Jökull - 01.01.2012, Side 89

Mass and volume changes of Langjökull ice cap, Iceland, ∼1890 to 2009
above assumptions are in good agreement with the
elevation difference obtained between both the 1986
and 1997 DEMs, and 1997 and 2004 DEMs at the
highest areas of the glacier. Furthermore, the uncer-
tainty due to this ’ad hoc’ method yields only minor
errors in total volume change estimates as the eleva-
tion changes in the lower regions are by far larger.
Comparison of the elevation maps at mountain tops
and other prominent features did not suggest signif-
icant vertical shift. We assume the accuracy of the
1945 DEM to be ∼5 m and ∼7 m for the 1937 map.
We believe, however, that the average vertical bias is
much smaller, less than a few metres in all the recon-
structed maps. In error estimates of volume change
based on DEM difference we assume possible bias as
half of the random error.
Mapping of the LIA maximum margin
The LIA maximum extent of Langjökull was delin-
eated from geomorphological field evidence such as
lateral or terminal moraines, ice cored hummocky
moraines, fluted terrain and trimlines and mapped
from high resolution aerial photographs using remote
sensing software ArcGis. Historical documents, maps
and photographs from the 19th century to the early
20th century, along with field observations, detailed
oblique and aerial photographs support the estimated
LIA maximum extent (e.g. Wright, 1935; Sigbjarnar-
son, 1967; Geirsdóttir et al., 2009; Kirkbride and
Dugmore, 2006; Larsen et al., 2010)
Meteorological observations
In this study we primarily investigate the sensitivity
of the mass balance to temperature changes, using
data from the meteorological station Hveravellir in
central Iceland (location in Figure 1). The meteoro-
logical data from Hveravellir reach back to the year
1966. Hence, this data is supplemented with observa-
tions from a meteorological station at Stykkishólmur
in W-Iceland (the longest temperature record in Ice-
land, reaching back to the year 1822; location in Fig-
ure 1, e.g. Sigurðsson and Jónsson, 1995; Hanna et
al., 2004). The climate record from Stykkishólmur is
however damped due to the proximity to the ocean,
while the station at Hveravellir reflects inland temper-
atures (Björnsson et al., 2005).
RESULTS AND DISCUSSION
Mass balance from in situ observations
The average measured 1996–1997 to 2008–2009 mass
balance at sites along an approximate central flow line
down a south outlet of Langjökull is shown in Fig-
ure 6. The winter and summer balance is highly vari-
able; the standard deviation of measured balance at
each survey site is between 0.25 and 0.70 mwe yr−1
for both the winter and the summer balance, higher in
the accumulation zone for the winter balance and in
the ablation area for the summer balance. The bw gra-
dient (dbw/dz) is roughly linear by 0.4 mwe yr−1 per
100 m in elevation until reaching the highest peaks
where some of the winter snow is blown off. The
overall summer balance gradient (dbs/dz) is ∼0.7 mwe
yr−1 per 100 m in elevation, somewhat higher in the
lowest part and significantly lower in the upper accu-
mulation area; the gradient is mostly controlled by the
surface albedo (Gudmundsson et al., 2009b). The net
balance has a gradient (almost linear) of 1.1 mwe yr−1
per 100 m in elevation (the deviating net balance at the
highest elevation is excluded). The average ELA of
1996–1997 to 2008–2009 is ∼1090 m on Langjökull
southern dome, but about 1300 m on the north dome
(Figures 4d and 7). In Figure 7, the zero net balance
contour of the 2008–2009 bn-grid mostly coincides
with the dark/light boundary of an ENVISAT image
from 20 October 2009. The dark/light boundary is in-
terpreted as the boundary between ice or old firn and
the last winter snow residue.
The specific winter-, summer- and net balances
have varied between 1.1 and 2.1 (mean 1.74, std. dev.
0.33), -2.1 and -4.0 (mean -3.00, std. dev. 0.55), -0.4
and -1.9 (mean -1.26, std. dev. 0.48) in mwe yr−1, re-
spectively, from 1996–1997 to 2008–2009 (Table 1;
Figure 8). During these 13 years, the net balance has
always been negative and the total cumulative mass
loss 16.4 mwe yr−1 (Figure 9). Hence, the glacier has
lost 8.6% of its mass during this 13 year survey pe-
riod. Scatter plots, demonstrating the relationship of
bn to both bw and bs (Figure 10), indicate the zero
mass balance turnover b0−bal (bw = -bs) for the cur-
rent topography of Langjökull to be ∼1.8 mwe yr−1.
The average winter balance has been 1.73 mwe yr−1
or 96% of b0−bal while the summer balance average is
JÖKULL No. 62, 2012 87