Jökull - 01.12.1977, Qupperneq 5

Fig. 1. ERTS-1 satelite image showing the
Bárdarbunga area (detail from Fig. 1 in Williams
& Thorarinsson, 1973). The features are ex-
plained in Fig. 2.
Mynd 1. Bárðarbunga og umhverfi séð frá
gervihnetti i 920 km hœð. Örnefni og drœttir
myndarinnar skýrðir i tnynd 2 (úr grein Wil-
liams og Sigurðar Þórarinssonar, 1973).
deals with the methods and results of the tephro-
chronological effort, and some inferences that
can be drawn from the profile and the tephras
themselves.
METHODS USED
The tliree methods used to identify the
tephras were (a) comparison to volcanic history,
as gleaned front the study of written documents,
and from tephrochronological studies, (b) grain-
size analysis, and (c) chemical and optical ana-
lysis.
Volcanic history
A nurnber of volcanoes could potentially con-
tribute tephra to Bárdarbunga. A list of known
or suspected eruptions, drawn from Jónsson
(1945) and Tliorarinsson (1974), is shown gra-
phically in Fig. 4. together with the eruptions
represented in the ice core, according to the
interpretation of the profile presented here. In
the following what is known of the character-
istics and histories of the chief volcanoes in
question is briefly discussed (see Figs. 1 and 2).
GRÍMSVÖTN is a subglacial caldera in
Vantajökull, some 23 km SE of the drilling site
(Table II, nos. 3, 4, 5). It has erupted at least
twice in this century, in 1934 (Áskelsson 1936)
and 1922 (Thorkelsson 1923). Thorarinsson
(1974) has recently published a comprehensive
account of its activity throughout the ages, since
1332.
The eruptions in Grímsvötn are accompanied
by jökulhlaups, caused by the catastrophic
draining of the caldera lake through subglacial
tunnels into the river Skeidará. Jökulhlaups
need not be accompanied by an eruption, how-
ever, and since 1934 they have occurred peri-
odically every 5 or 6 years without an eruption
taking place at the same time. This also has
been the case in the past, even if the length of
the period separating the hlaups has varied.
The mechanism of jökulhlaups from Gríms-
vötn has been under dispute for a long time.
The presently accepted explanation is that of
Helgi Björnsson (1974). It may be summarized
thus (op. cit. p. 1):
„In a period of five or six years the water
level in the ice-covered caldera lake rises about
100 m. The glacier is lifted off a subglacial
ridge east of the lake and water is forced sub-
glacially through a 50 km long route beneath
Skeidarárjökull, causing vast floods. When the
water level ltas fallen about 100 m the water-
way is sealed by rapid plastic deformation of
the ice at the eastern edge of the Grímsvötn
depression. The level is not lowered to the
subglacial rim.
The primary cause of jökulhlaups front Gríms-
vötn is the melting of ice in the geothermal
area. The slope of the glacier surface is altered
and ice and water flow into the Grímsvötn de-
pression."
It has been suggested (Thorarinsson 1953) to
account for the observed apparent causal rela-
tionship between Grímsvötn eruptions and
jökulhlaups in Skeidará, that the release of
pressure in the caldera upon draining away tlie
water, sometimes triggers an eruption. The low-
ering of the level of the caldera lake by 140 m
JÖKULL 27. ÁR 3