Jökull - 01.01.2016, Qupperneq 44

Jonathan L. Carrivick et al.
Glacier surface morphology
Surge-type outlets of Drangajökull dip on aver-
age 6.5–7.5◦and small cirque surge-type glaciers in
Tröllaskagi peninsula dip 12–14◦(Björnsson et al.,
2003; Björnsson and Pálsson, 2008). Since the slope
of the vast majority of the surface of Kverkjökull is
>4◦, a surge there is not typical in an Icelandic con-
text. Furthermore, none of the glaciers mentioned in
Table 1 are generally steeper than 4◦ except perhaps
in small localised parts. The exception is the surge of
Horcones Inferior glacier because that glacier is steep
and fed by ice falls (Pitte et al., 2015). The surging
glaciers in the Karakoram are also generally low an-
gle in their ablation areas but with steep ice falls to
connect with accumulation areas.
The orientation of crevasses is normally depen-
dent on the direction of maximum extending strain-
rates, with a fracture typically forming perpendicu-
lar to the direction of maximum extending strain-rate
(Cuffey and Paterson, 2010). In our data, crevassing
in 2007 was generally restricted to the margins of the
glacier and to a few centrally-situated zones associ-
ated with proximity to bedrock outcrops (Figure 2A).
In contrast, in 2011 crevassing extended throughout
the central portion of the glacier and was intense along
the eastern margin and in the south-western part of
the terminus region (Figure 2B). Between 2008 and
2011, the evolving orientation of the crevasses close
to the centreline of Kverkjökull demonstrates differ-
ential flow between the middle of the glacier and its
margins. Longitudinally compressive stresses are in-
terpreted in the highly-crevassed central lobe, whereas
longitudinal extension is prevalent in the reservoir
zone in the upper glacier.
Our surface velocities derived from manual fea-
ture tracking in the terminus area (Figure 8) and in the
mid-elevation parts of the glacier near the Kverk (Fig-
ure 9) both suggest a deceleration of the glacier be-
tween 2012 and 2013 of ∼20 to 25%. This slowdown
is from velocities of up to 80 m yr−1 (Figure 8) but the
fastest sections of the glacier are probably those most
crevassed and they were moving so quickly that our
imagery did not have sufficient temporal resolution to
confidently track features (Figure 8). The fastest ve-
locities that we could measure relate to a 21 m dis-
placement over 30 days in 2012 (Figure 9), equiva-
lent to 0.7 m day−1, which is not uncommon for ‘nor-
mal’ (non-surge) ice flow. The bias in our velocity
measurements towards slower velocities probably ex-
plains why the surface speeds that we have measured
are low in comparison to the surface velocity of land-
terminating surging glaciers elsewhere (Table 1), but
might also be an indication that we are measuring the
final stages of the surge.
Our surface observations did not reveal any
change in the number or position of glacier river out-
lets, which have been noted to commonly occur dur-
ing other Icelandic glacier surges (Björnsson, 1998;
Björnsson et al., 2003).
Surge timing and duration
We are limited by data availability for determining
exactly when the Kverkjökull surge started and how
long it lasted. However, we can say that it started af-
ter 2008 (Figure 3) and before 2011. The surge front
was well progressed by 2009 (Figure 7) and had not
reached the glacier terminus in 2010 (Figure 6) but
had by 2011 (Figure 2C). The terminus continued to
advance through 2013 (Figure 8). Ice surface veloci-
ties apparently slowed between 2012 and 2013 (Fig-
ures 8 and 9). The discrepancy between the daily dis-
charge and seasonal volume of meltwater runoff be-
tween the years 2008, 2009 and 2014, and the years
2010, 2011, 2012 and 2013, in the Jökulsá á Fjöl-
lum and the Kreppa suggest that the Jökulsá á Fjöllum
river received an unusual amount of meltwater that
was not seen in neighbouring catchments and so un-
likely to be due to climate. Considering these obser-
vations together, the timing and duration of the surge
of Kverkjökull might therefore be tentatively put at
2009 to 2013, and 4 years, respectively.
The duration of the Kverkjökull surge is longer
than those at many temperate surge-type glaciers in
regions such as Alaska, which typically surge for 1 to
2 years (Jiskoot et al., 2000). Other Icelandic glaciers
that surge typically do so over ‘several years’ (Björns-
son et al., 2003). The Kverkjökull surge is, how-
ever, apparently rather short in duration in comparison
with some other high-latitude and polythermal glacier
surges, such as those in Alaska, Svalbard and Green-
land, which often last 10 to 15 years (Jiskoot et al.,
44 JÖKULL No. 66, 2016