Jökull - 01.01.2012, Side 85

Mass and volume changes of Langjökull ice cap, Iceland, ∼1890 to 2009
0 200 400 600 800 1000
Cumulative area (km2)
400
800
1200
1600
E
le
va
tio
n
(m
a
.s
.l.
)
0 20 40 60 80 100
Area (km2)
0 40 80 120 160 200
Cumulative volume (km3)
400
800
1200
1600
E
le
va
tio
n
(m
a
.s
.l.
)
0 4 8 12 16 20
Volume (km3)
Figure 2. Langjökull 1997: Upper graph: Surface
(light blue) and bedrock (grey) area distribution
(50 m elevation steps), and cumulative area (black,
dark blue). Lower graph: Volume distribution by
elevation (50 m slices) (red), and cumulative ice
volume (blue). – Langjökull: Dreifing og upp-
safnað flatarmál botns og yfirborðs (efri mynd) og
rúmmál íss (neðri mynd) með hæð.
DATA AND METHODS
In situ mass balance measurements
The mass balance of Langjökull has been measured
using a stratigraphic method; changes in thickness
and density are measured relative to the summer sur-
face (e.g. Paterson, 1994; Björnsson et al., 2003b).
The survey sites are situated along a number of ap-
proximate flow lines that cover the elevation range of
the ice cap, selected to describe the spatial variability
(Figure 4a). The mass balance values span the time in-
terval between given survey dates, which are not fixed
annually. The dates in the autumn are separated by
approximately one calendar year, which roughly co-
incides with the hydrological year between October
1 and September 30. The surveys in the spring were
carried out between late-April and mid-May.
Digital winter and summer mass balance grids
(Figures 4b-d) are produced by manually interpolat-
ing between the observed balance values. Mass bal-
ance contour lines are hand-drawn, digitized and a
matrix of grid cells (200x200 m) calculated using
kriging interpolation. Volumes are calculated by inte-
grating over the digital maps. Error limits for the area
integrals of the mass balance are cautiously assigned
as 5–15%.
Surface maps; construction and evaluation
In the bedrock radio echo sounding campaign of
Langjökull in 1997, the surface was also surveyed
with differential GPS along profiles about 1 km apart
(Figure 3); on average there are 10 m between points
along the profiles, a data redundancy that allows low
pass filtering along the profile to reduce random noise.
GPS base data for post-processing were collected at
the base camps (the first at the top of the ice cap’s
northern dome, the second 3 km south of the southern
dome summit; distance between base and rover from
0 to ∼25 km. The base stations were tied to several
permanent GPS stations. The point vertical accuracy
is estimated ∼1–3 m, mostly random noise; statisti-
cal analysis of ∼200 profile crossover points yields
standard deviation of 1.06 m. A surface DEM was
constructed from this data (Figure 3), with an average
elevation accuracy estimated <2 m. This is the first
surveyed map of the ice cap above 1100 m elevation.
JÖKULL No. 62, 2012 83