Jökull - 01.01.2012, Side 129

Glacier changes in the marginal zone of Sólheimajökull
(Figures 3-4). This places the margin c. 4 km fur-
ther south than today. According to Dugmore (1989),
some of the lateral moraines overlie soils formed 3.1–
1.7 kyr BP but are overlain by soils of ages up to 1.2
kyr BP. Firmly established tephrochronology by Dug-
more et al. (2000) indicate that the lateral moraines
were deposited 1.4–1.3 kyr BP. Maizels (1991) de-
scribes a fan of jökulhlaup sediments (Húsa Fan) with
an apex suggesting that meltwater drained through
Ystagil and the neighbouring canyon to the east, Hóls-
árgil, at the time of deposition, which is consistent
with an extended glacier. 14C datings of wood in the
fan indicate that it is younger than 2060 yr BP. The di-
rect dating of the lateral moraines with CED therefore
suggests ages that are 500–600 years older than the
tephrochronologically established ages (Dugmore et
al., 2000). The ’outlier’, SOL-02, however, suggests
an age 200 years younger than the previously estab-
lished ages. This highlights the challenging large un-
certainties associated with such young samples (Ta-
ble 4). Sample SOL-01 of striated basalt from the
top of Jökulhaus indicates that this bedrock knob was
deglaciated 1.7 kyr BP. Because this predates the LIA,
it is suggested that the very top of Jökulhaus was not
ice-covered during the LIA, unless subglacial erosion
was too little to zero the ’cosmogenic clock’ in the
rock.
The two 14C datings from section 2 indicate de-
position of organic material in AD 1512–1539 (Ta-
ble 3). This is in the early LIA, and section 2 sug-
gests that Sólheimajökull advanced after that time, de-
positing the end moraine ridge. According to Thor-
arinsson (1943), the LIA advances of Sólheimajök-
ull culminated in AD 1705, 1794 and 1820. Dug-
more (1989) concluded that the end moraine ridges
and other glacial landforms in the outermost part of
the valley originate from advances in the 10th century
AD. This implies that the upper tills (units 9-11) of
section 1 are most likely from the 10th century AD
and, hence, the sediments below are older than that.
The time series of DEMs (Figures 9–11) display a
spatial view of the glacier snout changes 1960–2010
and are important supplements to the ice marginal
measurements (Figure 2). Future work on the DEM
time series from Sólheimajökull should focus on com-
pletely covering the entire glacier surface in order to
be able to produce more spatially complete change
maps and derive the total volume change and geodetic
mass balance. Similar maps have successfully quanti-
fied the thickness of surged glacier margins of Vatna-
jökull (Magnússon et al., 2005) and long-term spa-
tial changes of glaciers (Barrand et al., 2010; Mo-
tyka et al., 2010). Recently, Aðalgeirsdóttir et al.
(2011) pointed out the value of time-series of DEMs
to force and constrain models of glacier mass balance
and ice flow. Thus, good data covering the 20th cen-
tury glacier history are important prerequisites to sim-
ulate future glacier response to climate in the 21st
century. As more high-resolution and high-accuracy
DEMs become available, changes in glacier forefields
(erosion/deposition) can also be identified and quanti-
fied on difference DEMs (e.g. Schomacker and Kjær,
2007; Schomacker, 2008).
CONCLUSIONS
The forefield of the Sólheimajökull outlet glacier
was shaped by glacier oscillations during the late
Holocene. The most prominent landforms are numer-
ous series of end moraines, some of which have a saw-
tooth pattern.
Southeast of the present glacier, very distinct lat-
eral moraines are present. They are up to 1300 m long
and document a much larger extent of Sólheimajökull
than presently.
Cosmogenic exposure datings suggest that the
lateral moraines formed c. 1.9 kyr BP which is
500–600 years older than previously documented
by tephrochronology. Hence, where accurate
tephrochronological dating is available, it becomes
obvious that cosmogenic dating is not comparable in
either precision or accuracy. However, in many set-
tings cosmogenic exposure dating is the only option,
and the method has a good potential although the ac-
curacy should not be expected to be as high as with
tephrochronology, particularly not for young samples.
14C datings of organic material document that the
LIA advances took place after AD 1539. The Little
Ice Age advances were much more restricted than the
advance delimited by the prominent lateral moraines.
JÖKULL No. 62, 2012 127