Jökull - 01.01.2013, Page 58

Da̧bski and Tittenbrun
and Thorarinsson (1937) and Thorarinsson (1943), as
well as lichenometrical investigation using four differ-
ent growth rates (Da̧bski, 2002). The oldest moraine
ridge found on the Hólmsárgarður profile, the moraine
I (Figure 1) according to Da̧bski (2002, 2007), is com-
plex and it is possible to distinguish a narrow and
a small ridge further east of the Hólmsá river, and
slightly more distal to the glacier front. The test site
Ia, used in this study, was located on this potentially
older ridge, and the moraine I is represented in this
study by site Ib. Consecutive numbers of test sites fol-
low previous numeration of moraines (Da̧bski, 2002,
2007), the youngest (site VII+) being situated on a
proximal slope of the VII moraine ridge (in its north-
ern part), on glacial till deposited in the first decade of
the 21th century.
Fláajökull has retreated over 1500 m since the end
of the 19th century. The retreat was interrupted by
numerous short advances or still-stands in years AD:
1907, 1913, 1920–1925, 1932, 1941, 1948, 1951–
1952, 1958, 1968–1969, 1992, and a relatively major
advance in AD 1980–1989 (Thorarinsson, 1943, Den-
ton, 1975, Sigurðsson, 1998, Da̧bski, 2002).
The fluted surface of a huge moraine ridge V and
examination of the Fláajökull retreat history testifies
that the core of the moraine had been created proba-
bly as a terminal moraine due to pre-LIA glacial ac-
tivity, and then it was overridden by the glacier which
deposited ground moraine on top of the ridge (Snorra-
son, 1984).
Linking the Fláajökull LIA maximum with its fur-
thest moraine ridges is difficult, because of possible
erosional censoring of earlier glacial deposits (Kirk-
bridge and Winkler, 2012). However, hypothetical
increase in micro-roughness, weathering rind thick-
ness and decrease in Schmidt hammer rebound val-
ues on moraine Ia in reference to moraine Ib, apart
from lichenometrical dating, lends support to a pre-
late 19th century age of the Fláajökull LIA maximum.
METHODS
The analyses were performed in AD 2011 on 90
basaltic boulders, 15 boulders in each of the 6 test
sites located on moraine of different age deposited
since the end of the LIA (Da̧bski, 2002, 2007). The
sites correspond to moraine ridges by Da̧bski (2002,
2007) with some modifications (Figure 1). All anal-
ysed boulders, apart from being the same petrography
(Tertiary grey basalts), had clear signs of glacier abra-
sion, indicative of transport in the subglacial traction
zone. We thus infer that the abraded surfaces were
subject to weathering only since their release from
glacier ice and that any former weathering rinds had
been eroded prior to deposition. Glacially-abraded
surfaces were facing the glacier (N, NW), with only
few exceptions of west-facing or horizontal surfaces.
The lengths of the smallest axis of the boulders were
at least 30 cm and weight exceeded 80 kg. Such large
sizes inhibited any boulder movements upon impact
of the Schmidt hammer, that could result in inconsis-
tency of R-values (Sumner and Nel, 2002).
After boulder selection, the following procedure
was employed. Firstly, any loosely lying rock parti-
cles (silt or single grains of fine sand) were removed
with a strong blow of air. Micro-roughness was mea-
sured in three smoothest lichen-free places found on
the abraded surface, giving the total of 45 readings
per test site, along profiles parallel to striation in or-
der to omit visible erosional features on the run of
the profilometer. We thus measured weathering rather
than erosional micro-relief, however, total exclusion
of micro-erosional features is impossible. Calibra-
tion of the profilometer was checked at every test
site using the reference roughness specimen (part of
a standard equipment of the Handysurf). Secondly,
the abraded surface of a boulder was subjected to
ten Schmidt hammer blows at least 1 cm apart from
each other and at least 6 cm away from any visible
irregularity on the rock surface according to the rec-
ommendations of Day and Goudie (1977). R-values
were corrected according to inclination of the rock
surface to account for gravitation force (Runkiewicz
and Brunarski, 1977). Schmidt hammer blows were
performed on the same boulders chosen for micro-
roughness measurements; however, we limited the
number of the boulders to ten, in order to avoid de-
calibration of the hammer. Therefore there were 100
readings of R-values per test site, meeting the require-
ments of a minimum sample size for this analysis ac-
cording to Niedzielski et al. (2009). Thirdly, a part
58 JÖKULL No. 63, 2013