Fróðskaparrit - 01.01.1993, Qupperneq 150

154
MEASUREMENTS OF SEA BOTTOM TEMPERATURES
ments which were considered more accu-
rate; but the consistency with the Rolland
Black values decreases the probability of
erroneous values considerably.
From 1906 until August 8th 1949 the re-
sistance measurements were made in a test-
house just by the landingpoint in Tórshavn.
When the landing point was moved at the
end of this period, the measuring system
was installed at the telegraph station in Tór-
shavn and for the rest of the period mea-
surements were made there. Four land ca-
bles connected the Telegraph station to the
landing point, one for each of the sea-ca-
bles and two spare cables which were
grounded at the landing point. During each
cable measurement the resistances of the
two spare cables were measured also and
the deviations used to correct the resis-
tances of the other two land cables. These
were then subtracted from the seacable re-
sistances. Thus the resistance values pre-
sented are for only the marine part of the
cables also after 1949.
Temperature calculation
In the temperature ranges involved the re-
sistance of a copper conductor increases al-
most linearly with temperature. If Rt is the
measured resistance of such a conductor at
a fixed temperature t and R24 is the stan-
dard resistance at 24°C then the tempera-
ture will to sufficient accuracy be given by:
t = 24°C - (R24 - Rt)/(a*Rt)
where the coefficient a for the cables used
was 0.0041 (°C)"‘. The linearity further-
more implies that if there are temperature
variations along the cable - which was cer-
tainly the case here - then the temperature
deduced by the above formula should be a
fair representation of the average tempera-
ture of the whole cable.
In the start of the period the resistances
were reported in BA Ohm (British Associ-
ation, 1864) and the standard resistance
was referred to 75° Fahrenheit (= 23.89°C).
From Febr. 2.nd 1913 Intemational Ohms
were used and 24 °C used as standard tem-
perature. In the tables presented here the
values before 1913 have all been converted
to Intemational Ohms and referred to 24°C.
Results
The measured cable resistances are shown
in Table 1 and Table 2 together with infor-
mation on cable lengths and standard resis-
tances from the splicing lists and calculated
cable temperature. The measured resis-
tances and temperatures calculated from
them are copied from the SNTS joumal
(S.N.T.S joumal for elektriske málinger)
but all temperatures have been recalculated
for checking purposes. The temperatures
are furthermore plotted in Fig. 3 and Fig. 4
as timeseries for the whole period.
Discussion
The regular seasonal variations in the cal-
culated temperatures, evident in Figs. 3 and
4 for most years, are indications that the
values do reflect sea temperature to some
extent. This does not guarantee long-term
accuracy, however. Hansen et al. (1994)