Jökull - 01.01.2005, Page 145

Mass balance of Vatnajökull outlet glaciersreconstructed back to 1958
Figure 6. Change in terminus position versus cumu-
lated bn over 1973-1992. Glaciers are identified in
Table 1. icelandic
For all five series the shift was between 1994 and
1995, which is expressed in Figure 5 by change in
mean slope of the cumulative curves. The 5-glacier
mean of the change from 1958–1994 to 1995–2003
was ∆b∗
n
= −0.66 m/a w.e., most of it coming from
increased ablation. From the first part to the second,
∆b∗s = −0.54 and ∆b∗w = −0.12. The negative shift
in the piecewise-constant fit of b∗
n
(t) corresponds to a
negative change of mean slope in cumulative
∫
b∗
n
(t)
shown in Figure 5.
CONCLUSIONS
The upper-air model gives a good fit to observed bal-
ances, comparable to published results from two other
models (Table 3), one using observations form exist-
ing climate stations, the other using meteorological
observations made in one season on the ice. Its cal-
culated temperature sensitivity of net balance bn is
intermediate between those determined by the other
models (Table 4). Its calculated precipitation sensitiv-
ity is less than that from the other models.
The upper-air model has the advantage of using as
its input a database maintained as an integral part of
an ongoing major scientific enterprise. Calculations,
therefore, can easily be extended as future mass bal-
ance measurements become available. Moreover, the
model can be used to estimate mass balance for years
in which it is not measured. It could also be easily
applied to mass balance records for other glaciers in
Iceland. Using upper-air data only at 0 UTC would
probably give results nearly as good as those obtained
here by using values four times a day.
Spatial variation of cumulative reconstructed bal-
ance
∫
b∗
n
(Figure 5) closely matches that in bn over
the period of observations (Table 1). The correspond-
ing differences of
∫
b∗
n
from that of Brúarjökull for
glaciers 1-5 are -37,-12, +10, 0,-28 m w.e. Model cal-
ibration, however, is over rather short periods of ob-
servation (Table 3).
Mean change of slope of the reconstructed
∫
b∗
n
(Fig. 5) corresponds to a change of net balance b∗
n
between the mean over 1958–1994 to the mean over
1995–2003. The five-glacier average of the change
was -0.66 m/a w.e. It was caused by mean winter bal-
ance becoming 0.12 less positive and summer balance
becoming 0.54 more negative between the means for
the two periods.
Spatial variation of critical direction φ′ of the 850-
hPa wind at the Reanalysis gridpoint (Figure 4) seems
to be physically reasonable. For each glacier, the
optimum direction for calculating precipitation flux
(Eqs. 1,2) avoids flow over the top of the icecap. The
range of mean bw is small, 0.39 m/a w.e. between
glaciers 2 and 5 (Table 1), but spatial coherence of
interannual variation is relatively weak (Table 2), re-
flecting the varying sensitivity to wind direction from
glacier to glacier.
By contrast, spatial variation of bs is large,
1.07 m/a w.e. between glaciers 3 and 5, but interan-
nual variations are much more coherent (Table 2). In-
terannual variation of temperature affects bs similarly
at all glaciers, whereas interannual variation of wind
direction has an effect that depends on the critical di-
rection φ′ for each glacier.
Acknowledgements
This work was funded from US National Science
Foundation grant OPP-0240861. Thoughtful com-
ments by two anonymous reviewers led to a much
JÖKULL No. 55 145