Jökull - 01.01.2013, Blaðsíða 101

Mass balance of Mýrdalsjökull ice cap
Based solely on observational data, the precipita-
tion at Mýrdalsjökull can be compared with that on
the 1800–1850 m high plateau of the Öræfajökull ice
cap in Southeast-Iceland. There, the annual precip-
itation was also estimated to be approx. 8 mwe us-
ing similar methods, i.e. a comparison of five years
(1993–1996,1997–1998) of mass balance data and
lowland observations of precipitation (Guðmundsson,
2000). Summer temperatures are low at the eleva-
tion of the plateau and most of the summer precip-
itation is solid. There is little ablation and presum-
ably little water leaks through the snow layer, except
during warm summers when the temperature at the
plateau of Öræfajökull exceeds 0◦C. Öræfajökull is
approx. 500 m higher than Mýrdalsjökull and may
hence cause greater orographic uplift of the airmass
with the potential for larger amounts of precipitation.
However, due to its greater height, the airflow is more
likely to be to forced around Öræfajökull, than over
it, opposite to what may be the case when the airflow
impinges on a lower mountain like Mýrdalsjökull. It
is also of relevance in this context that an airmass
impinging on Öræfajökull will already have felt the
orographic (deflecting) effects of the Vatnajökull ice
cap. Mýrdalsjökull is the largest massif at the south
coast of Iceland and it is broader than Öræfajökull,
which may partly compensate for its lower elevation
with regard to the potential for orographic uplift. That
is, the impinging airmass, or larger parts of it, may
on average be more likely to be lifted over Mýrdals-
jökull compared to Öræfajökull, than to be forced
around it. An in-depth investigation into the atmo-
spheric dynamics controlling the interaction of the air-
flow with the complex orography during large precip-
itation events is necessary to describe this sufficiently.
This is beyond the scope of the current paper, but in
this context, unexplored data from the ÖREX obser-
vational project as well as numerical data from the
RÁV-project may hold some answers (Rögnvaldsson
et al., 2011).
Previous numerical studies are not decisive on
whether Mýrdalsjökull or Öræfajökull receives more
precipitation (see e.g. Rögnvaldsson et al., 2004,
2007; Crochet et al., 2007). Here, both the compar-
ison of winter mass balance within the accumulation
area of the ice cap with lowland observations of pre-
cipitation, as well as the atmospheric simulations, in-
dicate that the annual precipitation at Mýrdalsjökull
plateau is of similar magnitude (up to approx. 8 m of
water) as that at Öræfajökull (Guðmundsson, 2000).
This is the highest reported in Iceland; however, the
periods studied are not the same. The atmospheric
simulations reveal that the total precipitation (i.e. rain
and snow) may be even greater just off the south-
east edge of the plataeu of Mýrdalsjökull, or close to
10 m (Figure 5). These values are higher than those
found on Mýrdalsjökull by downscaling the analysis
of the ECMWF based on the linear model of Cro-
chet et al. (2007) for a different period; the maxi-
mum is slightly smaller than 7.5 m for 1961–1990
but it exceeds 7.5 in a more recent dataset valid for
1971–2000. Rögnvaldsson et al. (2007) found maxi-
mum values slightly greater than 8 m on the ice cap
during another period (1988–2003), based on results
from a dynamical downscaling of the analysis of the
ECWMF with a numerical atmospheric model, sim-
ilar to that used here but less advanced and run at
a coarser resolution of 8 km. Unfortunately, these
datasets do not cover the period of the mass balance
measurements on Mýrdalsjökull or have an overlap
with the simulated data presented here. The high reso-
lution of the linear model of Crochet et al. (2007) and
that in the current study is expected to be sufficient
so that the models may reproduce the spatial struc-
ture of atmospheric flow and the precipitation fields
on the ice cap (Ágústsson and Ólafsson, 2007). Inter-
estingly, the precipitation structure of the two methods
is similar with a maximum in the southeast of the ice
cap plateau. In Crochet et al. (2007), the maximum
is elongated to the northwest on the plateau, similar
to the winter precipitation simulated here while the
simulated annual precipitation has a more northward
elongated maximum (Figure 5). The simpler mois-
ture physics and airflow dynamics used in the linear
model of Crochet et al. (2007) compared to the cur-
rent model are expected to impact the precipitation
distributions. These effects can not be separated from
those introduced by differences in spatial resolution
and the periods covered by the models. Namely, the
precipitation distribution on the plateau will in gen-
JÖKULL No. 63, 2013 101