Jökull - 01.01.2017, Síða 27

Svavarsdóttir et al.
Figure 3. cont. (d) The Kvía-
hraun lava near Kiðagil.
(e) Holocene lava flows of
Dyngjuháls to the southeast
of Gæsavötn. (f) The pahoe-
hoe lavas of Trölladyngja. –
Ljósmyndir af hraunum norðan
Vatnajökuls.
Sciences, University of Iceland. In the case of highly
plagioclase phyric samples, plagioclase macrocrysts
were separated from the groundmass by handpicking
under a binocular microscope. Preferentially, only
fresh samples were used, excluding samples show-
ing evidence of secondary fillings, organic material or
signs of oxidations. Details regarding preparation of
the samples prior to ICP-OES analyses can be found
in Svavarsdóttir (2017). Assessment of analytical un-
certainty associated with these analyses was made by
multiple analyses of the rock standard K1919 (a Ki-
lauea basaltic standard collected from the same flow
as BHVO-1 and provided by Ken Rubin at the Uni-
versity of Hawaii).
Multi-Collector Inductively Coupled Plasma Mass
Spectrometer (MC-ICP-MS)
Methods adopted for isotopic analyses are in general
similar, but improved, to those described by Halldórs-
son et al. (2008) and Sigmarsson and Halldórsson
(2015) with further details given in Svavarsdóttir
(2017). Sr-, Nd-, Hf-, and Pb-isotopic ratios were
determined on powdered rock samples with a Nu
Plasma MC-ICP-MS at the Institute of Earth Sci-
ences, University of Iceland. Strontium isotope ra-
tios were normalized to NIST SRM-987 87Sr/86Sr ra-
tio of 0.710248 after correction for Rb and Kr inter-
ference following Konter and Strom (2014). Sim-
ilarly, 143Nd/144Nd was normalized internally to a
22 JÖKULL No. 67, 2017