Jökull - 01.06.2000, Qupperneq 7

Guðrún Larsen
Figure 3. Main axes of thickness for some histor-
ical basaltic Katla tephra layers. Partly based on
Thorarinsson (1975) and Larsen (1978). Thin ar-
rows indicate minor tephra layers. The distribu-
tion is in fairly good agreement with prevailing wind
patterns at the 500 mb level (Jónsson, 1990). –
Meginþykktarásar nokkurra gjóskulaga frá Kötlu. Að
hluta til samkvæmt Sigurði Þórarinssyni (1975) og
Guðrúnu Larsen (1978). Grannar örvar tákna litla
gjóskugeira. Stefna gjóskugeira er í allgóðu sam-
ræmi við tíðni vindátta í 500 mb fletinum yfir Íslandi
(Trausti Jónsson, 1990).
An average eruption frequency of two eruptions
per century during the last 11 centuries is implied by
20 documented eruptions and/or tephra layers (Table
1). The maximum observed frequency is three erup-
tions in the 15th and 17th centuries. A similar erup-
tion frequency since ca. 7000 14C yrs BP is implied
by the number of tephra layers in proximal soil sec-
tions. A prolonged period of repose after the 10th
century Eldgjá event may have exceeded 200 years.
Katla tephra is coal-black to brownish black and
consists mostly of highly fragmented, poorly to mod-
erately vesiculated glass with grain sizes in the ash
and lapilli range. Crystals are scarce. The lithic com-
ponent, when present, consists of small light grey sub-
rounded rock fragments. The glass composition (Ta-
ble 2) of tephra from Katla is normally homogenous in
a single layer (the notable exception, layer K-x, being
part of the 10th century Eldgjá eruption). Layers from
individual Katla eruptions are difficult to distinguish
from each other on major element chemistry alone.
Most Katla tephra layers show distinct bedding
due to intermittent deposition and shifting wind
strength and wind directions during the eruption. A
fine grained lower part and a coarser upper part char-
acterize some of the layers (e.g. K 1357, Einarsson
et al., 1980), implying that the first erupted tephra
is more highly fragmented than that of later stages,
probably as a result of abundant meltwater at the erup-
tion site during the early stages of the eruptions. The
opposite has also been observed (e.g. K 1755, Guð-
mundsdóttir, 1998), indicating less favourable water
to magma mass ratio in the early stages, possibly as a
6 JÖKULL No. 49