Jökull - 01.01.2016, Blaðsíða 36

Jonathan L. Carrivick et al.
only the Upptyppingar site receives meltwater directly
from Kverkfjöll so the pair of sites together provide a
useful assessment of whether meltwater runoff from
Kverkfjöll catchment is unusual.
RESULTS AND INTERPRETATION
The hill-shaded DEMs revealed considerable struc-
ture, mainly crevasses, on the ice surface, because
snow cover on the glacier is either absent or relatively
thin, at least below ∼1500 m a.s.l. (Figure 2A) due
to the late summer data acquisition. Additionally, the
hill-shaded DEMs indicated roughening of the ice sur-
face due to increased crevassing, notwithstanding the
possibility of more snow on the surface in 2011 than
in 2007, and expansion laterally of the lower parts of
the glacier against lateral moraines and cliffs (Figure
2B). The DoD (Figure 2C) showed a pattern of sur-
face elevation changes between 2007 and 2011 that
was negative at higher elevation and positive at lower
elevation, which is remarkable because it is the oppo-
site of what could be expected due to an overall neg-
ative mass balance regime of a normal glacier and ice
dynamics. The magnitude of these surface elevation
changes; tens of metres, and most importantly the spa-
tial pattern of elevation changes along the glacier, can-
not be explained by snow cover being thicker in one
year or the other. Based on these variations in surface
topography we infer that Kverkjökull surged during
the 2007 to 2011 period. The boundary between the
reservoir and receiving zones, which can be identified
by a zero elevation change isoline, was complex and
ranged from 1400 to 1528 m a.s.l. and outlines at least
three lobate-shaped areas of the glacier (Figure 2C).
Comparison of our dGPS measurements of spot
elevations in 2007 with the ALS-derived DEM of
2007 have excellent agreement, and comparison of
our 2008 dGPS elevations with the 2007 ALS data
show little change (Figure 3), together giving us con-
fidence that the surge had not yet started, or at least
not reached the terminus, in August 2008. Further-
more, comparison between our dGPS measured el-
evation and those in the 2008 and the 2011 DEM
demonstrate that by 2011 the north-eastern and south-
western portions of the Kverkjökull terminus were
behaving differently. Specifically, the north-eastern
part of the glacier showed surface lowering, i.e. thin-
ning, as would be expected due to a negative mass
balance, or during the quiescence period of a surge-
type glacier, whereas the south-western portion of the
terminus had thickened, with the greatest thickening
towards the terminus, as indicative of the passage of a
surge front (Figure 2C).
Figure 4. Ice thickness estimated via a 1D steady state
‘perfect-plasticity’ model. – Ísþykkt metin með ein-
víðu, æstæðu, plastísku líkani.
The asymmetric impact of the surge on ice sur-
face elevations in the terminus area was conspicu-
ous and warranted further investigation. Although our
overall pattern of estimated ice thickness, based on
a one-dimensional centreline flow model, looks rea-
sonable (Figure 4) we have no direct measurements
of the detailed form of the bed geometry. How-
ever, the increased crevassing and amplified hum-
mocky ice-surface undulations in the 2011 DEM (Fig-
ure 2A,B) that extend from the centre of the termi-
nus in a south-easterly direction do not correspond
36 JÖKULL No. 66, 2016