Jökull - 01.01.2016, Síða 78

Þorsteinsdóttir et al.
Figure 7. Graph that shows CaO wt% values plotted against FeO wt%. The SILK-UN, N3, N2 and A8 have
higher values in CaO than the other 9 layers (Larsen et al., 2001). – Meðaltalsgildi (wt%) CaO á móti FeO
(Larsen o. fl., 2001).
DISCUSSION: CHANGES THROUGH
THE HOLOCENE
The main changes in the grain size of the 12 Holocene
SILK layers (2800 and 8100 years ago) from the Katla
volcano are that the younger layers (2800–5800 year
old) appear to be coarser grained than the older ones
(6000–8100 year old). An exception is the SILK-A7
which is coarser grained than the other older layers.
The factors that could cause differences in grain
size and grain morphology are 1) external factors, in
particular the presence and availability of meltwater
and 2) internal factors, such as changing composition
of the magma and the volume erupted.
Tephra produced in phreatomagmatic eruptions is
often characterized by being highly fragmented. The
ratio of water to magma affects the explosive activ-
ity and fragmentation of the magma (e.g. Wohlets,
1983; White and Houghton, 2000; Morrisey et al.,
2000; Francis, 2001; Francis and Oppenheimer, 2004)
and for this reason reduced fragmentation of erupting
magma and consequently coarser grained tephra in a
sub-glacial environment, could be the result of thinner
ice cover and therefore less available meltwater.
A general trend towards coarser-grained tephra
in the younger SILK layers is shown by the data in
Table 3. This could suggest a thinner ice cover 2800-
5800 years ago compared to when the older SILK
layers were forming. This is supported by the end-
moraines of Sólheimajökull which suggest a receding
ice cap since at least 4500 years ago, and the outer-
most moraine could be as much as 7000 years old
(Dugmore, 1989). The two largest SILK layers are
amongst the coarser grained younger layers, so the
duration of an eruption could also be important, with
longer eruptions creating a less wet environment due
to a reduction in the amount of ice in contact with
the magma over time. The location of eruption sites
may also have changed within the caldera to an area
of thinner ice. It should also be kept in mind that not
all tephra samples came from axes of greatest thick-
ness which could affect the result.
Grain shape analyses revealed that the SILK-A11
tephra layer is dissimilar to the other analyzed SILK
tephra layers. Firstly, SILK-A11 grains are less elon-
gated and this is supported by the circularity value of
A11 indicating that they are closer to being circular
in shape compared to the more elongated grains of
the other SILK layers. SILK-A11 grains also have
much rougher surfaces than those from the other lay-
ers. The reasons why this particular layer is so dif-
ferent from the other SILK layers are not obvious.
Perhaps there was less water present than in the other
eruptions or maybe the chemical composition of the
SILK-A11 tephra layer is important. Chemical analy-
sis of the SILK-A11 reveals that it differs from nearly
all the other SILK layers, but is very similar to SILK-
A12. However, as SILK-A12 has the typical elon-
gated grains found in the other SILK tephra layers,
78 JÖKULL No. 66, 2016