Jökull - 01.12.1991, Qupperneq 57

may be at the bottom of the normal zone in the
Mosfellssveit area, but this direction does not fit the
pole path found by Sigurgeirsson (1957 ; Kristjáns-
son and Sigurgeirsson 1992) for the R3-N3 transition
in lavas in Hvalfjörður. It is therefore possible e.g.
that UK 7 belongs to a geomagnetic excursion within
N3.
The lowermost zone of normal-polarity lavas is
followed by the relatively thick reverse lava series in
the remainder of GB and UE as well as in UF.
Continuous outcrops in this series end at the foot of
Úlfarsfell due to scree, but we consider that profile
UH covers the stratigraphic gap below the cliff pro-
file UO in Úlfarsfell.
Most of the upper part (UB) of Úlfarsfell is of nor-
mal magnetization, which may correspond to the
Olduvai subchron or possibly the elusive Gilsá sub-
chron of approx. 1.6 M.y. age. We correlate this zone
with the youngest normal zone in Esja, occurring in
intrusions and hyaloclastites in the Svínaskarð
(Móskarðshnúkar) profile SC of Kristjánsson et al.
(1980) and in other late products of the Stardalur cen-
tral volcano (Friðleifsson 1973, 1985).
The very top of Úlfarsfell (Stórihnúkur peak), as
well as profile UD, is reversely magnetized. The pat-
tem of easterly declinations of remanence in UC 2 -
8 is remarkably similar to that in UD 1 - 5 (or 7) ,
and the inference that they are contemporaneous
agrees well with the geological strike. In a more
detailed survey of the Úlfarsfell area, it may become
possible to correlate other small groups of lavas, or
even individual lava units, over some distance with
the aid of their paleomagnetic directions and other
characteristics, for example between the profiles UL
and UA/UB.
Secular variation. Despite generally close within-unit
directional agreement (95% confidence angles < 10°
in Table 1), the lavas described here are much less
satisfactory material for paleomagnetic secular varia-
tion studies than are other lava sequences described
so far from Iceland. The reasons for this are mostly
geological:
a) Tectonic tilt is quite large, with a correspond-
ingly large uncertainty.
b) Outcrops are not very good, possibly causing
repeated sampling of the same flow in some
instances, such as in the flat-lying Gufunes profile
GB.
c) Undetected faults, and lack of traceable hori-
zons (such as extensive tuffs or petrographically dis-
tinct lavas) may have caused duplicate sampling of a
few lavas in adjacent profiles or even within the same
profile: thus, the similar remanence directions in
lavas GB 8, 12/13, 21 and 32/33 indicate that there
are repeated segments (see Fig. 3). Similarly, the sites
UB 13,14,15 and UB 15A,16,17 may actually repre-
sent the same three (or two) flows.
d) Relatively close grouping of directions in two
to several successive lavas is common; it does not
appear to be due to secondary regional remagnetiza-
tion. Examples include UC 2-8 as pointed out above,
UF 13-16, UD 10-20 and even UH 3-14. This serial
correlation indicates that each such group of lavas
was erupted in a short episode compared to secular
variation time scales.
For these reasons, we feel it is not appropriate at
this stage to state any mean paleomagnetic directions
or between-lava angular statistics for our collection.
CONCLUSIONS AND DISCUSSION
We have mapped a composite 800 m thick section
along a stretch of approximately 8 km from Gufunes
to Hafrahlíð in the Mosfellssveit area. The section is
mostly of reverse magnetization and most probably
of Lower Matuyama age, somewhere in the range 1.6
to 2.2 million years. A normally magnetized zone of
a few tens of m thickness overlies hyaloclastite at the
base of the profile, and another normally magnetized
zone of over 100 m thickness is found in the upper
part of Úlfarsfell. We correlate (Fig. 5) the lower nor-
mal zone with Einarsson's N3 zone in Kistufell and
other locations farther east, and the upper zone with
the final phase of the Stardalur volcano. These may
correspond in time respectively to the Reunion and
Olduvai subchrons, but additional radiometric dates
in the area are needed to test this suggestion.
The publication of the present pilot study is
intended to stimulate further interest in the mapping
of stratigraphy and structure in the Kjós-Mosfells-
sveit-Reykjavík region. This mapping can be done
JÖKULL, No. 41, 1991 55