Jökull - 01.12.1978, Qupperneq 12

sediments are lying at a lower altitude than
the proximal end of the delta. Thus, ice must
have lain close to the delta to allow water to
flow over the delta flat.
No ice marginal deposits could be
correlated with the proximal end of Vind-
heimamelar. A localized stillstand there is
caused by a bedrock threshold of 30—40 m
altitude, into which the present river (Svartá)
has eroded a 15 m deep and 800 m long gorge.
At the head of the gorge there is a waterfall
named Reykjafoss (Fig. 9). To the north of
Vindheimamelar an outwash plain is situated
at about 15 m altitude. During the ice
recession period, before this outwash plain was
built up, the land was lying still lower there.
When the receding ice front approached the
Reykjafoss threshold, calving was highly
reduced as the ice floor rose from more than 30
m depth north of the threshold to about 10 m
depth.
DEGLACIATION OF THE VALLEYS
After the formation of the Vindheimamelar
delta, the ice retreat was nearly continuous.
No end moraines have been observed in the
valleys to the south of Vindheimamelar. On
the other hand other sendiments and land
forms indicate how the ice retreat proceeded.
Widespread deposits of ablation till and ice
contact fluvioglacial sediments are located in
the shallow valley, which opens at Vind-
heimamelar (Fig. 9). These sediments are
mainly of fluvioglacial origin in the area from
Vindheimamelar towards Mælifellsá in the
south. From Mælifellsá and onwards to the
south, ablation till dominates.
In the deep valleys at the head of Skaga-
fjördur (Austurdalur, Vesturdalur and Svart-
árdalur), there are neither ablation moraines
nor end moraines. The valley bottoms of
Austurdalur and Vesturdalur are covered by
coarse fluvioglacial valley trains with boulders
up to half a metre in diameter. The coarseness
indicates that these sediments were formed
near the retreating ice front. Finer grained
gravelly foreset beds were found at one
locality in Vesturdalur. This indicates that
lakes sometimes formed in front of the
retreating ice front and were later filled by
fluvioglacial sediments. In both of these val-
leys the valley trains are deposited behind
thresholds. Later, canyons were cut through
these thresholds and thereby lowering the
river floors. The Lateglacial valley train in
Austurdalur has been eroded away for the
most part but in Vesturdalur the valley floor is
broader and extensive terraces have been
preserved.
Fig. 10 shows a longitudinal section and
some cross sections of the Vesturdalur valley.
Two thresholds can be seen near the entrance
of the valley. The valley train has been built
up behind the inner and lower threshold
(170—175 m altitude). A gorge must, there-
fore, already have existed through the outer
threshold (reaching up to 190 m altitude) be-
fore the valley train was formed. South of the
175 m high threshold the valley bottom of
Vesturdalur is covered by fluvioglacial sedi-
ments. Extensive terraces are left in the
northernmost part of the valley, but at the
Litlahlíd farm and southwards, only minor
terraces remain close to the hillsides. These
southernmost terraces lie 20— 40 m above the
present river, and represent the height of the
valley train which once filled the bottom.
Fig. 12 shows a longitudinal section of the
valley bottom of Austurdalur. A 90 m deep
canyon lies at the mouth of the valley. This
canyon has been cut into a glacially eroded
valley bottom with some thresholds. Two
thresholds can be seen in the long section. A
fluvioglacial valley train has been built up
behind the inner one. The through between
the two thresholds is occupied by thick till.
The valley train can be reconstructed by small
but prominent terraces which rise con-
tinuously from 230 m altitude at the threshold
to 290 m at the deserted Árbær farm. No rem-
nants of the terraces are found upstream from
Ábær. Another system of terraces rises
upstreams from 260 m altitude at Ábær. This
indicates that the ice front was stationary for
some time at Ábær. This stillstand seems to
10 JÖKULL 28. ÁR