Jökull - 01.12.1984, Page 62

Fig. 5. Acid fallout from the Eldgjá eruption in
934 A.D. as seen in 3 Greenland ice cores.
Sampling as in Fig. 3.
Mynd 5. Súr úrkoma frá gosinu í Eldgjá 934 í
þremur ískjörnum frá Grœnlandi.
explosive phase in the eruption, which must have
been more pronounced than in the case of the
Lakagígar eruption. The long duration of the
fallout is also, at least partly, supporting this
point of view.
Another interesting feature is the chemical
composition of the acid: During the Dye 3, 1980
deep drilling season the Eldgjá signal was
encountered and samples were collected on loca-
tion for chemical analysis by M. Herron over the
ice segment, where the acidity increased due to
the eruption, Fig 5. The results are interesting,
because they show a quite different acid composi-
tion, than the Lakagígar eruption, as well as a
change of chemical composition with time, (Her-
ron 1982). The deposited acids are first consisting
of mainly HF, then in the peak acidity H2S04
dominates and finally a mixture of HF, H2S04
and HCl.
These chemical data must be interpreted with
some care, because in the Créte acid record there
is an additional high level of HCl in the Eldgjá
fallout (Hammer, 1980b and M. Herron, personal
communication 1981). This partly explains the
very pronounced maximum in the Créte acidity
profile. The interpretation of the difference
between the Dye 3 and Créte acid composition is
probably to be found in either aeolian differentia-
tion and/or in the actual phases of the eruption.
In both cases meterorological conditions at the
time of the eruptive activity, could have played a
role.
As there can be little doubt, that we are deal-
ing with the same event in both cores, one must
conclude, that “fingerprinting” the eruption by
chemical analysis of the anions is not a
straight-forward procedure, though it may be
helpful. In Fig 5 part of the Eldgjá acid signal in
the Camp Century core is shown. The data are
incomplete due to bad quality of the core. Only
data, which are comparable with the quality of
Dye 3 and Créte data are shown. The high acidi-
ties and the dating of the ice segment to 925 ± 15
justify the notion “Eldgjá signal”; thus the Eldgjá
eruption has been revealed in all the 3 Greenland
ice cores, which reach back to the time of the
eruption.
The Katla 1179 eruption
According to Thorarinsson (personal com-
munication, 1979) the only information we have
of a Katla eruption in 1179 derives from an old
script, in which it is mentioned, that a quarrel
took place in 1179 between the bishop Þorlákur
Þórhallsson and the mightiest temporal chief on
Iceland, Jón Loptson: The quarrel came up due
to a jökulhlaup in Höfði (i.e. from Katla), which
had destroyed several farms of which two had
churches. The jökulhlaup must have been large,
indicating an eruption in Katla, but no tephra
layer has yet been identified with certainty. The
short duration of the acid fallout, the precise
dating of the signal and its occurrence in all 3 ice
cores, Fig. 6, “smells” like an Icelandic eruption.
This is almost, what one can derive from the
ice cores, had it not been for the Dye 3 R value of
the eruption. The R value is quite low and the
peak value of the acidity profile is almost equal to
the Créte value. It must have been a fairly strong
eruption with respect to gas release, but not very
explosive with respect to sending large amounts
of acid gases into the stratosphere. Is it too much
to call it the Katla 1179 eruption? At least I will
do so, until someone disproves it. In order to give
the reader an impression of the non-volcanic acid
background a period of years before the 1179
eruption, is also presented in Fig. 6. The some-
what high acidities around 1168 (Créte) could be
due to an unknown eruption, as there is a fairly
60 JÖKULL 34. ÁR