Jökull - 01.01.2015, Qupperneq 34

Þorsteinsdóttir et al.
EXPLOSIVE VOLCANISM:
WET VERSUS DRY ERUPTIONS
Explosive eruptions are of two different kinds, mag-
matic explosive eruptions were the explosive ac-
tivity is first and foremost caused by gas ex-
pansion in the magma and secondly hydromag-
matic/phreatomagmatic eruptions which are caused
by interaction between external water and magma
(White and Houghton, 2000; Morrisey et al., 2000;
Vergniolle and Mangan, 2000; Cashman et al., 2000;
Francis, 2001; Francis and Oppenheimer, 2004).
Magmatic eruptions
Magmatic eruptions (sometimes called dry eruptions),
are divided into several types, for example Hawai-
ian, Strombolian and Plinian eruptions. They are as
mentioned earlier, eruptions where only volatiles take
place in the explosive activity. The two basic pro-
cesses of tephra formation are fragmentation of the
magma and the cooling of the fragments or clasts
(Cashman et al., 2000; Vergniolle and Mangan, 2000;
Francis and Oppenheimer, 2004).
In the following only silicic magma is considered.
Fragmentation of magma may occur when the ten-
sile strength of the melt is exceeded. If strain rates
are very high the magma cannot deform fast enough
and the melt can break as a solid matter would, in
a brittle fashion (e.g. Cashman et al., 2000; Zhang,
1999). Fragmentation in magmatic eruptions where
the magma is high-viscosity silicic magma may result
either from rapid acceleration or rapid decompression
of magma according to Cashman et al. (2000).
Hydromagmatic eruptions
Hydromagmatic eruptions are divided into three cat-
egories: Phreatic eruptions where magma is not
directly involved, phreatomagmatic eruptions and
phreatoplinian eruptions where the fragmentation is
partly or solely the result of interaction with water
and can produce extremely fine-grained tephra. Frag-
mentation in hydromagmatic eruptions is much more
complex process than in magmatic eruptions. The
presence of external water can affect both fragmenta-
tion and cooling at various stages in an eruption (e.g.
Morrisey et al., 2000; White and Houghton, 2000; Zi-
manowski et al., 1997; Wohletz, 1983).
Fragmentation in most of the phreatoplinian erup-
tions described by Houghton et al., (2000) was
brought about by vesiculation and bubble expansion
as well as by quenching by external water. The
pumice clasts were vesicular enough to imply that
the magma had already formed foam and perhaps
begun to fragment or disintergrate when it first en-
countered external water. The fragmentation resulting
from the magma-water interaction could be the result
of thermal contraction upon quencing, brittle failure
caused by high shear rate resulting from expansion of
steam or both these mechanisms. Lithics appear to be
less common in phreatoplinian eruptions than in the
basaltic ones. This could suggest that the fragmenta-
tion caused by water takes place at the interface be-
tween the magma and overlying ice/water, rather than
deeper in the conduit (Dellino et al., 2012).
METHODS
Tephra samples were collected along the thickness
axes of the SILK-LN and H-1947 tephra layers from
undisturbed tephra. Samples in two cross sections
of SILK-LN tephra were also collected. At each lo-
cation the tephra layers were cleaned, photographed,
measured and macroscopic features described such as
bedding, grading, colour, texture, grain size and grain
types (for details see Þorsteinsdóttir 2015). Following
logging, bulk samples and/or samples from particu-
lar units within the layer were collected. Samples of
the Katla SILK-LN tephra (Figure 2) were collected
from 10 previously logged soil sections (Larsen un-
publ. data). The H-1947 samples were collected in 3
places (Figure 4) in addition to the field and laboratory
measurement of Thorarinsson (1954).
Grain size analysis
Sieving (by hand to avoid breaking/abrading) was
used for size fractions larger than 4Φ (0.063 mm) and
settling velocity for smaller than 4Φ. A Sedigraph
(Micromeritics, 2010) was used for grains smaller
than 4Φ. The results from both methods were com-
bined and plotted on a graph, showing the complete
grain size distribution of each tephra sample.
Grain size distribution can provide important in-
formation about an eruption because different erup-
tion conditions influence both size and shape of the
34 JÖKULL No. 65, 2015