Jökull - 01.01.2016, Síða 20

Eyjólfur Magnússon et al.
with 2001–2006 surge of Reykjarfjarðarjökull (Sig-
urðsson, 2003; 2007; Þrastarson, 2006). D6 was at
similar locations in 2011 as D2 in 1946, 8 years af-
ter the 1933–1938 surge of Reykjarfjarðarjökull (Sig-
urðsson, 1998). D6 seems to be formed slightly up
glacier from a step in the bedrock topography (Figure
15c). Assuming that D1–3 were formed during the
1930s surge, their forming locations may also be re-
lated to the same step in bed topography, which passes
the 1946 locations of all these depressions. It is how-
ever difficult to relate the formation location of D4,
D5 and D7 to any particular feature in the bed topog-
raphy. Even though the depressions are initially large
crevasses, they are then mostly filled by snow accu-
mulation; many crevasses in 2005, appearing similar
to the preserved depressions, had disappeared in 2011.
The ones preserved strongly effect the surface topog-
raphy in their vicinity in the SW-NE direction (or up
and down glacier), likely through snow drift and wind
scouring (Figure 11) that probably also play a crucial
part in preserving the depressions.
Vectors indicating the observed rate of displace-
ment were obtained for depression D3 (Figure 15a).
For comparison, surface velocity vectors from Stokes
ice flow model obtained using the Icetools library
(Jarosch, 2008) run with the finite element pack-
age Fenics, were also calculated. The model ob-
tains 3D velocity fields caused by ice deformation
for the whole ice cap, using the bedrock DEM and
the glacier surface DEMs at corresponding dates (ex-
cept for 1994, where 1985 DEM was used due to
only partial coverage of the 1994 DEM). The same
ice flow rate factor as in Belart et al. (in open review,
2016) was used; the rate factor was scaled to fit ob-
served stake displacement on Leirufjarðarjökull and
S-Drangajökull in the summer 2013–2014. Compari-
son of the modelled surface velocity and the observed
displacements indicates similar magnitudes of dis-
placement except during the period 1994–2005, that
includes the last surge of Reykjarfjarðarjökull. There
is however a significant difference between the ob-
served and modelled displacement direction. The ob-
Figure 15. a) The area of steep surface depressions in the accumulation area of Reykjarfjarðarjökull (red box
in bottom right corner image). The ice cap surface in 2011 is shown as a shaded relief image with contours
overlain (10 m contour interval). The dot-line polygons indicate outlines of the depressions (named D1-5), as
seen in 1946 aerial photographs (Belart, 2013; Magnússon et al., 2016) and at various times since then. The
color-coding with dates shown for D3 applies to all traced depressions. Depressions D2 and D3 are still clearly
detected in Pléiades images in October 2014 (Belart et al., in open review, 2016), while D1, D4 and D5 have
disappeared from the surface. The black arrow indicates the rate of displacement for D3, while red arrow in-
dicates modelled surface motion applying Stokes ice flow model. b) Orthorectified aerial photograph of the
same areas acquired 27 July 2005 by Loftmyndir ehf. showing numerous new similar depressions including
D6 and D7 (also shown in a). Black lines indicate locations of RES-profiles and blue line the location of the
RES-profile shown in c, near the tracked path of D2 (blue line in b). The roman number in b and c indicates
locations where the radar passed cauldrons. Features in the RES-profile at shallow depth under the numbers
are probably side reflections from the depressions. – a) Skuggamynd ásamt hæðarlínukorti (10 m hæðarlínu-
bil) sem sýnir jökulyfirborð, í júlí 2011, á hluta ákomusvæðis Reykjarfjarðarjökuls. Þar er að finna sérstæðar
krappar dældir í jökulyfirborði. Punktalínuform sýnir staðsetningu dældanna á ýmsum tímum frá 1946 til 2014.
Litir á punktalínum gefa ártal staðsettra dælda sem sýnt er fyrir dæld D3. Ferla D2 og D3 í yfirborði jökulsins
má rekja í gegnum allt tímabilið meðan D1, D4 og D5 hafa horfið á tímabilinu. D6 og D7 sjást hins vegar ekki
í tiltækum myndum fyrir 2005. Svartar örvar sýna mældan yfirborðshraða og stefnu D2 en rauðar örvar hraða
og stefnu hans reiknaða með ísflæðilíkani. b) Upprétt loftmynd af sama svæði þann 27. júlí, 2005, sem sýnir
Reykjarfjarðarjökul í framhlaupi og fjölda svipaðra nýrra dælda í yfirborðinu þ.m.t. D6 og D7. Svartar línur
sýna legu íssjársniða. Blá lína sýnir legu íssjársniðs sýnt á c. Rómverskar tölur tengja saman staði í sniði (c)
við staðsetningar á loftmynd (b).
20 JÖKULL No. 66, 2016