Jökull - 01.01.2020, Page 54

The 1845–46 and 1766–68 eruptions at Hekla volcano
after the new vent opening at March 18, 1767, and
the observed lava-flow morphology from the aerial
orthophotos confirms these descriptions (Figure 3a,
b). Therefore, the high-SiO2 lavas of Hringlanda-
hraun was emplaced late in the eruption. This dis-
covery challenges the conventional model of tapping
of a single, large, layered magma chamber with the
most evolved (silica-rich) magmas at the top. The
simplest explanation is that the NE end of the summit
ridge, at least in 1766–68, tapped a different source,
but this should be investigated further requiring more
detailed sampling of the different phases of the erup-
tions. However, based on our results the eruption of
the 1766–68 lava-flows are inconsistent with a sin-
gle, stratified magma chamber model from which the
erupted products become poorer in SiO2 with time.
CONCLUSIONS
The estimated bulk volume based on the planimetric
method for the 1766–68 Hekla eruption is 0.7 km3
and for the 1845–46 eruption is 0.4 km3. Based on
volume comparisons between the planimetric method
and digital elevation model (DEM) thickness maps for
the 1947–48 eruption, it is suggested that the plani-
metric method may underestimate the lava bulk vol-
ume by 40–60%. Hence, the true bulk volumes of
the 1766–68 and 1845–46 eruptions are more likely
1.0–1.2 km3 and 0.5–0.6 km3, respectively. We esti-
mate viscosities representative of pre-eruptive (aver-
age 1.3 wt% H2O) and degassed magmas (0.14 wt%
H2O). The undegassed melt viscosity ranges from
1.8×102–3.1×102 Pa s (1060◦C) for the 1766–68
lava-flows, and 1.6×102–2.8×102 Pa s (1060◦C) for
the 1845–46 lava-flows, and the degassed melt viscos-
ity ranges from 1.4×103–3.6×103 Pa s (1060◦C) for
the 1766–68 lava-flows, and 1.1×103–2.9×103 Pa s
(1060◦C) for the 1845–46 lava-flows. This presents
one order of magnitude difference between pre-
eruptive and degassed magma, and confirms that the
pre-eruptive magma is more fluid than the degassed
magma. The emplacement time-lines of the lavas
combined with the silica content of collected sam-
ples show that SiO2 decreased from ca. 58 to 54 wt%
during eruptions in 1845–46 and 1947–48. In con-
trast, our results from the 1766–68 eruption suggest
that SiO2 fluctuated increasing from ca. 55 to 58 wt%
during the emplacement of Hringlandahraun, reveal-
ing that not all Hekla eruptions necessarily follow a
decline in SiO2 during the eruption. This implies
that the model assuming magma being tapped succes-
sively from the top of a stratified magma chamber may
not be applicable to all Hekla eruptions.
ACKNOWLEDGEMENTS
We would like to acknowledge the Icelandic Research
fund, Grant of Excellence No. 152266-052 (Project
EMMIRS) for supporting the field trip and access
to the remote sensing data. This research was also
funded by the Danish National Research Foundation
Niels Bohr Professorship grant 26-123/8. Thanks to
Landgræðsla ríksins for their help during fieldwork.
The first author also acknowledges Selskabet Arktisk
Forskning og Teknologi (SAFT) for funding of trav-
els connected to this project. Extended thanks go to
editor Bryndís Brandsdóttir and two anonymous re-
viewers for their thorough and constructive reviews.
Lastly, thanks to Páll Einarsson for help with the Ice-
landic translation of the abstract and for proof-reading
the manuscript. Also thanks to Stefán Már Melstað
for help with the Icelandic abstract and to Ásta Rut
Hjartardóttir for help with the Icelandic translation of
the figure- and table texts.
Ágrip
Rúmmál hrauna sem runnu í Heklugosunum 1845–
1846 og 1766–1768 er endurmetið út frá nýjum fjar-
könnunargögnum, sögulegum heimildum, bergfræði
og seigju hraunanna, sem reikna má út frá efnasam-
setningu þeirra. Ef reiknað er út frá flatarmáli hraun-
anna og ætlaðri þykkt þeirra fæst rúmmál 0,4 km3 og
0,7 km3 fyrir gosin tvö. Ef sömu aðferð er beitt á
gosið 1947 kemur í ljós að hún vanmetur rúmmálið
um 40–60%. Því eru rúmmál eldgosanna 1845–46
og 1766–68 áætlað 0,5–0,6 km3 og 1,0–1,2 km3. Við
áætlum seigju kvikunnar út frá efnasamsetningu og
gasinnihaldi. Ætla má að seigja kvikunnar fyrir gosið
1766–1768 hafi verið 2,5 x 102 Pa s en 2,5 x 103 Pa s
eftir að gasið rauk úr henni. Fyrir gosið 1845–1846
er seigja kvikunnar metin 2,2 x 102 Pa s en eftir að
gasið rauk úr er hún metin 1,9 x 103 Pa s. Seigja
kvikunnar vex nær tífalt þegar hún gýs og losar sig
JÖKULL No. 70, 2020 51