Jökull - 01.01.2012, Side 96

F. Pálsson et al.
Table 4. Estimated mass balance sensitivity of Langjökull to 1◦C temperature rise (e.g. Jóhannesson, 1997)
at both Hveravellir and Stykkishólmur, using averages over i) all days in the time interval and ii) using only
the corresponding summer months (June-August). Changes in precipitation are ignored in the error estima-
tions (locations in Figure 1). – Næmni afkomu Langjökuls fyrir breytileika meðalárshita og meðalsumarhita í
Stykkishólmi og á Hveravöllum.
All seasons Summer
Reference periods ∆t1 and ∆t2 δbn/δT δbn/δT
(mwe yr−1 ◦ C−1) (mwe yr−1 ◦ C−1)
(a) Using temperature at Hv: Hveravellir
∆t1: 1986 – 1997; ∆t2: 1997 – 2004 -1.35 ± 0.35 -0.90 ± 0.20
∆t1: 1986 – 1997; ∆t2: 1997 – 2009 -1.15 ± 0.25 -0.90 ± 0.20
(b) Using temperature at St: Stykkishólmur
∆t1: 1937 – 1945; ∆t2: 1945 – 1986 -2.15 ± 0.90 -2.20 ± 0.90
∆t1: 1986 – 1997; ∆t2: 1997 – 2004 -2.35 ± 0.55 -1.70 ± 0.40
∆t1: 1986 – 1997; ∆t2: 1997 – 2009 -1.60 ± 0.40 -0.85 ± 0.20
the annual averages. Precipitation and temperature
records from the Stykkishólmur and Hveravellir sta-
tions are highly correlated (Figure 15). However, the
higher mass balance sensitivity to uniform tempera-
ture rise at the coastal station Stykkishólmur than of
the inland Hveravellir is explained with oceanic con-
straint of the coastal temperatures, demonstrating that
the mass balance sensitivity calculations may strongly
depend on the location of a meteorological reference
station. In a model simulation study for Langjökull,
Guðmundsson et al. (2009a) obtained mass balance
sensitivity of -1.15 mwe yr−1 to an annual 1 K temper-
ature rise at Hveravellir, which is in a good agreement
with our results in Table 4.
In another study, Gudmundsson et al. (2011) used
the Hveravellir meteorological station to investigate
the mass balance sensitivity of the small Eyjafjalla-
jökull, Tindfjallajökull and Torfajökull ice caps (Fig-
ure 1) to uniform temperature rise. Their highest
sensitivity number was obtained for the maritime ice
cap Eyjafjallajökull, or -2.80 mwe yr−1 K−1 com-
pared to -1.37 to -1.15 mwe yr−1K−1 (using all sea-
son average) found in the present study for the inland
Langjökull ice cap. The mass balance sensitivity for
the neighbouring Hofsjökull ice cap (location in Fig-
ure 1) is however around 75% that of Langjökull, ex-
plained by the 200–300 m higher elevation range of
Hofsjökull (Gudmundsson et al., 2009a). Jóhannes-
son et al. (2011) found -1.90 mwe yr−1 K−1 (using
Stykkishólmur) for Snæfellsjökull an ice cap in W-
Iceland (Figure 1), an ice cap with similar elevation
range as Langjökull but much smaller. Anderson et al.
(2010) obtained a mass balance sensitivity of -2 mwe
yr−1 K−1 for the maritime Brewster Glacier in New
Zealand. Their number is comparable to the mass bal-
ance sensitivity obtained for Icelandic ice caps.
CONCLUSION
Although old surface elevation maps of glaciers may
be distorted laterally and shifted vertically due to er-
roneous triangulation sites and sparse or incomplete
survey, some may be corrected sufficiently and used
to realistically deduce volume change estimates and
average mass balance. This is particularly the case
for differential DEMs representing long time spans.
Average specific mass balance derived from low error
surface DEMs (7 years apart), produced from dense
GPS profiles and SPOT5 HRG images, is in close
agreement with the average specific mass balance ob-
served with in situ measurements. This indicates that
the set of 23 mass balance site are successfully se-
lected to describe both the lateral and vertical mass
balance variability on the ∼900 km2 Langjökull ice
cap. The observations of average mass balance pre-
sented in this paper, span more than 110 years with
94 JÖKULL No. 62, 2012