Jökull - 01.01.2015, Page 18

Sigmarsson and Halldórsson
term volcanic systems in the Eastern Volcanic Zone.
This latter approach has thus far not given a unique
answer to the origin of several lava flow fields at the
edge of Dyngjujökull, an outlet glacier of Vatnajökull.
Further complications arise due to the fact that several
volcanic systems at play in this region are partially
covered by ice (Figure 1).
The complicated tectonic structure of the region
N of Vatnajökull glacier, more or less directly above
the Iceland mantle plume, was highlighted during the
2014–2015 rifting event. This event was characterized
by two-week seismicity progressing over 40 km and
up to 2 m spreading as monitored by Icelandic Mete-
orological Office and Institute of Earth Sciences, Uni-
versity of Iceland and collaborators (e.g. Sigmunds-
son et al., 2015). The northward propagation of the
rifting from the Bárðarbunga central volcano along
the northern branch of the associated fissure swarm,
caused seismicity in the neighbouring volcanic sys-
tem of Askja several days before the fissure eruption
at Holuhraun. This tectonic evolution caused uncer-
tainty about where an eruption would occur, with an
explosive eruption at Askja as a possible scenario.
Fortunately, the magma came up through existing
craters at the Holuhraun lava field, which had previ-
ously been associated with the southern fissure swarm
of Askja (e.g. Einarsson and Sæmundsson, 1987),
a proposition questioned by Hartley and Thordarson
(2013). The question of illegitimate magma trans-
ferred from one volcano to another, such as observed
at Galapagos and between Kilauea and Mauna Loa
(e.g. Rhodes et al., 1988; Geist et al., 1998), and in
this case from Bárðarbunga to Askja, is thus a pos-
sibility. Alternatively, as discussed by Hartley and
Thordarson (2013) and developed below, the delimi-
tation of the Bárðarbunga and Askja volcanic systems
needs revision.
Radiogenic isotopes have proven useful in trac-
ing the provenance of magma in Iceland. For exam-
ple, the Sr- and Nd-isotope ratios reflect fractionations
from their parental elements (Rb and Sm) and the time
elapsed since the fractionation. Their respective half-
lives, far exceed the young age of Icelandic crust (i.e.
109 years vs. >106 years) and, therefore, limited
crustal contamination will not affect the Sr- and Nd-
isotope ratios in Icelandic basalts, which are princi-
pally mantle-derived. Mantle heterogeneity, and melts
thereof, is thus directly sampled by basaltic volcan-
ism. Nevertheless a given volcanic system (i.e. Jak-
obsson, 1979) appears to deliver magmas to the sur-
face with relatively uniform Sr- and Nd-isotope ratios
(e.g. Sigmarsson et al., 1992; Furman et al., 1995;
Kokfelt et al., 2009; Chekol et al., 2011), an observa-
tion that may be helpful with delineating volcanic sys-
tems. An example where this approach was adopted
on Icelandic volcanics, is the eruption at Gjálp in 1996
that produced magma originating at Grímsvötn vol-
cano, despite a precursory seismicity at Bárðarbunga
central volcano (Sigmarsson et al., 2000 and refer-
ences therein). In this paper, the aim is to further test
this "fingerprinting" method on several recent craters
and Holocene lavas from the complex intersection of
Askja and Bárðarbunga fissure swarms in order to as-
sign various eruptive units to their parental volcanic
system.
GEOLOGICAL SETTING
The ice-covered central volcano of Bárðarbunga is lo-
cated more or less directly above the presumed cen-
tre of the Iceland mantle plume. Several other cen-
tral volcanoes are also found in the vicinity of Bárð-
arbunga (Figure 1) as identified on the most recent
geological maps of active volcanic systems in this
region that combine tectonic observations, including
recent seismicity distribution and subglacial topog-
raphy (Einarsson and Sæmundsson, 1987; Björns-
son and Einarsson, 1990; Jóhannesson and Sæmunds-
son, 1998; Sigurgeirsson et al., 2015). This region
also marks the triple junction between the Northern
Rift Zone (NRZ), the Eastern Rift Zone (ERZ) and
the Mid-Iceland Volcanic Belt (MIVB), where the
strike of the fissures changes direction from SW-NE
to the South to progressively more N-S direction in
the North. Associated with this change, appears to be
an overlap in erupted magma compositions, or inter-
fingering of fissure swarms belonging to two different
volcanic systems. This region is characterised by very
high magma production with, for example Bárðar-
bunga being one of the most productive volcanic sys-
tems in Iceland during the Holocene with estimated
18 JÖKULL No. 65, 2015