Jökull - 01.01.2012, Blaðsíða 50

Dugmore and Newton
Landnám erosion. Some areas have truncated se-
quences of Holocene soils because of episodes of ero-
sion driven by geomorphic processes such as glacia-
tion and fluvial action, but in areas where a high den-
sity of soil sections have been dug through to the
underlying substrate (such as south of lake Mývatn,
close to Öræfajökull and around Eyjafjallajökull), it is
apparent that this is not a complete explanation (Guð-
mundsson, 1998; Dugmore, 1987; Ólafsdóttir and
Guðmundsson, 2002). Within the surviving areas of
soil cover, basal ages show that soil cover has become
more extensive through the course of the Holocene,
and indicates an increasing cumulative total of su-
perficial fine sediment across the island as a whole.
Tephra production through the Holocene is likely to
have been a major driver of this change, especially the
very large tephras (such as Hekla 3 and Hekla 4) that
were deposited across the sparsely vegetated central
highlands (Larsen and Thórarinsson, 1977; Óladóttir,
2011b). In the uplands, cubic kilometres of tephra
would have remained potentially mobile for decades
to centuries after their initial deposition. This would
have provided large-scale sediment banks that could
be winnowed-out to create a fine-grained flux of aeo-
lian dust over the surrounding lowlands, and the raw
material for soil formation. With the eruption of more
tephra, more extensive soils could form. Modern ana-
logues for this process can be observed with both the
2010 Eyjafjallajökull and 2011 Grímsvötn tephras.
The consequence for tephrochronology is that older
records are more spatially-fragmented because with
increasing age the soils necessary to preserve tephra
are more limited in extent and increasingly patchy.
Despite impressive recent progress identifying
a very large number of the Holocene pre-Settlement
tephra layers from the Katla, Grímsvötn, Bárdarbunga
and Kverkfjöll volcanic systems that have dispersed
into the lands around the icecaps, only a small propor-
tion of pre-Landnám tephra layers have been mapped
in detail (Larsen et al., 2000, 2001; Óladóttir et al.,
2005, 2011b). When a lack of time, resources or incli-
nation mean that it is not possible to analyse the ma-
jor, minor and trace element compositions of all the
tephra layers encountered in a study, one response is
to effectively ignore the problematic tephras and con-
centrate on the well-known marker horizons such as
Hekla 3 and Hekla 4; this may provide sufficient res-
olution to tackle the questions being posed, and so be
entirely justified. There may, however, be significant
gains to be made from using the less straight-forward
deposits. For example, a prominent pre-Little Ice Age
’Eystriheiði’ high stand of Sólheimajökull can be con-
strained using the 871±2 AD Settlement tephra layer,
which lies on top of the outermost moraine and the
SILK YN tephra that is buried beneath it (Dugmore,
1989b; Dugmore et al., 2000; (Figure 3). The use
of well-known marker horizons alone would date the
glacier high stand to between c. 410 AD and c. 871
AD. It is however possible to achieve a better reso-
lution because around Sólheimajökull, SILK YN is
overlain by a basaltic tephra, both of which underlie
the moraine, and Landnám lies above a narrow black
tephra both of which overlie the moraine (Dugmore
1989). Although these two black tephras have only
been mapped in a limited area around Sólheimajökull,
their distribution across different geomorphological
settings shows that they are primary tephra deposits
and despite their unknown provenance (and indeed,
poorly known individual ages), they can be used to
narrow the likely age of the Eystriheiði stage to the
6th-7th centuries AD (Dugmore et al., 2000).
The comparatively stable, non-tephra, aeolian
sediment depositional regimes that existed before
Settlement mean that aggradation rates can be used
to successfully interpolate dates, an approach that
has been tested with independent radiocarbon dat-
ing (Dugmore 1987, 1989b; Óladóttir et al., 2005,
2011b). In the case of the Eystriheiði stage, non-
tephra sediment accumulation rates alone could have
been used to estimate the moraine ages. However, be-
cause of the variable contact between the moraine and
the underlying sediments onto which it was emplaced
and the uneven surface of the boulder moraine that
was later covered by soil, the uncertainties of such
age estimates would have been considerable. The un-
provenanced tephras lying stratigraphically close to
the moraine do, however, give a very good guide to
where effective applications of accumulation rate age
estimates can be made.
48 JÖKULL No. 62, 2012