Jökull - 01.01.2016, Side 69

Reviewed research article
Grain characteristics of silicic Katla tephra layers indicate a
fairly stable eruption environment between 2800 and 8100
years ago
Edda Sóley Þorsteinsdóttir?, Guðrún Larsen and Esther Ruth Guðmundsdóttir
Institute of Earth Sciences, University of Iceland, Sturlugata 7, 101 Reykjavík, Iceland
?Corresponding author, eddas@mannvit.is
Abstract — Grain size and shape analyses were performed on silicic to intermediate Katla tephra (SILK)
formed 2800–8100 years ago to examine whether the grain characteristics had changed with time, and in
such a case, could they reflect changes in the eruption environment and/or changes in chemical composition.
No systematic changes with time were observed in the grain shape parameters (elongation, ruggedness and
circularity), however, the second oldest tephra layer SILK-A11 does not have the typical elongated grain shape
that characterizes the other SILK tephra layers. Chemical analyses indicate that the SILK layers can be divided
into three subgroups but no correlation between chemical composition and the grain parameters is observed.
Changes in grain size of the SILK tephra layers with time indicate an apparent increase in grain size occurring
about 6000 years ago, where largest grains are of category -3 Φ compared to 0 Φ in the older layers, all
sampled at similar distance from source. This change in grain size could result from variation in ice thickness
in the Katla caldera, with finer grain size between 6000–8200 years ago being due to thicker ice cover and
greater availability of meltwater for magma fragmentation. Conversely, the younger coarser grained SILK
layers may have formed under a thinner ice cover. A shift of the eruption sites to an area with thinner ice is also
a possibility. However, no radical changes in the eruption environment 2800–8100 years ago are demonstrated
by variations in grain characteristics, a conclusion further supported by large jökulhlaups from Mýrdalsjökull
ice cap during this period.
INTRODUCTION
It has been postulated that ice caps and glaciers in Ice-
land receded dramatically or even disappeared dur-
ing the Holocene climate optimum (e.g. Björnsson,
2008; Flowers et al., 2008). Óladóttir et al. (2007)
argued that the Mýrdalsjökull ice cap (Figure 1) has
been present for at least 8400 years, based on sul-
phur content of basaltic Katla tephra that indicates ar-
rested degassing of the erupting magma by quenching
in meltwater. Dugmore (1989) dated the outermost
moraines of Sólheimajökull outlet glacier to be over
3300 years old (3100 14C years BP), indicating a large
ice cap at that time. According to model calculations
presented by Björnsson (2008), the Mýrdalsjökull ice
cap did not exist 4000 years ago, but Björnsson as-
sumes that a caldera lake was present during the ice-
free period. However, 11 volcanogenic jökulhlaups
from the Mýrdalsjökull massif carrying volcanic ash
and pumice westwards into Markarfljót are known
throughout the 8400 year period (Larsen et al., 2005;
Gröndal et al., 2005; Smith and Dugmore, 2006; Egg-
ertsson, 2013). At least eight of the west-going jökul-
hlaups, dated between 7500 and 1200 years ago, em-
anated from the caldera via Entujökull, and the two
largest ones, ∼4400 and ∼3500 years ago, had peak
discharge of about 200,000 m3 s−1 (Gröndal et al.,
2005). This high peak discharge is better explained
by a sudden meltwater input than by rapid volume in-
JÖKULL No. 66, 2016 69