Jökull - 01.12.1977, Qupperneq 26

place. Similarly, snow drifts (landscape) on the
glacier surface at tlie time of the ash-fall miglit
cause deviations in the X’s, as compared to the
ideal model in whicli the glacier is perfectly
flat. To estimate the magnitudes of these ex-
traneous effects a simple calculation may be
made: Let a data point in the rniddle of a ten
year periocl be misplaced by one year. It would
cause a pair of spikes, in opposite directions,
] .88 m/yr and 2.82 m/yr from the mean value
2.26 m/yr. In Fig. 8 the most conspicuous pair
of spikes occurs for the periods I88S—87 (Xi =
4.85) and 1887-92 (\2 = 0.88). To “iron out”
these spikes the 1887 layer must be brought to
1890 (\i = 2.77, Ao = 2.2), which still would
yield the same relationship (positive — negative).
Only by bringing the 1887 layer up to 1891 the
relationship is reversed. Similarly for the early
1700’s.
Recalculating the annual balance into snow
gives 2 and 11 m/yr respectively, for the periods
1883—87 and 1887—92. Tliis seems totally un-
reasonable. The winter balance of Vatnajökull
was first measured in 1936—38, and lias been
measured annually since 1953. In Grímsvötn
the balance has ranged from 5 to 2 m/yr where-
as the maximum balance measured on the
eastern part of Vatnajökull in the thirties was
about 7 m/yr (Thorarinsson 1978, pers. comm.).
In fact, as discussed earlier, the exact year of
the “1887” eruption may be held in some doubt,
for there were references to “eruptions in Vatna-
jökull” both in 1887 and 1889. Using the latter
date, that of last mention of eruptive activity
in Vatnajökull prior of the Grímsvötn erup-
tion of 1892, gives Xi = 1.47 and X2 = 3.23 m/yr,
or in terms of snow 3.3 m/yr for 1883—89, and
7.4 m/yr for 1889—92. For the entire period
1883—92 the mean annual balance is 2.64 m/yr
ice (5.94 m/yr snow). In view of this it seems
inevitable that the Bb 196 tephra was deposited
not earlier than 1889.
The largest snow drifts observed on Vatna-
jökull may form of the order of half a meter’s
relative relief (Thorarinsson 1978, pers. comm.).
Reduced to ice this represents about 22 cm —
an insignificant deviation compared to other
variations.
In conclusion, the X—t profile supports the
age-depth model of the Bárdarbunga core pre-
sented liere. The deviations in Xj from the
24 JÖKULL 27. ÁR
mean seem in most cases to represent real varia-
tion in mean annual balance, which may prove
to be correlatable to climate (in preparation).
ACKNOWLEDGEMENTS
The work described here is a part of a large
cooperative venture on retreaving and inter-
preting the ice core from Bárdarbunga. The
chief instigator of that enterprise was Prof.
Bragi Arnason, whose idea it originally was to
employ tephrochronology to date the core, and
to involve the author in the project. In inter-
preting the tephra layers the guidance of Prof.
Sigurdur Thorarinsson has at all stages been
invaluable; Gudrún Larsen took active part in
the initial work, and did much of the descrip-
tive work on the samples. The author is
grateful to Helgi Björnsson for his interest in
the project, and for many fruitful discussions
on ice dynamics and glaciology, that aided the
modelling of the profile. In particular, Helgi
Björnsson drew my attention to eqn. (a), p. 21
in the text, and Sven Sigurdsson suggested the
moclifications discussed in eqns. (b)—(e). Jakob
Yngvason suggested the analysis in p. 11 of the
maximum grain size distribution in terms of the
laws of Stokes ancl Newton. Þorgeir Helgason
did most of the microprobe analyses.
REFERENCES
Árnason, Bragi, 1976: Groundwater Systems in
Iceland. Soc. Sci. Islandica XLII, Reykja-
vík. 236 pp.
Arnason, Bragi, H. Björnsson ir P. Theodórs-
son, 1974: Mechanical drill for deep coring
in temperate ice. J. Glaciol. 13: 133—139.
Áskelsson, Jóhannes, 1936: On the last eruption
in Vatnajökull. Soc. Sci. Islandica RIT 18.
Barth, T. P. W., 1937: Volcanic ash from Vatna-
jökull. Norsk Geol. Tidsskr. 17: 31-38.
Benjamínsson, Jón, 1975: Oskulag „a“. [The
tephra layer “a”]. Unpublished B.S. thesis,
University of Iceland, 38 pp. (in Icelandic).
Björnsson, Axel, K. Sœmundsson, P. Einarsson,
E. Tryggvason ir K. Grönvold, 1977: Cur-