Jökull - 01.01.2016, Blaðsíða 34

Jonathan L. Carrivick et al.
Table 2. WorldView imagery used in this project. – WorldView gervihnattamyndir sem notaðar eru í greininni.
Catalogue Id. Imaging Space- Acq. Total max Area max Area Area Total Area Res.
bands craft date off nadir off nadir min max cloud cloud after
angle (◦) angle (◦) sun elev. GSD cover cover resamp.
(km2) (km2) (%) (%) (m)
1020010016648200 Pan WV01 22nd Sept. 18.68 18.04 25.50 0.55 1 1 0.5
MS1 2011
MS2
103001001A4FCF00 Pan WV02 2nd Aug. 14.14 10.86 42.80 0.48 22 0 0.5
MS1 2012
MS2
103001001B396A00 Pan WV02 7th Sept. 15.13 14.90 31.01 0.50 1 2 0.5
MS1 2012
MS2
1020010023C5D200 Pan WV01 25th July 27.19 24.25 44.21 0.59 16 0 0.5
2013
1020010023851F00 Pan WV01 19th Aug. 24.37 19.54 37.67 0.56 26 19 0.5
2013
Einarsson, 1990), and (ii) in the immediate vicinity of
Gengissig (Montanaro et al., 2016).
Therefore, in this study we have estimated the bed
elevation of Kverkjökull by application of a simple
steady state ‘perfect plasticity’ one-dimensional (1D)
model applied along a centreline profile of the glacier
surface as represented in the 2007 DEM, and extrapo-
lated in 3D across the 2007 glacier area, following the
workflow of James and Carrivick (2016). Specifically,
we used a valley-wide yield stress of τ = 130 kPa
for the model, as suggested by Hoelzle et al. (2007),
from Haeberli and Hoelzle (1995), as representative
of alpine (European Alps) glaciers and including con-
sideration of a shape factor, f (James and Carriv-
ick, 2016). We interpolated our bed elevation from
along the centreline over the entire area of the glacier
using the ANUDEM algorithm (Hutchinson, 1989),
which is an interpolation routine designed to generate
hydrologically-correct surfaces from surrounding ele-
vation data and which is commonly applied to recon-
structing subglacial bed topography (James and Car-
rivick, 2016).
This bed elevation estimate must be considered
as a first-order estimate only and it will not repre-
sent possible near flow-parallel features that may be
present in the bed geometry. In terms of the appli-
cability of this model, in lieu of any other suitable
data for making bed elevation estimates, we consid-
ered firstly that a visual inspection of the 2007 DEM
(Figure 2A) showed no evidence of the onset of a
surge, such as a bulge as reported at other glaciers by
(Clarke et al., 1984, for example), or crevassing in-
dicative of recently enhanced ice flow such as shear-
ing of ice along the glacier margins or longitudinal
crevassing (c.f. King et al., 2015). Secondly, quan-
titative comparison of the contours generated from
the 2007 DEM with contours from other (lower res-
olution) mapping efforts (see Carrivick and Twigg,
2005) show good agreement between the two datasets.
Thirdly, the Kverkjökull surge (as we have measured
it between the 2007 and the 2011 DEMs) only con-
stitutes elevation changes of up to a few tens of me-
tres (Figure 2C). Fourthly, we note that the surface
elevation at the margin of Kverkjökull next to lateral
moraines in 2007 was typically just ∼50 m lower than
the symmetric crests of LIA lateral moraines and thus
has apparently changed little over the past one hun-
dred years or so, although we accept that the lateral
moraine is probably ice-cored and could have lowered
as well. These four sets of observations in combina-
tion lead us to regard that the 2007 DEM represents
a glacier in a non-surge state. We assumed that the
bed elevation as estimated from the 2007 data was the
same in 2011 and thus permitted calculations of ice
thickness in 2007 and in 2011 as well as an evaluation
of the change in driving stress, τ :
∆τ = ρgH2 tanα2 – ρgH1 tanα1
34 JÖKULL No. 66, 2016