Jökull - 01.12.1984, Qupperneq 46

Geothermal water from high temperature
areas in Iceland normally contains less than 0.1
mg/kg of iron. Exceptions are highly saline
waters as on the Reykjanes peninsula (2-3 mg/kg)
and waters strongly influenced (acidified) by vol-
canic activity like in the Krafla geothermal area.
During volcanic activity in Krafla very high con-
centrations of dissolved iron (up to 60 mg/kg)
were observed in well discharges. Precipitation of
iron silicates and sulphides or oxides was
observed, either within the wells or in surface
pipes (Kristmannsdóttir 1984). The normal con-
centration is only a fraction of those values, i.e.
0.1 mg/kg. In water from Reykjanes the con-
centration is up to 0.5 mg/kg.
We suggest the following explanation of the
iron concentration in the water from the jökul-
hlaup in 1983. Very acid water has leached iron
either from formation rocks or basaltic glass on
the lake floor. The high acidity must have been
caused by input of acid magmatic gases to the
geothermal water. Ferrous iron may remain in
solution as long as the water is reducing and the
pH is low. High acidity is more likely to remain
within the formation rock than on the lake floor
where mixing occurs with meltwater. Precipita-
tion of iron is expected due to reactions during
storing and oxidation in the lake and the magma-
tic effects will be masked out. Oxygen in the
freshwater will cause oxidation of H2S to sul-
phate and ferrous iron to ferric iron. We do not
know the time scale for this process.
PREMATURE TRIGGERING OF
JÖKULHLAUPS
In general, jökulhlaups from Grímsvötn occur
when the water has risen up to a critical level and
a pressure barrier has been eliminated. Then
water from Grímsvötn escapes beneath the ice
east of the lake, 1-2 km to the north of the roots
of Grímsfjall. This was discussed by Björnsson
(1974) and seems to have applied to the trigger-
ing of jökulhlaups in 1976 and 1982.
The triggering of the jökulhlaup in 1983.
The jökulhlaup in December 1983 occurred at
a water level 20-30 m below the critical level for
triggering jökulhlaups (Fig.3). Therefore, water
was not able to penetrate out from the bottom of
the lake. We suggest that this jökulhlaup was
triggered by the opening of waterways along the
slopes of Grímsfjall where the collapsed ice caul-
dron was observed northeast of Grídarhorn (see
Figs. 2 and 4). There increased geothermal or
volcanic activity had melted ice in places and for
two months meltwater with a sulphurous odour
had drained from this area down to Skeidarár-
sandur. Finally, this melting managed to open the
channels into the lake and the jökulhlaup started.
The leakage out of the lake took place at the
slopes of the mountain where the water pressure
was far from being high enough to open subgla-
cial waterways. This may also explain why the
surface of the lake did not fall down to the same
level as is usual for jökulhlaups that are triggered
at the critical water level.
The small jökulhlaups in 1939, 1941, 1945 and
1948 may all have been triggered through open-
ing of waterways by melting of ice when geoth-
ermal activity increased at the northeastern
slopes of Grímsfjall. These jökulhlaups are
exceptions to the rule that the odour of hydrogen
sulphide on Skeidarársandur is detected only a
few days before the water discharge increases.
(Thorarinsson 1974, Ragnar Stefánsson, pers.
comm.). In 1941 the beginning of a jökulhlaup
was noticed in early April but the main flood was
in mid-May. In 1945 the jökulhlaup was at max-
imum in late September but geothermal odour
was first noticed in June. The jökulhlaup in 1948
peaked in late February but the sulphurous odour
had been perceptible since early January.
In addition to this, we may add that the explor-
ers of Grímsvötn in the fifties seem to have
believed that the outlet from Grímsvötn was
along the eastern slopes of Grímsfjall. There they
described a trench in the glacier surface that has
not been observed in recent years (information
from Sigurdur Thorarinsson’s diaries, kindly pro-
vided by his son Sven Th. Sigurdsson 1983).
Increased thermal activity, frequency
and volume of jökulhlaups.
The effect of increased geothermal or volcanic
activity on the frequency and volume of jökul-
hlaups from Grímsvötn is an important problem.
Subglacial melting at the threshold east of the
lake may open waterways into the lake and trig-
ger jökulhlaups. If this takes place on the slopes
of Grímsfjall, the occurrence of a jökulhlaup is
hastened. Such a triggering mechanism might
increase the frequency of jökulhlaups and reduce
their volume. Volcanic activity at the threshold
east of the lake may trigger jökulhlaups whose
44 JÖKULL 34. ÁR