Jökull - 01.01.2013, Page 45

Stratigraphy, 40Ar–39Ar dating and erosional history of Svínafell, SE–Iceland
younger than the host lavas but older than both forma-
tions SV10 (i.e., the Svínafell sediments) and SV11
above. Dyke D2, that is reversely magnetized based
on measurements with a handheld magnetometer, was
sampled and dated by 40Ar-39Ar incremental heating
methods.
We prepared whole rocks for age determinations
by crushing, sieving to obtain the 0.1–0.5 mm size
fraction, ultrasonic washing in distilled water, dry-
ing and hand-picking under a binocular microscope
to obtain the freshest, phenocryst-free samples possi-
ble. From three normal polarity lavas (units SV17 and
SV23) lying above the Svínafell sediments, we loaded
approximately 500 mg of each sample in quartz vials,
evacuated and sealed them, then irradiated these for
6 hr near the central core of the 1 MW Oregon
State University TRIGA reactor, along with neu-
tron flux gradient monitor FCT-3 biotite (28.03 Ma;
Renne et al., 1998). Samples and monitors were an-
alyzed in single-step, total fusion experiments using
an AEI MS-10 mass spectrometer. All gases were ex-
tracted by radio-frequency heating of samples in Mo-
crucibles under vacuum, followed by removal of ac-
tive gases via hot metal getters. We made corrections
for mass discrimination based on frequent measure-
ments of atmospheric Ar from an on-line reservoir,
and for interfering isotopes produced during irradia-
tion (Wijbrans et al., 1995).
From reverse polarity dyke D2 cutting the Skjól-
gil lavas lying below the Svínafell sediments, we sep-
arated feldspar and, after identical irradiation condi-
tions, incrementally heated the sample in 6 temper-
ature steps from 500◦C to fusion using a 10W CO2
laser. Step heating gas compositions were analyzed
using a MAP 215/50 mass spectrometer.
Three samples from lithologic units SV17 and
SV23 in group S4, lying some 300 m above the Svína-
fell sediments, produced consistent and stratigraphi-
cally acceptable age determinations. The measured
ages (reported in Table 2) are not statistically distin-
guishable (at the 2 s.d. level), so we have calculated
a mean, weighted by inverse variance, which is 698
(±54) ka. Other samples, from lava flows below the
Svínafell sediments, produced ages that were unrea-
sonably old (up to 10 Ma) and imprecise, which we at-
tribute to alteration, with addition of excess 40Ar, pos-
sibly trapped with hydrothermal quartz (Seidemann,
1988). However, dating of dyke D2 that cuts the
Skjólgil lavas and terminates at the uppermost lava
(unit LA), gave a plateau age of 1.67 (±0.15) Ma.
Table 2. 40Ar–39Ar total fusion and incremental heating ages for Svínafell strata. – Niðurstöður Ar–Ar aldurs-
greininga fyrir Svínafell.
Sample no. coordinates stratigraphic age ± 2s (Ma) elevation, m a.s.l. comment
height (m) total fusion plateau
20–9–93–8 N63◦58.94599 1038 0.650 ± 0.184 628–765 Subglacially erupted
W16◦49.79781 volcanic unit (SV23).
20–9–93–7 N63◦58.89426 880 0.724 ± 0.066 598–613 Tholeiite to basalt andesite
W16◦49.83128 lava flow (SV17).
20–9–93–6 N63◦58.83366 870 0.643 ± 0.110 586–598 Tholeiite lava flow (SV17).
W16◦49.91939
D2 N63◦58.88486 253 1.67 ± 0.15 240 Reversely magnetized dyke D2
W16◦50.60318 cuts lavas of SV9 in Skjólgil.
Strike N18◦E, 0.7 m thick.
HV N64◦0.52176 1679 0.215 ± 0.012 240–345 Pl–phyric lava formation
W16◦52.47622 HM36 in Hafrafell.
K–Ar age; 0.54 % K
Ages calculated with the following decay and abundance constants: λε = 0.580 x 10−10 yr−1; λ = 5.530 x 10−10 yr−1;
40K/K = 1.17 x 10−4.
JÖKULL No. 63, 2013 45