Jökull - 01.12.1990, Qupperneq 109

Table IV. Analytical methods. — Efnagreiningaaðferðir.
Element
PH
Na
K
Ca
Mg
C1
S04
F
Si
HCO3
Analytical Method Relative error a Reference
Combination pH glass electrode High pressure liquid ion chromatography 3%
High pressure liquid ion chromatography 7%
Atomic absorption spectrophotometry 19% Skougstad et al. (1979)
Atomic absorption spectrophotometry 7% Skougstad etal. (1979)
High pressure liquid ion chromatography 29%
High pressure liquid ion chromatography 34%
Ion sensitive electrode Skougstad etal. (1979)
”Spectrophotometric, molybdate blue“ 20% Skougstad et al. (1979)
”Calculated, assuming pure water and air“ Plummer and
saturation at the measured pH and temperature. Busenberg (1982)
“The relative error for C1 and SO4 is the maximum one between two of up to 16 replicates.
1987-1988 precipitation layer can be studied by sam-
Ples 88-1.1, ”88-2.1“, ”88-3.0“, 88-4.1, 88-5.1, 88-
6-l, 88-7.1, 88-8.1, 88-9.1 (Table I, Fig. 2). The pH
°f what is left of the 1987-1988 layer, at the time of
sanrpling, ranges from 5.47 to 5.73. The pH is highest
ln eore 9 which is closest to the edge of the glacier but
Ihe pH reaches lowest value in core 7 at 1300 m.a.s.l.
altitude. At higher altitude it ranges from pH 5.50
,0 5.67. Chloride is always the most important anion
6ut of the cations the sodium ion is the predominant
°ne. This suggests a primary marine origin of the ions.
The concentration of dissolved sea-salts increases with
mcrease in altitude, contrary to what has been docu-
fiiented for precipitation in Iceland and other parts of
the world (Sigurðsson and Einarsson, 1988; Herron
and Langway, 1985). The concentration of chloride
'h the 1987-1988 layer is greatest near Grímsvötn,
6-9 ppm (samples 1.1 and ”2.1“, Table I and Fig. 2)
but decreases downslope towards west and is at its
minimum, 0.3 ppm Cl” in core 9 closest to the base
°f the glacier at 1120 m.a.s.l. (sample 9.1 in Table
1 and Fig. 2). The concentrations of dissolved silica
and fluorine are at or below the detection limits of the
analytical methods where tested (Table I).
Downcore changes in pH, Na+ and Cl“ are shown
in Fig. 4 for core 1 an(j jn pjg g for core 2 but down-
c°re concentrations for other chemical constituents are
shown in Table II for core 1 and Table III for core 2.
The pH is highest in the top part of core 1 (Fig. 4),
it decreases gradually down to about 140 cm depth
where the pH is equal to 5.4. Below this level down
to the bottom of the 1987-1988 precipitation layer the
pH ranges from 5.4 to 5.5. There is an abrupt shift to
a higher pH at the top of the 1986-1987 precipitation
layer (Fig. 4). The pH is always higher in the older
layer if the first two samples in the younger layer are
excluded. The chloride and sodium ion concentrations
vary with depth in similar manner as described for the
pH. The concentrations are relatively low at the top of
the core, they reach maximum at about 190 cm depth
and the ion concentrations in the 1986-1987 precipi-
tation layer are lower than in the 1987-1988 precipi-
tation layer above. Similar trends can be observed in
core 2 (Fig. 5). The ion concentrations reach maxi-
mum at 190 cm depth but that is the depth at which
the cores were below 0°C at the time of sampling.
All the cores shown in Figs. 4, 5, and 6 have un-
dergone partial melting. The core taken closest to the
edge of the glacier, core 9 (Fig. 2), have undergone
more melting than cores from higher levels. When
collected, cores 1 and 2 were frozen below 190 cm
depth. In other words, cores 1 and 2 have undergone
partial melting down to about 190 cm depth and the
lower parts of the cores represented by the 1986-1987
precipitation layer have undergone partial melting dur-
ing the summer 1987. Core 9 is partially melted down
JÖKULL, No. 40, 1990 105