Jökull - 01.01.2013, Page 59

Surface deterioration of glacially abraded basaltic boulders by Fláajökull
of the abraded surface was chipped-off using a chisel
and the weathering rind thickness was measured with
the use of an electronic micrometer and a magnifying
glass in three representative places.
Data pertaining to micro-roughness, Schmidt
hammer R-values, and weathering thickness (Ap-
pendix 1) were treated with statistical tests using
Statistica 10.0 software. The Shapiro-Wilk test
proved that the populations of readings were predom-
inantly not normally distributed; therefore, the non-
parametric Kruskal-Wallis and post-hoc Dunn’s tests
were used to find whether populations of readings ob-
tained at one study site are significantly different from
other test sites (Appendix 2). Finally, the Spearman’s
rank correlation coefficients were calculated to check
correlation between the age-dependent test site rank
and measured parameters (Table 1). All tests were
performed assuming significance level α = 0.05.
In order to supplement lichenometric dating of
Da̧bski (2002, 2007), we measured enveloping cir-
cles of 260 relatively circular thalli of lichens sub-
genus Rhizocarpon (Benedict, 1988) exclusively on
site Ia. As the growth rate of the lichen is nonlinear
and slows down after the so called "maximum growth
period", we applied two survey methods: 1) the largest
five method (based on the mean of five largest thalli)
of Thompson and Jones (1986), according to which a
growth rate of 0.725 mm year−1 characterizes lichens
younger than 50 or 60 years (for convenience, we used
55 years for the growth rate change), and a rate of
0.585 mm year−1 for older thalli, and 2) the size-
frequency approach after Bradwell (2004), according
to which, the younger moraine surface, the steeper
the regression line is within the total lichen distribu-
tion (including only lichens from the modal class and
larger). According to the method, the date is obtained
using the following formula: y = 7.307 x−0.9, where y
is surface age and x is the lichen population gradient
multiplied by -1 (Bradwell, 2004).
Pilot petrographic analyses under an optical mi-
croscope and a scanning electron microscope (SEM)
equipped with a microprobe Cameca SX100 were per-
formed on three rock samples taken from test site II –
moraine exposed to weathering since AD 1907 and
from recently deposited moraines (post AD 2000, site
VII+), in order to check mineral composition, poros-
ity and texture of basalts and alternations within the
weathering rinds. This was done to get insight into
the nature and causes of the development of rock sur-
face micro-roughness and weathering rinds.
Handysurf E35-B electronic profilometer
According to our knowledge, the Handysurf E35-B
electronic profilometer has not been, apart from the
recent works of Da̧bski (2012a,b), previously used
in geomorphological research. It is equipped with a
skidded pick-up with a built-in stylus which can reg-
ister rock surface micro-roughness down to a frac-
tion of a micron (vertical resolution 0.01µm). The
diamond stylus tip with a radius of 5 µm is pressed
against the measured surface with a force of 4 mN
or less and run with a speed of 0.6 mm/s along a se-
lected profile, which is a distance on a surface up to
12.4 mm long; in this study it was set for 4 mm, a dis-
tance typically used by engineers measuring industrial
steel elements. Roughness elements are calculated
based on the evaluation length of the roughness pro-
file, which consists of five elementary segments (sam-
pling length, referred to as cut-off value), to produce
roughness parameters. One measurement produced 4
roughness parameters: Ra, Rz, Rzmax and Rsm (Fig-
ure 2). The profilometer is equipped with a micro-
processor, LCD display and a light portable printer
enabling quick print-outs of magnified surface rough-
ness profiles (Figure 3). A significant disadvantage
of Handysurf E35-B is that it can register micro-relief
amplitude only up to 320 µm, therefore is it rather un-
suitable for coarse-grained rocks.
RESULTS
Lichenometrical dating of site Ia
Lichenometric dating of the test site Ia, based on
the size-frequency approach of Bradwell (2004), re-
vealed unexpected results of AD 1975 or AD 1966.
Theses dates are significantly younger than the age
obtained in previous dating for the Fláajökull old-
est moraine along the Hólmsárgarður profile (Da̧bski,
2002, 2007).
Using the mean of the 5 largest thallus and a
changing growth rate, from 0.725 mm year−1 for
young thalli to 0.585 mm year−1 for thalli older than
JÖKULL No. 63, 2013 59