Jökull - 01.12.1990, Qupperneq 158

son and Sæmundsson (1987). Six central volcanoes
have been identifiedbeneath the ice cap: Bárðarbunga,
Grímsvötn, Þórðarhyma, Kverkfjöll, and Öræfajök-
ull, and tentatively Esjufjöll. The location of fissure
swarms extending from the various central volcanoes
is more uncertain. Six fissure swarms have been
traced from the ice-free areas towards Vatnajökull.
First, the swarm which connects Torfajökull, Veiði-
vötn, Vatnaöldur and Heljargjá disappears beneath
Sylgjujökull and Köldukvíslarjökull, where there is,
however, no topographic expression of it. It proba-
bly continues under the glacier, runs through Bárðar-
bunga, reappears near Kistufell and extends to Trölla-
dyngja (Dyngjuháls swarm). The Laki fissure swarm
is traced to the edge of Skaftárjökull. It may con-
tinue beneath the ice towards Hágöngur, Þórðarhyma,
Háabunga and Grímsvötn. Between the two swarms
a swarm of hyaloclastite ridges extends from Tungna-
árjökull along Tungnaárfjöll and Fögrufjöll towards
Torfajökull (Fögrufjöll swarm). Two fissure swarms
occur north of Vatnajökull to the east of the Dyngju-
háls swarm. One trends from Askja to Dyngjujökull,
the other is centred on Kverkfjöll (Jóhannesson and
Sæmundsson, 1989). All these fissure swarms have
been active in postglacial time except the swarm at
Fögrufjöll.
The central volcanoes and the fissure swarms can
be grouped together into volcanic systems, which will
now be discussed in more detail.
BÁRÐARBUNGA-VEIÐIVÖTN VOLCANIC SYSTEM
Bárðarbunga is a large volcanic edifice, which
rises to an elevation of 1500-1850 m, i. e. 500 -
900 m above its surroundings. It contains a 700 m
deep caldera, about 80 km2 in area. Its bottom is
at 1100 m a.s.l. The mountain massif stretches NE
in the direction of Dyngjuháls. From the southwest-
ern slopes of Bárðarbunga, a ridge strikes SW across
Köldukvíslarjökull towards northwestern Hamarinn.
Beneath Sylgjujökull a cone-shaped mountain that
rises above 1000 m is situated on the continuation
of this ridge. The ridge may be the expression of a
fissure swarm that connects Hamarinn and Bárðar-
bunga. Thus, the subglacial topography indicates
that the Bárðarbunga volcanic system, with a fissure
swarm that extends to the southwest through Trölla-
gígar and Heljargjá to Veiðivötn (Sæmundsson, 1978,
1979; Larsen, 1984), is connected to a separate central
volcano, Hamarinn.
Bárðarbungahas been among the most seismically
active volcanoes in Iceland since 1974, when a remark-
able series of earthquakes began there. The series has
so far included 10 events of magnitude 5 and larger.
Earthquakes of this magnitude did not occur in this
area for at least 50 years priorto 1974. The epicentres
are mostly located slightly NE of the caldera. Focal
mechanisms have been obtained from teleseismic data
for 5 of the large events at Bárðarbunga (Einarsson,
1986, 1987, 1988). They all show a large component
of reverse faulting, interpreted as the result of deflation
of a crustal magma chamber beneath Bárðarbunga.
Thus the seismic evidence shows that Bárðarbunga
contains an active magmatic system and suggests that
the pressure in this system has been decreasing since
1974.
A separate epicentral cluster is found NE of
Bárðarbunga, near Kistufell, where small swarms
sometimes occur. Seismic activity is, on the other
hand, absent in the fissure swarm SW of Bárðarbunga.
Volcanic activity in Bárðarbunga may cause floods
in rivers flowing in all directions, depending upon the
eruption site. Fig. 4 shows the water divides on the
glacierforthe various rivers (Bjömsson,inprep.). Wa-
ter from an eruption on the SW and W flanks would
drain to Kaldakvísl, and an eruption on the N flank
would produce floods in Skjálfandafljót. The water
divide between Skjálfandafljót and Jökulsá á Fjöllum
follows a ridge that extends from northeastern Bárðar-
bungatowards Kistufell. An eruption on the and E and
SE flanks would produce meltwater that would drain
to Jökulsá á Fjöllum. From a south flank eruption
some water might find its way to the water-filled vault
beneath the Skaftá ice cauldrons and even Grímsvötn.
and could trigger jökulhlaups in Skaftá and Skeiðará.
Volcanic activity in Bárðarbunga could certainly
cause catastrophic jökulhlaups, with high potential en-
ergy and great erosional power. Suchjökulhlaupsmay
explain the deep canyons in Jökulsá á Fjöllum. Melt-
water produced within the caldera would most easily
escape through a pass in the SE rim and drain to-
154 JÖKULL, No. 40, 1990