to air pressure variations, however the sea is not
found to be able to respond to rapid pressure fluctuations.
Acknowledgements
Most of this work has been done while the author was
recieving a grant from Nordisk Kollegium for Oceanografi and
on the Institute of Physical Oceanography at the University of
Copenhagen headed by Prof. Jerlov. The inspiration of Dr