Jökull - 01.01.2021, Page 15

The 1918 Katla eruption
around 3 PM the youngsters in the group were sent
home for safety’s sake. Shortly afterwards it became
clear that a jökulhlaup was approaching along the
Skálm river. The remaining farmers hastened across
the river as a greyish-black wave of flood water, 2–
3 m high, was closing in “about 10 minutes away”
from the ford (Jóhannsson, 1919). Although the herd-
ing farm hands made a narrow escape, no human lives
were lost and all reached safety (Jóhannsson, 1919;
Gísladóttir et al., this issue).
The branch that came down Hólmsá swiftly re-
moved the bridge in the canyon, around 3 PM
(Jóhannsson, 1919). It advanced down Kúðafljót and
reached the farm Sandar in Meðalland not later than
5 PM, where the farmer’s family made a narrow es-
cape. The water in Kúðafljót continued to rise and by
8 PM the flood had reached the homefields of Strönd
in Meðalland, not receding until 10 PM and leaving
2–3 m thick heaps of ice on its banks (Jóhannsson,
1919).
The northern fork flooded large areas in the dis-
tricts of Álftaver and Meðalland (Figure 9). In Álfta-
ver the flood water entered local rivers and brooks and
left up to 4 m thick heaps of ice and sludge on their
banks (Jóhannsson, 1919).
An excursion to the outlet of the 1918 jökulhlaup
at the southern tip of Kötlujökull (Jóhannsson, 1919)
found that the first phase of the jökulhlaup had trav-
elled on the surface of Kötlujökull and part of it
had overtopped the adjacent hills into the gully Re-
mundargil (Figure 9). Flooding over the surface of
the northern part of Kötlujökull was also confirmed
(Sveinsson, 1919).
Second phase of the jökulhlaup on October 12
A much larger flood was seen from Hjörleifshöfði on
Mýrdalssandur around 5 PM as it pushed through the
1.5 km wide gap between Selfjall and Hafursey (Fig-
ure 9). It carried huge amounts of ice, flooded the
area around Hjörleifshöfði and continued into the sea
(Jóhannsson, 1919). The farmer in Hjörleifshöfði de-
scribed it as if “snow-covered hills” were rushing over
the sandur plain. The width of the flood on the east
side of Hjörleifshöfði was about 5 km (extending to
the east of Lambajökull hill) and about 3 km on the
west side. The flood lasted into the evening, but re-
ceded during the following night.
From Vík it was noticed that the flood in
Múlakvísl had receded around 5:30 PM (by then
Múlakvísl river had been blocked at Selfjall by ice
carried by the jökulhlaup), but at the same time a
much larger flood had reached Hjörleifshöfði, car-
rying huge icebergs (Jóhannsson, 1919). The flood
formed a large spit of sediment and ice blocks into
previously navigable waters. Offshore, the icebergs
formed a large pack of ice slowly moving towards the
west.
The “main flood” broke up the edge of the glacier
and created an ice canyon. Members of an excursion
dispatched by the sheriff at Vík estimated this canyon
to be 1300–1700 m long and 300–500 m wide with ice
walls up to 130 m high (Sveinsson, 1919). This sec-
ond phase of the jökulhlaup was confined to the west-
ern part of Mýrdalssandur and did not reach Álftaver
or Meðalland.
The flow front of the jökulhlaup apparently ad-
vanced the 9 or 10 km from Hafursey to Hjörleifs-
höfði in about 30 minutes, a velocity of 20 km/hour
(∼6 m/s).
Flooding on Mýrdalssandur 14–28 October
On the morning of October 13, the jökulhlaup was
over and had left large icebergs on the sandur plain
(Figure 10), from “20 m to 60 m high” (Loftsson,
1930). The coast had advanced and a second spit
had formed from the flood deposits, with numerous
stranded icebergs.
The flood resumed on October 14, forming chan-
nels with fast flowing waters that buried or carried
off icebergs left on the sandur in the first two days
(Jóhannsson, 1919). During the following day the wa-
ter flow had decreased and bars had appeared between
the channels, some of which appeared quite deep. The
next four days saw similar conditions where the wa-
ter flow decreased and increased again. On October
20 the flood was smaller than before and continued
to decrease until October 26, when the sandur plain
was more or less dry (Jóhannsson, 1919). The last
flood occurred on October 28. About three weeks af-
ter the eruption was over, on November 26, a substan-
JÖKULL No. 71, 2021 13