Jökull - 01.01.2020, Page 39

Vestergaard et al.
2003; Gudnason et al., 2017; Janebo et al., 2018,
Einarsson, 2018; Barsotti et al., 2019). The effu-
sive eruption of Hekla lavas also represent a hazard
for the surrounding settlements and farmland, but has
remained understudied in particular the lavas erupted
before the 20th century (Jakobsson, 1979; Pedersen et
al., 2018b). There is therefore sparse information re-
garding lava-flow emplacement, lava morphology and
lava volumes. For example, the eruptions in 1845–46
and 1766–68 are some of the largest known eruptions
from Hekla since the settlement. They are comparable
in volumes to the 0.4 km3 eruption of Surtsey 1963–
67 (lava shield part) and to the recent 1.44 km3 2014–
15 eruption of Holuhraun, respectively (Thórarins-
son, 1967; Thordarson, 2000; Thordarson and Larsen,
2007; Janebo et al., 2016a,b; Pedersen et al., 2017,
2018a,b; Bonny et al., 2018; Gudnason et al., 2018).
Previous work has concluded that Hekla’s plumb-
ing system involves a single, zoned magma source
and that the most silica-rich magmas (rhyolite) are
tapped from the topmost layer during the initial explo-
sive phase (particularly the case for prehistoric erup-
tions) (Sigmarsson et al., 1992; Sverrisdottir, 2007).
During the subsequent effusive eruptions the SiO2 of
the erupted melts declines towards basaltic andesite
(ca. 54 wt%) (Sigmarsson et al., 1992; Thordarson
and Larsen, 2007). Following this hypothesis, a nega-
tive correlation of silica content and relative emplace-
ment age is expected during eruptions. The aim of
the present study is to extend the understanding of
the effusive activity at Hekla using remote sensing
data, petrology, geochemistry (with focus on whole
rock SiO2 values and their correlation with relative
lava emplacement age, and on estimating viscosity)
and historical sources of the 1845–46 and 1766–68
eruptions. The two eruptions are selected to improve
knowledge of large, effusive activity of Hekla. The
1947–48 eruption is also included as a benchmark
since its activity and eruptive products have already
been extensively recorded (e.g. Thórarinsson, 1976,
Pedersen et al., 2018a). In particular, the goals are
to (i) estimate the erupted volumes; (ii) qualitatively
outline the emplacement time-lines of the lava-flows;
(iii) discuss lava morphology and viscosity, and (iv)
discuss magma chamber dynamics of these eruptions.
BACKGROUND
The volcano Hekla
The volcano Hekla is located in the southern part
of Iceland at the intersection of the South Iceland
Seismic Zone (SISZ) and the Eastern Volcanic Zone
(EVZ) (Sæmundsson, 1978; Jakobsson, 1979) (Figure
1). The Hekla volcanic system consists of a NE-SW
trending fissure swarm and a central volcano rang-
ing from basalt to rhyolite in composition (Thórar-
insson, 1967; Jakobsson, 1979; Sigmarsson et al.,
1992; Sverrisdottir, 2007; Larsen et al. 2013; Tuller-
Ross et al., 2019). Today the central volcano forms
a 5–6 km elongated steep-sided ridge that reaches a
height of 1490 m a.s.l., superimposed on a basaltic
base and oriented parallel to the 60 km NE-SW fis-
sure swarm (Sæmundsson, 1978; Jakobsson, 1979).
The nature of the magma chamber beneath the central
volcano has been debated. Most authors argue that the
changing compositions of the tephras from rhyolite to
basaltic andesite demonstrate tapping of a chemically-
stratified magma chamber (Sigmarsson et al., 1992;
Sverrisdottir, 2007). Sigmarsson et al. (1992) initially
argued for a very large (20 km long, 5 km wide and
ca. 7 km deep) magma chamber with the top ca. 8 km
below the surface. Subsequently others have sug-
gested slightly modified architectures such as a dyke-
formed magma chamber (Sverrisdottir, 2007) or sep-
arate small, connected magma chambers undergo-
ing closed and/or open system geochemical evolution
(Chekol et al., 2011). Recent geophysical models
confirm that the magma chamber is indeed very deep,
probably at depths exceeding 10 km (Ofeigsson et al.,
2011; Geirsson et al., 2012; Sturkell et al., 2013).
Whereas the tephras range from rhyolite (ca. 75 wt%
SiO2) to andesite (ca. 57 wt% SiO2) (Sigmarsson et
al., 1992; Sverrisdottir, 2007; Janebo et al., 2018),
the compositional range of the lavas is more restricted
from andesite (ca. 58 wt% SiO2) to basaltic andesite
(ca. 54 wt% SiO2), which is the typical end compo-
sition (Thórarinsson, 1967; Sigmarsson et al., 1992;
Larsen et al., 1999; Sverrisdottir, 2007). All the major
eruptions occur in the main fissure along the summit
ridge, and historical sources describe how the ridge
sometimes split in two during eruptions; more com-
plex fissure opening have also been described (Thór-
36 JÖKULL No. 70, 2020