Jökull - 01.01.2013, Side 8

L. Kristjánsson
cation-deficient magnetite had exsolved from larger
grains of titanomagnetite during cooling
d) The ambient geomagnetic field being strong during
this eruption episode.
RESEARCH IN 1990–2008
Helgason et al. (1990) and Steinþórsson et al. (1992)
presented results of Mössbauer spectroscopy on re-
spectively 8 and 20 samples from 15–175 m depth
in the Stardalur drill hole as well as on 10 sam-
ples of typical Icelandic basalts, mostly from out-
side the Late Quaternary areas. Their observations
confirmed that the strongly magnetized lava sequence
in Stardalur contains exsolved magnetite in a very
pure state. Helgason et al. (1990) state that practi-
cally no maghemite is present in their Stardalur core
samples, while Steinþórsson et al. (1992) found that
most of those lava samples from elsewhere in Iceland
which they studied, contain a good deal of maghemite.
The latter concluded that the maghemite is formed
from magnetite by secondary hydrothermal alteration.
In both papers the exsolution of pure magnetite in
Stardalur is also ascribed to such alteration, but no
explanation is offered for the absence of maghemite
there. Helgason et al. (1990) further suggest that
the strong remanence has resulted from the alteration.
However, we find it difficult to visualize how sec-
ondary processes could produce the uniform steep re-
manence inclinations observed by Búason (1971).
Later in the 1990s, magnetic field measurements
by satellites orbiting Mars revealed the presence of
strongly magnetic near-surface materials (Acuña et
al., 1999); a minimum value of Mt∼50 Am−1 was in-
ferred. This, as well as reports from Mars landers on
magnetic properties of exposed formations in 2004,
created new interest in the magnetism and petrol-
ogy of the Stardalur core. Gunnlaugsson et al. (2004,
2006) concluded that a part of the magnetite in the
core (presumably the small cubic grains of Steinþórs-
son and Sigvaldason (1971)) is very pure and did ex-
solve from olivine at high temperature during cooling.
A similar conclusion was reached on the origin of un-
usually high NRM intensity (∼40 Am−1) in an olivine
tholeiite lava flow MM 2 from Eastern Iceland (Kristj-
ánsson and Guðmundsson, 2005). Observations by
Gunnlaugsson et al. (2006, 2008) on magnetic sepa-
rates indicated that the natural remanent magnetiza-
tion in small magnetite grains from olivine might be
several times higher per weight unit than in an equal
amount of magnetite in typical Icelandic lavas.
Vahle et al. (2007) measured various magnetic
properties of nine samples from the strongly magne-
tized part of the original Stardalur core, and made a
detailed study of the mineralogy of two of these (at
101.0 and 135.15 m depth) by electron microscopy
(their Figs. 4 and 5) and other techniques. Vahle
et al. concurred with the suggestion of Gunnlaugs-
son et al. (2004) that the small magnetite grains
may have been generated from olivine by exsolution.
They also demonstrated that a minor proportion of the
magnetite in the larger grains had been oxidized to-
wards maghemite composition in situ, as evidenced
by shrinkage cracks in some of these. In thermomag-
netic curves obtained in air, the magnetization of sam-
ples decayed progressively with time spent at high
temperature, while curves obtained in an argon at-
mosphere were approximately reversible. These au-
thors concluded from their various observations that
the maghemite and some of the magnetite may have
formed during secondary hydrothermal alteration at
250–350◦C. However, evidence from a 1920 m long
core recovered from the lava pile of Eastern Iceland
(Bleil et al., 1982) indicates that secondary formation
of magnetite accompanies the appearance of epidote
at 250◦C or more (see Pálmason, 2005, p. 72). In the
Stardalur core, mixed-layer clays and chlorite which
presumably form at 200–250◦C are seen (Steinþórs-
son and Sigvaldason, 1971; Vahle et al., 2007, p. 123)
but no epidote. Hall (1985) states that the formation
of secondary magnetite occurs in the vicinity of dikes
at burial depths exceeding 2100 m and becomes rela-
tively important below 2900 m. The lava pile above
Stardalur is not expected to have reached such thick-
ness (Friðleifsson, 1985).
A comprehensive total-field magnetic survey of
the Reykjavík area and offshore at 500 m altitude
a.s.l. (Jónsson and Kristjánsson, 2002) confirmed that
a localized negative central-volcano anomaly east of
the city noted by Sigurgeirsson (1970a) extends up
8 JÖKULL No. 63, 2013