Jökull - 01.12.1967, Qupperneq 35

and in the Jarlhettur Valley as can be clearly
seen on photograplis from the years 1929 and
1934. Wright (1935) measured the discharge
through the Leynifoss Col under high water
conditions in the year 1934 and it was then
only about 6 m3/sec.
“Jökulhlaups” from Hagavatn
As mentioned five “jökulhlaups” have been
reported from the Hagavatn area. The last two
were investigated, and it is therefore known
what changes in drainage were caused by these
events, except that the drainage conditions be-
fore the “jökulhlaup” of 1929 are unknown.
The “jökulhlaup” of 1929 escaped directly
through the Leynifoss Col and the lake level
was lowered 6.3 m (Ólafsson 1929, Wright 1935),
but in tlie year 1939 Hagafellsjökull Eystri had
receded so far that the water could be released
through the present outlet, the Nýifoss Col, and
the lake level was lowered 9.5 m, from level
3 to level 4 on Fig. 5 (Thorarinsson 1939). The
released volume of water in these two latest
“jökulhlaups” was probably greater than in the
earlier ones, so that they cannot be explained
by a recurrence of the same events. As men-
tioned earlier, the height of the shore line be-
fore 1929 corresponds approximately to the
height of the channel along the north slopes
of Brekknajökull. The “jökulhlaup” of 1929
probably occurred because the drainage moved
from this channel and the lake level was lower-
ed from 2 to 3 on Fig. 5. As mentioned above,
there is a good correlation between the highest
one of the shore lines and the delta terrace
at the NW corner of Brekknafjöll. For some
time the terrace has determined the elevation
of the lake level. The glacier burst in the year
1902 probably occurred because the glacier
snout had retreated a little from the delta
terrace and tlie water could then drain on the
north side of it. In this way the lake level sub-
sicled frorn 1 to 2 on Fig. 5 or about 4—5 m,
and about 45 million m3 of water were releas-
ed, wliich is in fairly good correspondence with
what is told about its magnitude in comparison
with the “jökulhlaup” of 1929 (Ólafsson 1929),
when 65 million cubic meters of water were
released from the Hagavatn during 24 hours
(Wright 1935). A comparable volume of water
was released in the “jökulhlaup” of 1939, but
this time distributed over three days. The latter
therefore caused much less damage than the
former (Thorarinsson 1939). No attempt will
be made here to explain the causes of the
“jökulhlaups” in the years 1708 and 1884, but
they most likely represent some recession of the
Flagafellsjökull Eystri, allowing the water to
escape through some lower subglacial tunnels.
Thorarinsson (1939) suggests that the “jökul-
hlaup” in 1702 escaped through the Nýifoss Col
and in 1884 through the Leynifoss Col, but
they might also have originated from ice-damm-
ed marginal lakes in the Jarlhettur Valley.
The changing drainage
The glacial retreat has brought about changes
in drainage conclitions at the southern margin
of Langjökull. Here the bedrock is mostly post-
glacial lavas and other very permeable rocks,
where no surface drainage is found. It is al-
most certain that the ground water drains to
the rivers Sog and Brúará, while the surface
run-off from the glacier flows through Haga-
vatn and from there into the Tungufljót river.
As formerly discussed the lake level of Haga-
vatn was up to 22 m higher than the present
one and three other marginal lakes were form-
ed along the glacier margin west of Hagavatn
(Fig. 7) during the time of the greater extent
of the glacier. The glacial retreat caused succes-
sive lowerings of the Hagavatn lake level and
the emptying of the three western lakes, which
took place sometimes between 1950 and 1960,
but in that year they had all disappeared. It
can be asserted that great leakage through the
bottom of these lakes was yielded to the ground
water storage, especially during their formation
before the glacial silt and clay had tightened
their bed (Tómasson 1964). During this time
a great part of the melt water from the south-
ern part of Langjökull has been drained as
ground water. At the present time the surface
drainage from Langjökull flows on compara-
tively impermeable moraine. The decrease in
the discharge of the river Sog can at least part-
ly be explained by this process (Sigbjarnarson
1966). It can be stated that during the last
decades Tungufljót has been capturing water
from the Sog and the Brúará.
JÖKULL 17. ÁR 269