Jökull - 01.01.2005, Page 128

Sverrir Guðmundsson et al.
Figure 7. Deviation between the 2004 total summer melting calculated from the energy budget maps and the
in-situ stake measurements of the summer balance, in m of water (a) and percent of the total melting (b);
see Figure 1 for location of stakes. – Munur á leysingu metinni útfrá orkubúskap annars vegar og aflestri á
jöklastikum hins vegar.
ance map. This deviation of∼1 m of water is however
close to the total summer accumulation measured by
the sonic echo sounder at the 1525 m AWS.
RESULTS
Energy balance and weather parameters at the
AWS sites
Annual variation in the average summer melting
(June-August) during the period 1996–2004 is mainly
caused by variation in net radiation (R), which is
closely connected to the surface albedo and the previ-
ous year’s winter mass balance (bw), i.e. low bw tends
to result in low albedo, and consequently a high R
(Figure 8a,c). Exceptions occurred during the sum-
mers 2002 and 2003 (Figure 8c). The mean summer
values of the incoming solar (Qi) and long-wave (Ii)
radiation did not vary considerably from 1996–2004
(Figure 8d). The turbulent heat fluxes (H) closely fol-
lows the annual temperature (TG) and wind-speed (u)
variations (Figure 8a-b) observed within the downs-
lope wind layer of the glacier (e.g. van den Broeke
1997; Oerlemans, 1998; Björnsson et al., in press).
For a melting glacier surface, the emitted long-wave
radiation (Io) is fixed at 315 Wm−2, resulting in
slightly negative net long-wave radiation balance (on
average Ii − Io ≈ −30 Wm−2) from June-August
1996–2004 (see Ii in Figure 8d).
Mean values of energy, surface and weather pa-
rameters during the summer months for 1996–2004
are shown in Figure 3a-d; the station at the ELA (1200
m) is chosen as an example because it contains the
longest time series on the glacier and has proven to be
representative for the whole outlet.
Good consistency was obtained between the daily
melting measured directly from a sonic echo sounder
and that calculated with Eq. (2) (Figure 4) during the
summer of 2004. Data for August 3 was, however, ex-
ceptional, with much higher melting measured at the
lowest station by the sonic echo sounder than calcu-
lated with Eq. (2). If precipitation of 80 mm d−1 is
assumed, no more than 8% of this difference may be
explained by heat supplied by precipitation (Hp) and
2% by frictional heat from water flowing everywhere
on the surface of Brúarjökull. The differencemight be
explained by frictional heat if it is assumed that water
flows in channels covering only 2–3% of the glacier
surface. Thus, an extensive amount of water flowing
in a channel formed below the sonic echo sounder is
128 JÖKULL No. 55