Jökull - 01.12.1967, Qupperneq 71

recovered during the third pumping test with
the lower ð-value (— 62%0) is derived from an
aquifer at a depth of more than 900 m but
the water with the higher ð-value from an
aquifer at less depth. A correlation between
the water level in the well and the ð-value of
the water was observed. During the first pump-
ing test the ð-value was less than — 60%o with
the exception of the first sample that was
probably contaminated. During the second
pumping test when the water level was at a
depth of 35—38 m, the ð-value was consistently
Depth /77 Temperature °C Before pumping 14. tO'. 65 S °/oo Simptified geotogicat section. After pumpmg 2.// .'65 S °/oo
150 - 180 ■ /3 -53 S/l'/t/ f/U '7/7//. -55
777/ZZ7Í

.1
640 ■ - 42 - 62 -56

720 - . 46 - 63 1 -56
020 — — 53 -59

900 - - 57 - 62 Basa/t -62
Fig. 5. A section of the well. The ð-value of
the water before and after the third pumping
test ancl the temperature before that pumping
are shown in relation to depth. The dashed
part of the profile is cased.
Mynd 5. Alyndin sýnir snið af holunni, b-gildi
fyrir og eftir þriðju dœlingu og hita fyrir þá
dcelingu. Fóðurrör er i skástrikaða hluta sniðs-
ins.
— 55.5%c or higher. Summing up it appears
that when the water level is above 13 m, water
with a ð-value of — 60%o or less is recovered,
but when the water level Itas been lowered
to a depth of 35 m or more a new aquifer
opens in the well at less depth (see Fig. 5)
with a higher ð-value, probably — 'ó2°/00.
The chloride content of the water increased
with every pumping test. After the first pump-
ing test it was 8540 ppm, after the second 9124
ppm and after the third 9540 ppm. This in-
crease is correlated with the water level in the
well. It is thought that the filtering properties
of the sediments are responsible. In the case
of a small drawdown the loosest retained water
in the coarsest sediments is extracted, but as
the drawdown is increased, the more tightly
retained water frorn finer sediments is released.
The Ca-content of the water is rather vari-
able, 389—600 ppm, i.e. similar or higher than
in sea-water. The Ca/Cl-ratio in sea-water is
0.02 but in the water frorn the well it is 0.05—
0.145. This increase in calcium is very common
in ground-water which takes its chloride con-
tent from sea-water (see Fig. 8). There seems
to be no connection between the chloride
content and the ð-value. On the other hand
there is a definite correlation between the Ca-
content ancl the ð-value and especially between
the Ca/Cl-ratio and the ð-value. The only
exception to this rule is the second sample
from the last pumping test. This can be ex-
plained as follows: Before the pumping test
the water with the lower ð-value reached at
least up to 640 m depth (see Fig. 5), indicating
that the lower ð-value aquifer has a higher
pressure than the higher ð-value aquifer. The
water from the higher pressure aquifer seems
to have penetrated into the lower pressure
aquifer between pumping tests and reacted
with the well rock and so obtained same Ca/Cl-
ratio as the higher ð-value water.
The chloride content in the water from the
well was about half as great as tliat of sea-water.
This chloride content can be explained by
mixing of sea-water and ground-water clerived
from the precipitation on the mainland. Such
mixing coulcl take place in two ways, either
by mixing with ground-water from the main-
land after deposition of the sediments or by
mixing of interstitial fluvial water during the
JÖKULL 17. ÁR 305