Jökull - 01.06.2000, Page 21

Guðrún Larsen
Entujökull gaps (e.g. Larsen, 1978; Dugmore, 1987;
Sigurðsson, 1988), as well as the Kötlujökull gap,
while post-Eldgjá jökulhlaups have been confined to
the Kötlujökull gap.
Lasting effects: The post-Eldgjá changes
When volcanic activity was resumed after the  200
year repose following the Eldgjá eruption, new con-
ditions had developed on the sandur plain in front of
Kötlujökull. The Álftaver lava was now a high area
that formed a barrier obstructing meltwater flow to
the east. Hyaloclastite flow(s) at Kriki had, possibly,
raised the topography on the northern side of Kötlu-
jökull by tens of metres and changed the topography
below the glacier significantly. The combined effect
was to direct jökulhlaups escaping through the Kötlu-
jökull pass in a southerly direction, into the lower ly-
ing areas to the west of the lava fields.
Some recovery of the areas damaged by the Eld-
gjá eruption took place during the long repose. The
areas along the edges of the Eldgjá lava fields are
likely to have developed in the same way as Bruna-
sandur in front of the 1783 Skaftáreldar lava, which
became vegetated within decades and farmable within
two centuries (Thoroddsen, 1911). In the following
centuries, revegetated areas and farmlands within and
along the borders of the present sandur plain were
gradually destroyed (e.g. S.t.s. Ísl. IV, 1907-15).
Tephra fall affected the neighbouring areas, some-
times severely, but the effects were temporary. Per-
manent damage on the scale inflicted by the Eldgjá
tephra has not reoccurred. Compared to the jökul-
hlaups, airfall tephra plays a minor role in the post-
Eldgjá changes.
Changes caused by the post-Eldgjá jökulhlaups
can be assessed from many sources. The first doc-
umented occurrence of a jökulhlaup after more than
200 years repose was Höfðárhlaup that flooded the
western part of Mýrdalssandur shortly before 1179
AD, destroying several farms and two churches within
the Höfðabrekka parish (Biskupa Sögur 1878). The
comment in the Book of Settlement that a sandur
plain, Höfðársandur, now lies where there was pre-
viously a fjord, suggests that significant changes had
taken place when the comment was written in the late
12th or early 13th century (Íslensk Fornrit I). Farms
within the Álftaver lava were flooded and abandoned
in the late 15th century (Gestsson, 1987; Árnadótt-
ir, 1987), indicating that by then jökulhlaups could
flood the lava field. This implies that the surface of
the sandur plain west of the lava field had been raised
to a point where jökulhlaups were no longer deflected
by the lava edge and that the sandur was progress-
ing onto the lava field. Continuation of this develope-
ment is seen in all the documented jökulhlaups since
1625 (S.t.s.Ísl. IV, 1907-15). The shoreline at var-
ious times since the Norse settlement can be recon-
structed with reference to the Book of Settlement and
younger sources (e.g. S.t.s. Ísl. IV, topographic maps
from 1904 onvards). An extension of up to 4 km
from the pre-Eldgjá shore is implied by such recon-
structions. The permanent addition to Mýrdalssandur
along the 30 km of shore between Vík and Álftaver
lava is at least 60 km

.
Evidence of the changes is also preserved in the
soil in areas within and adjacent to Mýrdalssand-
ur. The soil that began to form on the Álftaver lava
shortly after the eruption, is devoid of windblown
sand size material until the early 15th century. This
change becomes noticeable above the tephra layer
from the 1416 Katla eruption and increases drastically
above the 1625 Katla layer. The increasing influx of
windblown material is best explained by sand blowing
in from areas close to the Álftaver lava, i.e. a growing
sandur plain.
Frequent renaming of the rivers on Mýrdalssand-
ur implies that both their courses and character have
repeatedly changed through time. Some have disap-
peared while others have appeared temporarily fol-
lowing major jökulhlaups. The most recent example
is the disappearance of the Sandvatn river, a glacial
river in the middle of Mýrdalssandur, following the
1918 jökulhlaup.
The picture emerging from the pieces of infor-
mation is the following: When volcanic activity re-
sumed after a long repose, jökulhlaups accompanying
Katla eruptions repeatedly flooded the area between
the Álftaver lava field and the hyaloclastite mountains
of Höfðabrekkuheiði. Different parts were affected in
different eruptions but the overall effect of repeated
deposition of water-transported debris was to fill in
20 JÖKULL No. 49