Jökull - 01.12.1991, Side 36

Figure 12 shows the distribution of ash in
Grímsvötn according to the vertical air photos. Two
months after a jökulhlaup, in September 15, 1954,
patches of ash were located in the northwestem part
of Grímsvötn. In August 9, 1960, seven months after
a jökulhlaup, isolated patches of ash could be seen in
the northwestern part, located at the same places as in
1954. No similar ash layers could be detected on air
photos from August 10, 1972. However, the firn was
dirty with ash near Naggur and at one location in
Svartibunki (Fig. 1). In August 22,1984 the ash layer
seen on earlier photos could no longer be detected
but ash was spread over parts of Grímsfjall and
around the 1983 crater. This ash layer is from the
1983 eruption and has not been observed after 1984.
During the period from 1934 to 1960 at least, it
depended on the time of the year whether tephra was
observed in Grímsvötn. With the exception of 1934
and 1935, very little or no tephra was detected in
Grímsvötn in winter and spring. In contrast, all obser-
vations in late summer and autumn show sizable
parts of the noithwestem corner of the area covered in
tephra. Moreover, photos show that the surface of the
ash-covered area was uneven and rough in late sum-
mer, similar to ablation areas of some outlet glaciers.
These observations raise the question whether the
very same ash layer was being observed throughout
this period (1934-1960). That would imply the whole
winter accumulation in the northwest part of
Grímsvötn being melted during the summer and that
the area was an ablation area indeed. For such an
ablation area to exist, surface melting would have to
be considerably greater than typically at the elevation
of 1400 m a.s.l. in Vatnajökull (Björnsson, 1988).
The mean winter balance in Grímsvötn is about 4.5
m of snow, equal to 2450 mm of water (Björnsson,
1985) whereas the summer balance under normal
snow conditions with albedo of 0.6 is of the order of
-500 mm.
Dispersal of ash over the glacier surface, decreas-
ing the albedo, may explain melting of the entire win-
ter accumulation. An average solar radiation flux of
200 W/m2 (as measured by Björnsson (1972) on
Bægisárjökull, North Iceland during July and August;
1100 m a.s.líj could melt the whole winter accumu-
lation in 70 days if the albedo were 0.2.
From 1934 to the 1960’s low albedo values were
maintained in the westem part of Grímsvötn during
summer. Wind-blown ash and loose material were
dispersed from nunataks and ash cones and ridges
which became free of snow early in the summer. On
these nunataks the geothermal heat flux was high and
the winter accumulation was reduced by snowdrift on
the uneven surfaces (see Figs. 7-8 and 13). This is
e.g. confirmed by photogrammetric measurements of
vertical air photos (H. Kristinsson, Hnit engineering
consultants) which show that the surface of the ash
covered area in the northwestern part of Grímsvötn
was uneven with a relief of a few metres on
September 19, 1946.
Figure 12 shows a gradual decrease of the ash cov-
ered areas between 1946 and 1972. Photos from the
area show that a gradual decrease in the extent of the
ice-free areas was also apparent by the early 1950’s.
That decrease was accelerated in the late 1960’s and
1970’s (Bjömsson, 1988). Apparently increased albe-
do retarded the melting of the winter accumulation
and the ash layers finally became buried by snow.
Further, the trend towards cooler summers in Iceland
since 1946 (Einarsson, 1989) and reduced ablation
may have accelerated the burying of the ash layer.
On the basis of these considerations we suggest
that the northwestem part of the Grímsvötn depres-
sion was an ablation area, at least over the period
1934-1960 and the very same ash layer was observed
by expeditions to the area throughout this period.
This ash cover dates back to the eruption of 1934 but
it may contain ash dispersed in earlier eruptions,
especially that of 1922. According to Fig. 12 the size
of this ablation area was about 20 km2 in 1945 and
1946. By 1972 the ablation area no longer existed.
Effects oftephra layers
on glacier mass balance
The existence of an ablation area within Grímsvötn in
the period 1934 to 1960 calls for an explanation of
greater surface melting than at present. As the data on
surface ablation within the Grímsvötn depression are
limited and sporadic, estimates of total ablation have
been fairly rough. Björnsson (1988) used available
data to estimate the mean ablation at present as being
34 JÖKULL,No. 41, 1991