Jökull - 01.06.2000, Qupperneq 43

Helgi Björnsson et al.
Our model predictions of the delineation of the
water drainage basins suggest that from an area of 60
km

of the caldera floor, meltwater is drained down to
Mýrdalssandur, as did 18 of 20 recorded jökulhlaups
in historical times; from about 20 km

down to Sól-
heimasandur as did 2 of the jökulhlaups. The third
route down Entujökull was taken by the jökulhaup in
1600 B.P. (Haraldsson, 1981).
The location of the watershed predicted by our
data may change during volcanic eruptions. However,
if our calculations apply for normal conditions and
at the start of an eruption we consider it likely that
meltwater from the eruption site will continue to
drain through pre-existing channels. After the erupti-
on breaks through the ice cover the basal water
pressure will be determined by the water level at the
crater and the energy used to transport the water down
the glacier.
CONCLUSIONS
We present the first maps of Mýrdalsjökull that descri-
be with known accuracy the surface and bedrock
topography. The maps provide basic data for various
studies in geology, glaciology and hydrology. They
describe the shape of the subglacial part of the Katla
volcanic system, the geometry of the central volcano,
the location of recent eruptive sites and their conn-
ection with structures in the surrounding landscape.
The caldera encircles an area of 100 km

, is 600 to
750 m deep and its highest rims reach 1380 m a.s.l.
The northern part of the caldera floor is smoother and
lies deeper than the southern part, which in contrast
is characterized by subglacial ridges and individual
mounts rising from ca. 750 to 1100 m a.s.l. A row of
mounts trending NNW, presumably active in recent
eruptions, lies 2 km within the eastern rim of the
caldera beneath 400 m thick ice.
A number of ridges radiate out from the caldera,
however, none towards south. One ridge strikes W
towards the neighbouring volcano Eyjafjallajökull,
and a second ridge strikes E from the eastern rim
of the caldera. Ridges also radiate from the caldera
rim towards NW, N, and NE. A linear depression,
bounded by steep slopes, 200-250 m deep and 1.5 km
wide, strikes NE towards the volcanic fissure Eldgjá.
The maps describe the geometry of the ice cap and
its flow and provide data to evaluate the drainage of
meltwater from the ice cap during normal conditions
and volcanic eruptions. All but two of the 20 jökul-
hlaups in historical times have taken a path through
a pass in the caldera rims southeastward, down to
Mýrdalssandur. An area of 60 km

within the caldera
drains now to Mýrdalssandur, and an area of about
20 km

to the southwest, down to Sólheimasandur.
Two jökulhlaups are known to have taken this rou-
te to the river Jökulsá á Sólheimasandi in historical
times. A third route, westward into Fremri Emstruá
and the Markarfljót river, was taken by a prehistoric
jökulhlaup in 1600 B.P.
The maps form the basis for studies of glacier-
volcano interactions and provide a reference datum
for monitoring temporal changes in the geometry
and flow of the glacier in response to basal geot-
hermal activity as well as to climatological impact.
Presently, geothermal activity is displayed by sever-
al small cauldrons in the glacier surface, 0.5-1 km in
diameter. Meltwater accumulates beneath two or three
of these cauldrons and is frequently drained in small
jökulhlaups.
Over the last four centuries eruptions have occur-
red on single vents and volcanic fissures trending E-
W and S-N. We suggest that the 1755 eruption, the
largest eruption in Mýrdalsjökull of the millennium,
took place on a several kilometers long fissure trend-
ing east from Goðabunga. The fissure extended far
into the drainage basin of Kötlujökull and therefore
meltwater was directed eastward to Mýrdalssandur.
We suggest that the 1823 eruption took place on the
2-3 km long ridge striking NNW from the eastern rim
of Háabunga (Figure 12), and that the 1918 eruption
site was situated in the same area.
Acknowledgements
Members of the rescue group Víkverjar at Vík ass-
isted during field work. Axel Einarsson surveyed the
optically levelled profile across the ice cap. Georg
Guðni Tómasson, Halldór Gíslason and Torfi Hjalta-
son assisted with field work. The work was supp-
orted by the Icelandic Research Council. The National
Power Company provided a hut on a sledge during the
field work.
42 JÖKULL No. 49