Jökull - 01.01.2020, Blaðsíða 45

Vestergaard et al.
A, B and C are adjustable parameters, though A is as-
sumed to be a constant (see Giordano et al. (2008) for
arguments therein), which implies that the composi-
tional controls are merely on B and C. The melt tem-
perature is estimated from the relationship between
melting temperature and MgO content based on the
experimental work on evolved ferrobasalt by Thy et
al. (2006). Dissolved volatiles like water play a great
role in influencing melt viscosity. Since the sam-
pled whole rock are partly or fully degassed, we esti-
mate the primary, pre-eruptive water content from the
K2O-H2O relationship similar to Lucic et al. (2016).
It has been shown that H2O and highly incompati-
ble elements, like K2O, correlate positively in pri-
mary, undegassed melt inclusion of primocryst min-
erals of Hekla eruptions (Portnyagin et al., 2012; Lu-
cic et al., 2016). The study of Lucic et al. (2016)
focussed on the volatile compositions of four Hekla
eruptions: H3, 1104, 1845 and 1991, which represents
a range in repose periods, compositions and volumes
of erupted material. They found that H2O/K2O ra-
tio of the melt inclusions varied, and generally dis-
played lower H2O/K2O ratio of < 2 as opposed to
what have been suggested by earlier studies (e.g. Port-
nyagin et al., 2012). This suggested loss of water
through degassing. The H2O content representative
of degassed magma is measured from glass rims on
separated feldspars from basaltic andesite samples of
the 2000 Hekla eruption by Höskuldsson et al. (2007).
RESULTS
The 1766–68 eruption
The following account is based on the eyewitness de-
scription of Finnsen (1767) and Thórarinsson (1967)
together with this study’s observed lava morphologies
and vent structures on the Hekla ridge in the aerial
photographs from 1945–46. In the morning of April
5, 1766 Hekla awoke with rumbles and earth tremors
and spew "fire, pumice and stones" from the volcano
(vent opening not specified). Descriptions of the first
days of the eruption focus mainly on the tephra fall
and on fluctuations of explosiveness. Two craters
were active from the initial phase of the eruption (de-
noted April 5, 1766–?): the middle crater (more to the
SW) on the summit (Figure 3b pink outline) and the
crater on the SW shoulder of the Hekla ridge called
Bjallagígur that is surrounded by a horseshoe shaped
scoria ridge called Höskuldsbjalli (Figure 3b red out-
line). There are vague descriptions of a third crater
also active from the initial phase of the eruption, how-
ever, this crater has more likely developed into the
ca. 6 km long fissure during the initial phase (Figure
3b purple dashed outline). All three craters were main
lava sources for the lavas flowing S-SW (Figure 3a
orange outline and yellow dashed outline). However,
the two first craters (Figure 3b pink outline and pur-
ple dashed outline) were mainly active during the first
days of the eruption and Bjallagígur was active dur-
ing the whole of the eruption. Mentions of at least
seven craters near the main three craters at the mid to
the SW shoulder of the ridge are given. We interpret
these observations as if they were part of the fissure
opening (Figure 3b purple dashed outline) as we could
not detect them on the aerial orthophotos. Moreover,
outpouring of lava has likely drowned some of the fis-
sure and later modifications from the 1845–46 erup-
tion as well as subsequent erosion has changed the
summit ridge. The first description of lava outpouring
is on April 9. Lava was seen flowing towards SSW
stretching as far as the north end of Vatnafjöll with a
length of about 7 km (Figure 3a orange outline). On
April 15–21, a great flow of lava emerged resulting
in a flow field extending to the SW (Figure 3a yellow
3. mynd. – Kort sem sýnir líklegt hraunflæði í gosinu 1766–68, byggt á sögulegum heimildum. Heilar línur sýna
útbreiðslu hraunflákanna, byggða á nákvæmum lýsingum. Brotalínur tákna áætlaða útbreiðslu hraunflákanna.
Bakgrunnur kortsins er sá sami og á 1. mynd. B) Nærmynd af Hekluhryggnum sýnir að minnsta kosti fimm
gígar voru virkir á árunum 1766–68. Loftmyndin er frá 1945–46, þá hafði Hekluhryggurinn þegar breyst eftir
gosin 1766–68 og 1845–46, og vegna rofs sem varð í kjölfar þeirra. Heilar línur tákna gíga sem nokkuð öruggt
er að hafi verið virkir á þessum tíma, og brotalínur ógreinilegri gíga á loftmyndinni. Túlkun þessi er eingöngu
byggð á rituðum heimildum.
42 JÖKULL No. 70, 2020