Jökull - 01.01.2014, Qupperneq 57

Magnetostratigraphy, K-AR dating and erosion history of Hafrafell, SE–Iceland
glacial intervals, may reflect that Hafrafell was cen-
trally located at this time with regard to an ice sheet.
In this case it may be that erosion in Hafrafell out-
weighed accumulation that lead to greater erosion and
excess relief there during the Gauss chron. The greater
erosion in Hafrafell may even have obliterated the ev-
idence of at least 3 Gauss glacials that are present in
the Fljótsdalur area relatively close by.
The Hafrafell dike swarm, age and possible source.
The Hafrafell stratigraphy and lava-dike relations do
provide some information on the dike swarm age and
source. Thus, we find the most likely source to be
an old volcano that was active some 2 Ma and had
a location near the Hrútfjöll area or in that direction
with regard to Hafrafell. The dikes cut the top of
Neðri-Menn, a lava sequence of Matuyama age. Be-
low Neðri Menn, at 390 m a.s.l., we dated a lava flow
at 2.35 Ma that therefore must be maximum age for
the dike swarm. The swarm does not cut through the
"Hafrafell valley strata", that are slightly older than
Olduvai or 1.85 Ma. Most probably, the swarm was
active at about 2 Ma or in the interval 1.85 to 2.35 Ma.
The swarm trends into the Hrútfjöll lower sequence
and the Hafrafell strata show "southward" building of
lenses. The source for both the Hafrafell strata and
dike swarm could therefore be in the general Hrútfjöll
area.
SUMMARY AND CONCLUSIONS
The lowest strata in Hafrafell are lava flows that we
correlate with the base of the Gilbert chron (C2Ar or
3.596–4.187 Ma), erupted at about 4 Ma. Upwards the
lavas accumulated slowly during the Gilbert (C2Ar;
155-m-thick) and Gauss (C2An; 234-m-thick) mag-
netic chrons or during the interval 4.187–2.581 Ma.
Dating and stratigraphic correlation suggests that
during Gauss time strata accumulated only during
early Gauss (C2An.3n or 3.330–3.596 Ma). There-
fore, slow accumulation rates or more likely inten-
sive erosion lasted through the onset of Matuyama
chron (2.581–0.781 Ma) when a 739-m-thick se-
quence formed, predominantly of lavas.
We attribute slower accumulation rates in the
Hafrafell lower sequence in part to its rift flank lo-
cation, to the east of the axial rift zone. The Hafrafell
lower strata accumulated through forward stacking of
lenses that were added toward south or distally, as op-
posed to centrally constructed crust in the accreting
rift zones. Twelve erosion surfaces, HR1 to HR12, are
identified, based on tillite occurrences, subglacially
erupted pillow basalts and glacially striated lava sur-
faces.
We divide the erosion and landscape evolution of
Hafrafell into 6 stages. During the first two stages
(late Neogene) volcanism was continuous but accu-
mulation rates were slow and topographic relief was
less than 100 m. During stage 3, i.e. lower Matuyama
time, lava production increased and accumulation ex-
ceeded erosion. In stage 4, during upper Matuyama
time, the Hafrafell valley formed and a valley net-
work may have characterized the landscape. Stage 5,
in upper Matuyama time, marked the filling up of the
Hafrafell valley with lava flows and subglacial vol-
canics and evening out the established valleys. Dur-
ing stage 6, or the Brunhes chron, major subglacial
volcanism took place with continuous erosion within
∼2-km-deep valley network.
Acknowledgements
The people of Freysnes and Svínafell in Öræfi
are thanked for continued support over the years.
We gratefully acknowledge the generous use of Leó
Kristjánsson’s laboratory, Institute of Earth Science,
University of Iceland, for paleomagnetic measure-
ments. Friends of Vatnajökull, a non-profit asso-
ciation of the Vatnajökull national park, is thanked
for financial assistance to J.H. We thank Morten Ri-
ishus for many thought provoking comments on the
manuscript. Critical review by David McGarvie on
the manuscript led to significant improvements. An
early part of this work was supported to J.H. by the
Iceland Science Fund.
ÁGRIP
Jökulrof umhverfis eldstöðina í Öræfajökli, SA-landi,
hefur grafið yfir 2 km djúpa dali. Þar, í Hafrafelli,
bendir samsetning um 2.8 km þykks jarðlagasniðs til
þess að landslag hafi þróast frá frekar flötu landi til
mishæða og að lokum dalakerfa. Með kortlagningu
JÖKULL No. 64, 2014 57