Jökull - 01.07.2003, Blaðsíða 9

A calibrated mass balance model for Vatnajökull
Figure 4. Mean energy used for melting ( ) during the 1996 experiment. Only those stations are shown where
the net radiation was measured during the entire experiment. Shown are the values obtained from the observed
melt ( ), the simulated values without the effect of insulation ( ) and the simulated values with this ef-
fect ( , ). The simulated values are obtained from the observed net radiation and the turbulent fluxes
calculated from the observed 2 m values of meteorological variables. – Orka sem fór til leysingar á ýmsum
mælistöðum sumarið 1996 (mælt og reiknað).
a.s.l.) the ablation season lasts about 120 days and at
station U3 (169 m a.s.l.) about 220 days.
Both the ablation rate and the accumulation rate
of an ice body are largely determined by the condi-
tions of the atmosphere that surrounds it. We assume
all meteorological variables in the free atmosphere,
except precipitation, to be horizontally homogenous
around Vatnajökull. This seems justified, for the mean
temperatures of three stations that lie at very different
locations are nearly the same when they are extrapo-
lated to sea-level: 9.3, 9.4 and 9.3ÆC for A1, U10 and
R1, respectively. The tuning of the emissivity of clear
skies (Appendix) also indicates that the free atmo-
sphere was horizontally homogeneous, for all stations
display the same results. Because of this assumption
the model can be forced with data from a single per-
manent weather station. We use a permanent weather
station outside of the ice cap so that we can recon-
struct a long mass balance record and study the sen-
sitivity of Vatnajökull to external climatic changes.
There are three such stations close to Vatnajökull:
one in Hornafjörður, one in Fagurhólsmýri and one
in Kirkjubæjarklaustur (Figure 1). Hornafjörður is
much more influenced by clouds and damp air from
the ocean than most of Vatnajökull. The station in
Fagurhólsmýri has no long records of humidity, so the
station in Kirkjubæjarklaustur (named KIR hereafter)
JÖKULL No. 52, 2003 7