Jökull - 01.01.2020, Síða 41

Vestergaard et al.
arinsson, 1967; Larsen et al., 2013; Pedersen et al.,
2018b). In Holocene times, Hekla has mainly pro-
duced mixed eruptions which encompasses both ex-
plosive and effusive activity (Thordarson and Larsen,
2007; Thordarson and Höskuldsson, 2008; Pedersen
et al., 2018b). The eruptions in 1947–48, 1845–46 and
1766–68 were no exceptions (Figure 1). All three be-
haved rather similarly, exhibiting a highly explosive
subplinian to plinian initial phase (VEI 4) with sig-
nificant tephra fallout (Houghton et al., 2013; Janebo
et al., 2016a; Gudnason et al., 2018). The follow-
ing phases were: fire fountaining, strombolian activity
ending with an effusive phase generating large lava-
flow fields (Thórarinsson, 1967, 1976; Larsen et al.,
1999; Thordarson and Larsen, 2007; Thordarson and
Höskuldsson, 2008; Pedersen et al., 2018b). The
1947–48 and 1766–68 eruptions also showed exam-
ples of renewed, violent explosivity during the effu-
sive phase, but these events never increased to the
same force as the initial phase (Thórarinsson, 1967,
1976).
Lava-flow morphology
Lava is a common and persistent threat to settle-
ments and understanding its properties and emplace-
ment mechanisms is key for hazard assessment and
for predicting lava-flow behaviour and development
(Soule et al., 2004; Takagi and Huppert, 2010; Mon-
talvo, 2013). This includes the estimation of lava-flow
thickness, volume and emplacement style and history
(e.g. Wadge et al., 1975; Stevens et al., 1999; Har-
ris et al., 2000; Poland, 2014; Albino et al., 2015;
Kubanek et al., 2017; Pedersen et al., 2018a). The
development and emplacement of lava-flows hinge on
parameters such as rheology, effusion rate and erup-
tion duration, temperature, topography and surface
slopes and also total volume of lava extruded (Walker,
1973; Pinkerton and Wilson, 1994; Parfitt and Wil-
son, 2008). Furthermore, the rheology of the lava
evolves during emplacement, because of changes in
melt composition, oxygen fugacity and temperature of
the magma resulting from gas loss, cooling and crys-
tallisation (Hulme, 1974; Fink, 1980; Gregg and Fink,
2000; Kilburn, 2004; Kolzenburg et al., 2018). Hekla
typically generates ‘a‘ā lavas (Thórarinsson and Sig-
valdason, 1972; Grönvold et al., 1983; Höskuldsson et
al., 2007; Thordarson and Höskuldsson, 2008) which
are flows with autobrecciated exteriors and a coher-
ent liquid interior during emplacement, referred to as
the core. The autobreccia comprises irregular clink-
ers that have rough, sharp surfaces, and are derived
from the flow core (Macdonald, 1953). Common mor-
phologies that are mentioned in this study, are sheet-
flows which formed in a single surge of lava that
is not bounded by levées (banks of solidified lava),
and channelised flows of lava bounded between lev-
ées (Hulme, 1974; Peterson and Tilling, 1980; Row-
land and Walker, 1988; Hon et al., 1994; Harris et
al., 2009; Harris and Rowland, 2015). Morphologies
related to lava inflation include flat-surfaced plateaus
(formed like tumuli) called lava rise, collapsed de-
pressions called lava-rise pits, and marginal fractures
called lava-inflation clefts (Walker, 1991). They form
by the injection of lava beneath the surface layer
which leads to simple uplift of the surface without any
horizontal compression (Walker, 1991). The lava-rise
pits and lava-inflation clefts describe, respectively,
lava surfaces that failed to be uplifted, and actively in-
flating interior of the lava-flow that detaches from the
stagnated margin (Walker, 1991; Hon et al., 1994).
Inflation structures are commonly found in pāhoehoe
flows (e.g. Macdonald, 1953; Walker, 1991, Hon et
al., 1994), but inflation structures and lava tubes also
occur in ‘a‘ā flows (e.g. Calvari and Pinkerton, 1998,
1999; Pedersen et al., 2017). Lava tubes are insu-
lated conduits of still-molten lava beneath a cooler
(ultimately solidified) surface layer (Peterson et al.,
1994). The boundary of the inflated lava-flow may
breach, thus allowing lava breakouts and thereby new
lava lobes. Breakouts do not only occur due to rupture
of the cooled skin at weak points, but can also oc-
cur due to effusion rate fluctuations (Thordarson and
Self, 1998; Rowland and Harris, 2015; Pedersen et
al., 2017). The emplacement of breakouts may vary
depending on viscosity and effusion rate.
DATA AND METHODS
Remote sensing data and bulk volume estimates
The remote sensing data consists of orthophotos and
digital elevation models (DEMs) from 1945–46, 1960
and 2015 (Table 1). The remotely-sensed images
38 JÖKULL No. 70, 2020