Jökull - 01.12.1990, Blaðsíða 167

flight on November 30, it was observed that the con-
centric crevasses created in the cauldron during the
jökulhlaup were cut across by a rift through the centre
of the cauldron. This has not been observed in earlier
jökulhlaups. It is likely that the pressure release asso-
ciated with the jökulhlaup triggered a short eruption
that did not reach the surface of the glacier. The only
other eruption considered to have occurred on the Loki
Ridge was in 1910 (Jónsson, 1986).
The volcanic activity in 1753 which released melt-
water to Djúpá, Hverfisfljót and Skaftá, may have been
located on west- or south facing slopes of the fissure
swarm which links Pálsfjall, Þórðarhyma, Hábunga
and Grímsvötn.
No historical records exist of volcanic activity in
Bárðarbunga, and the mountain was apparently not
known as a volcano until quite recently. Larsen’s
(1982) tephrochronological and geochemical studies
indicate that the volcanic eruptions that caused jökul-
hlaups into Jökulsá á Fjöllum in 1717 and in the late
15th century occurred along the ice-covered part of the
Bárðarbunga-Veiðivötn fissure system. She also finds
evidence for several eruptions in this system in the
early centuries of historical times. Many tephra layers
with the chemical characteristics of Bárðarbunga were
also found in the Bárðarbunga ice core (Steinþórsson,
1978), particularly in the period 1697-1720, which
seems to have been a very active time in this vol-
canic system (ísaksson, 1984). The eruption in NW
Vatnajökull in 1797 may have been on this system
and the eruption in Tröllagígar 1862-64 (Þórarinsson
and Sigvaldason, 1972) took place on it, just outside
the glacier edge. Larsen (1984, p. 54) concluded
that episodes of rifting and magmatic activity have
affected the Bárðarbunga volcanic system at an aver-
age interval of 100 years during the last 1200 years,
accompanied by volcanic activity at the Bárðarbunga
central volcano. There is no evidence for eruptive ac-
tivity in Bárðarbunga in this century in spite of high
seismicity for the last 16 years.
Several eruptions have produced jökulhlaups in
rivers running north from Vatnajökull, Jökulsá á
Fjöllum (1477, 1684-85, 1711-12, 1716-17, 1726,
1902-03 and 1934) and Skjálfandafljót (1902-03, and
1934). The jökulhlaups of 1477 and 1717 have been
attributed to events in the Bárðarbunga volcanic sys-
tem (Larsen, 1984), and similarly, the jökulhlaups in
the early part of the 18th century (Isaksson, 1984).
Other jökulhlaups appear to be related to activity in the
Grímsvötn system (1684-85, 1726, 1934). The two
jökulhlaups in Skjálfandafljótcoincide with activity in
the Grímsvötn system. A causal relationship between
activity in Grímsvötn and jökulhlaups in Skjálfanda-
fljót is hard to reconcile with the topographic data,
however, since the tributary area of Skjálfandafljót
does not overlap with the Grímsvötn system. Either
the jökulhlaups were not related to the eruptions or
the volcanic activity was more widespread than as-
sumed. In both cases there was a flood in Jökulsá á
Fjöllum as well, so if it was caused by one eruption
we would expect the site to have been on the NE flank
of Bárðarbunga.
Jökulhlaups in Jökulsá á Fjöllum may be caused by
volcanic eruptions in any of the three volcanic systems
beneath thenorthern part ofVatnajökull: at Kverkfjöll,
Bárðarbunga or from the Askja fissure swarm beneath
Dyngjujökull. A potential for very large jökulhlaups
exists in these areas, for example in the large ice-
filled calderas of Bárðarbunga and Kverkfjöll. These
may have been the sources of the prehistoric catas-
trophic floods in Jökulsá á Fjöllum which occurred
about 7,100 B.P., 4,600 B.P., 3,000 B.P. and before
2,000 B.P. (Tómasson, 1973; Elíasson, 1977).
The only known jökulhlaup in Jökulsá á Breiða-
merkursandi occurred in 1927. It was accompanied by
a strong sulphuric stench and some ash fall, indicating
a volcanic origin (Sigurður Bjömsson, 1977). It is
considered possible that it may have been caused by
an eruption in Esjufjöll.
CONCLUSIONS
Based on the recently acquired knowledge on sub-
glacial topography and earthquake activity, combined
with results of earlier investigators, the following vol-
canic systems can be defined beneath Vatnajökull:
1. Bárðarbunga volcanic system. It includes the
fissure swarms of Veiðivötn (including Vatna-
öldur and Heljargjá) to the SW of the Bárðar-
bunga central volcano and Dyngjuháls to the NE.
JÖKULL, No. 40,1990 163