Jökull - 01.12.1974, Blaðsíða 36

Fig. 7. The course of
the jökulhlaup after it
subsided in early April
1972. The course is cover-
ed with black sand and a
silt layer. The area on
Fig. 6 is also shown on
this picture.
Mynd 7. Farvegur Grims-
vatnahlaupsins, eftir að
það hafði sjatnað snemma
í april 1972. Farvegurinn
er þakinn svörtum sandi
og mó. Svceðið á Mynd 6
er einnig á þessari.
normal grain size differs very much from one
locality to another. As a general rule it becomes
finer grained further downstream. This layer
is usually very coarse at top and is considered
to reach down to the next very coarse layer.
This stratification, which is very coarse in-
deed, is due to the mode of transportation of
the coarse material. Branches are cut down un-
til protected by a stone layer. Then gravel bars
accumulate on top of it and again some erosion
and formation of a protective stone layer takes
place again. This may be repeated many times
during the jökulhlaup. The fresh appearance
of this layer indicates that layer 2 is usually
new, redeposited in the jökulhlaup. The thick-
ness of this layer is usually about 1 m in the
upper part ancl thins out downstream.
The volume of this layer is estimated at 45
Gl. The distance it has been brought down-
stream is not known, but based on bedload
measurement and changes in the flood channel
it is estimated as 3 km.
Another effect is the widening of the channel
by bank erosion. Along Skeidará the right bank
was eroded in the upper reach, usually a few
tens of meters, forming a bank a couple of
meters high. At Gígja much higher banks were
eroded and much more material brought into
the channel that way. Both banks were eroded
there, but the right bank to a much greater
extent.
The net effect of these factors was measured
at Skeidará by observing the river level both
before and after the jökulhlaup in relation to
piles driven down by the Public Roads Ad-
ministration in a line along Skeidará from
Skaftafellsbrekkur and approximately 10 km
downstream. In the uppermost part there was
insignificant difference, about 15 cm net accu-
mulation midway on the line and an erosion
of a few tens of cm at the lower end of the
line. Along Skaftafellsbrekkur there seems to
be some net accumulation and from there to
the glacier a net erosion. In all these calcula-
tions the unstable layer 1 is left out. At Gígja
there was obviously considerable net accumula-
tion except in the uppermost part. This accu-
mulation was mainly due to bank erosion.
Sampling of suspended sediments at the main
outlets into the sea at the highest peak in sedi-
ment discharge indicates that most of the sedi-
ment load is cleposited on the Sandur plain.
This is mainly sedimented as layer 1 as already
discussed. About 10- 106 m3 have been aclded
to the sancl which makes a 4 cm thick layer
over the entire flood channel.
The remaining about 5 • 10° m3 of sediment
load is deposited close to the shore. Suspend-
ecl in the sea on the 27th of March were 1.1 • 10°
tons. 70% of the total clay is found there, 13%
of total fine silt and less than 1% of total
coarse silt. This indicates very rapid sedimenta-
34 JÖKULL 24. ÁR