Jökull - 01.12.1976, Blaðsíða 57

the temperature curves to greater depths, we
see that there seems to be no deviation from
the normal gradient below about 1000 m in the
hole. This means that the increased surface
temperature and pressure head is most prob-
ably caused by increased pressure in the aqui-
fers between 820 and 1000 m depth in the drill
hole.
Fig. 3. Temperature profiles in the drill hole
at various times. Also shown in the upwards
extrapolation of the thermal gradient below
900 meters, as measured in 1967.
Mynd 3. Nokkrir hitaferlar i holunni, mœldir
1973, og hitastigull undir Vestmannaeyjum fyrir
gos (1967).
DISCUSSION
Conduction of heat from the eruptive fissure
is easily ruled out as a cause of the observed
ground water heating, because of the long time
constant (here ~ 104 years) involved.
The most likely process responsible for the
initial temperature and water level increases in
the hole is pressure of volcanic gases in the high-
ly permeable aquifers between 800 and 1000 m
depth. Water will flow laterally towards the
ltole, seep througli the blockage at 821 m and
then flow upwards, losing heat to the hole
walls on the way (Fig. 3). The pressure changes,
however, do not accurately reflect activity in
the crater and we have found no convincing
explanation for the pressure minimum in March
1973.
Discharge from the drill hole probably start-
ed between 22 and 03 on January 24th accord-
ing to eyewitness accounts; deep premonitory
earthquake activity occurred overnight January
21—22 (S. Björnsson, pers. comm. 1974). How-
ever, extrapolation of the temperature curve at
the hole back in time indicates tliat the water
level may have been rising, even for several
days, before this. The rapid drop in pressure
around the end of activity in the crater (Fig. 2)
is likely to be due to the decline in pressure
from below or the closing of gas vents.
The pressure changes at the top of the hole,
both the transient changes during the eruption
and changes in the mean water level since 1964
may be partly caused by the opening and clos-
ing of aquifers having different hydrostatic
heads. The transient changes in water level are
also partly caused by buoyancy effects in the
water column, both by its thermal expansion
and the effect of gas bubbles: water samples
from the deep aquifers were very rich in dis-
solved gases (Pálmason et al., 1965) and during
the 1973 eruption, much gas of composition
Fig. 4. Plot of discharge or water level in the
Heimaey drill hole against temperature, using
10-day means from Figs. 1 and 2 (July 1—10
omitted for lack of data).
Mynd 4. Tengsl hitastigs og rennslis eða vatns-
stöðu i borholunni.
JÖKULL 26. ÁR 55