Jökull - 01.12.1976, Qupperneq 51

Subglacial water cupola may not form again
and meltwater would drain continuously from
tlie subglacial geothermal area. A lake was ob-
served in the Skaftá-cauldron during the sum-
mer 1976. This may explain why in December
1976 no jökulhlaup had occurred in Skaftá
since 1974 but sulphurous smell has frequently
been sensed from the river Skaftá.
CONCLUDING REMARKS
A summary has been given of glacier-margin
and glacier-surface lakes in Iceland. Jökulhlaups
are frequent from the lakes but the present size
of the glaciers does not allow the great jökul-
hlaups known from the first half of the century.
Glacier advance, in the first instance, will
form small lakes in gullies and ravines. Fre-
quent but small jökulhlaups will result. Recent
glacier advance has caused jökulhlaups from the
southern margin of Mýrdalsjökull to the rivers
Jökulsá at Sólheimasandur, Klifandi and Múla-
kvísl. This also may explain a jökulhlaup in
the summer 1976 from Vidbordsjökull, SE-Vatna-
jökull. A lasting glacier advance may form lakes
in tributary valleys. The last jökulhlaup from
Jökulsárgil at Sólheimajökull occurred in the
late 1930:s. The glacier would have to advance
over one kilometer to dam the gorge again. The
same applies for the proglacial lake Hagavatn.
— A glacier surge could dam up lakes quite
abruptly. This is probably the most drastic
eifect which a glacier surge can have on rural
districts in Iceland.
Thickening of the ice barrier of lateral lakes
will decrease the frequency and increase the
volume of jökulhlaups. At present, the jökul-
hlaups from Grænalón are triggered by under-
cutting of a shallow barrier. A thicker ice
barrier would force water to run beneath the
main glacier. Voluminous jökulhlaups would
occur again as during the 1940:s. The jökul-
hlaups from Grænalón and Vatnsdalslón would
cause again serious devastation of lands and
roads.
Lowering of the ice barrier may cause lateral
lakes to start dumping subglacially. Jökulhlaups
from the lake at Olafsfell, Hofsjökull would
flood the bogs at Thjórsárver. Further, glacier
retreat may uncover tributary valleys in which
lakes are formed. The lack of data on sub-
glacial landscape prevents further discussion of
tliis point.
All jökulhlaups from marginal lakes can be
prevented from occurring by providing con-
tinuous drainage out of the lakes. The drainage
outlet should be located at an elevation that
does not allow the ice barrier to be floated. Tlie
col can be lowered by explosion or by drilling
a tunnel through the rock barrier.
Glacier-surface lakes are so small that they
do not cause voluminous jökulhlaups. Their
existence and englacial drainage is a challeng-
ing problem in glacier hydrology.
ACKNOWLEDGEMENT
The author is indebted to Professor Sigurdur
Thorarinsson for encouragement and valuable
discussions.
REFEREN CES
Arnborg, L. 1955: Ice-margin lakes at Hoffells-
jökull. Geogr. Ann. 37, 202—228.
Aslielsson, J. 1936: On the last eruptions in
Vatnajökull. Societas Scientiarum Island-
ica, XVIII, Reykjavík, pp. 68.
Björnsson F. 1962: Fjallsárhlaupið 1962 og at-
huganir á lóninu i Breiðamerkurfjalli. The
glacier burst in Fjallsá 1962. Jökull, 12:
42-43.
Björnsson, H. 1974: Explanation of jökulhlaup
from Grímsvötn, Vatnajökull, Iceland. Jök-
ull, 24: 1-26.
Björnsson, H. 1975: Subglacial water reservoirs,
jökulhlaups ancl volcanic eruptions. Jökull,
25: 1-14.
Einarsson, E. H. 1973: Hlaup í Hafursá. Jökull,
23: 66.
Freysteinsson, S. 1972: Jökulhlaup í Köldukvísl.
Jökull, 22: 83-88.
Gestsson, G. 1956: Arnessýsla milli Hvítár og
Þjórsár. Arbók Ferðafélags Islands. 117 pp.
Reykjavík.
Green, R. 1952: Sedimentary sequence in the
Hagavatn basin. Jökull, 2: 10—16.
Kjartansson, G. 1938: Um nokkur jökullón og
jökulhlaup í Harðangri og á íslandi. Nátt-
úrufr. 8: 21-33.
Kjartansson, G. 1957: Selta í Jökulsárlóni á
Breiðamerkursandi. Jökull, 7: 39—41.
JÖKULL 26. ÁR 49