Jökull - 01.01.2016, Blaðsíða 10

Eyjólfur Magnússon et al.
Figure 6. Perspective view from SW of Drangajökull ice cap (top) and bedrock (bottom) with twofold vertical
exaggeration. The DEM of the area outside the ice cap and attached ice and snow patches is the Lidar DEM
from 2011 (Jóhannesson et al., 2013). The largest lakes within the Lidar DEM coverage are shown (dark green).
– Séð yfir Drangajökull úr suðvestri með (efri) og án (neðri) jökulskjaldar. Stærstu vötn í nágrenni jökulsins
eru sýnd sem dökkgrænir fletir.
sented the topography indicated by the survey pro-
files, e.g. at a bottom of a valley or at a peak of a
mountain. In such cases additional contour line with
elevation chosen to appropriately demonstrate the to-
pographic shape where digitized with 25 m spatial in-
terval along the contour line. The extracted 10 m ele-
vation contours, the additional digitized contours, the
positions and corresponding elevations of the glacier
margin (including attached ice and snow patches) in
2011 (Lidar DEM) and the same part of the 1985,
1994 and 2005 DEM (value for every 20 m×20 m
cell) as above, were now merged into one single data
file. This file was used as an input for calculating the
final bedrock DEM of Drangajökull with 20 m×20
m cell size using kriging interpolation in Surfer. By
excluding the survey profiles from the actual final in-
put for the kriging interpolation some level of details
is lost at the locations of the profiles; the RMS dif-
ference of the interpolated product compared to the
bed elevation from survey files increases from 0.75
m for the first draft to 1.75 m for the final products.
The level of details is however much more homoge-
neous for the whole grid, making the appearance of
the DEM much more realistic, when viewed as shaded
relief image (Figure 5a-b).
10 JÖKULL No. 66, 2016