Fróðskaparrit - 01.01.1975, Side 86

94
Sea Level Fluctuations in Tórshavn, Preliminary Results
TABLE V The correlation between Sea level (reduced) and air-
pressure and the Barometer factor for separate months of 1973.
Also listed the standard deviation of air pressure and the sig-
nificance P } that is the probability of obtaining the observed
barometer factor under the assumption that it has the theoreti-
cal value -1.01 cm/mb.
PERIOD CORRELATION BAROM.FACTOR cm/mb ST.DEV.(air pr.) (mb)2 P
JAN 73 -0.948 -0.788 14.79 o.di
FEB " -0.886 -0.732 14.87 »
MAR " -0.903 -0.880 13.32 "
APR " -0.924 -0.720 12.13 »
MAY " -0.688 -0.489 9.36 »
JUN " -0.851 -0.920 8.67 »
JUL " -0.868 -1.024 5.91 Not Sign.
AUG " -0.931 -0.899 9.32 0.01
SEP » -0.941 -0.898 10.32
OCT " -0.858 -0.931 8.16 0.05
NOV » -0.879 -0.803 14.21 0.01
DEC » -0.798 -0.689 11.52 »
fig. 7 that of air pressure. These spectra have not been obtained
using the MEM, rather a Fast Fourier Transform has been
computed and squared. It is seen from the spectra that not all
the diurnal and semidiurnal influence has been removed from
the reduced water level (this may to some degree be due to the
fact that corrections for nodal modulation were not included
in the calculation of reduced water level). The air pressure
spectrum displays a peak at semidiurnal frequency which is
probably the S,, the influence of this on sea level must be
intrinsic in the amplitude and phase of S2 given in tables III
and IV.
Fig. 8 shows a plot of reduced water level and air pressure
for a specific month chosen at random from the data series
The interrelation is obvious. Taking all the available sea level
and air pressure data in 1973 a linear regression analysis was
performed between the two variables and plotted in fig. 9.
The correlation coefficient obtained was —0.82. Theoretically
a pressure increase of 1 mb should lower the sea level by 1.01
cm. In contrast to this theoretical barometer factor of —1.01