Geology Reference
In-Depth Information
0.10
0.15
0.20
1. 0
0.9
0.8
0.7
λ
1
0.6
Tw o
bottom
sections
0.5
0.4
Figure 4.52
A macrograph of a horizontal thin section at a
depth of 1.59 m of unidirectionally frozen, columnar‐grained
sea ice in Resolute Bay (Canadian High Arctic). Brine spacing
λ
1
is shown. The double‐headed arrows indicate the average
c
axis of the grains.
0.3
0.5
1. 0
1. 5
2.0
Growth rate (cm/day)
Figure 4.51
Plot of average brine layer spacing (open circles)
versus the corresponding growth rate of curve
b
in Figure 4.50.
The solid circles indicate the bottom two sections. The broken
line shows a least squares fit assuming that the spacing is inversely
proportional to the growth rate [
Nakawo and Sinha,
1984].
field at Resolute where a microscope and transmitted
light were used (this is referred to as field measurements)
and the second was by applying image processing tech-
niques on digitized slides of the thin sections in a labora-
tory in the NRC (Ottawa). This is referred to as laboratory
measurements. Figure 4.52 is an example of a macropho-
tograph of a horizontal thin section showing brine along
grain and subgrain boundaries (white color) taken at
−16 °C, using combined cross‐polarized and scattered
light. The brine spacing
λ
1
is shown. Figure 4.53 shows a
profile of brine spacing from one of the ice cores. Large
scatter is observed in both field and laboratory measure-
ments with the two sets differing by more than 0.1 mm at
some depths. Field measurements tend to show larger
fluctuations than the estimates using the digital image
processing. The same reference presents a plot between
brine layer spacing and the salinity (Figure 4.54). The
average spacing tends to decrease with the increase in salt
entrapment (i.e., salinity) according to the following lin-
ear regression equation:
One of the studies on brine layer spacing that includes a
full record of the ice growth rate is presented in
Sinha and
Zhan
[1996]. The authors conducted a study in Resolute
Bay, Canadian High Arctic, in May 1992 and obtained
the record of the ice growth starting shortly after the ice
formation in November 1991 (when it was safe to operate
a light truck on it) from the Resolute weather station of
Environment Canada. The growth season of 1991-1992
was rather unique because the ice cover was flat and the
snow cover was very uniform—a good indication of uni-
directional freezing. Moreover, the mean daily air temper-
ature for the entire ice growth season stayed well below
the freezing point, which ensured almost steady growth.
The ice growth record allowed the authors to develop the
following relationship for the dependence of ice thickness
h
i
on the number of freezing day N:
52
83
h
0 0162
.
N
27110
.
N
1 48 10
.
N
(4.3)
1110093
.
.
S
(4.5)
i
1
which gives a growth rate of
This equation was developed from the available set
of data and therefore should not be extrapolated for salin-
ities higher than about 5‰ or lower than about 2‰. If
extrapolated to zero salinity, the equation gives a brine layer
spacing of 1.1 mm. This is obviously erroneous and may lead
to wrong deductions. For growth rates generally encountered
in nature, dendritic growth does not occur in freshwater.
Perovich and Gow
[1996] conducted a wide‐ranging
study to characterize brine and air inclusions in all ice
5
8
2
(4.4)
dh dN
/
0 0162 54210
.
.
N
4 44 10
.
N
i
With these two equations the growth rate at any depth
of the ice cover can be obtained and therefore related to
the estimated brine layer spacing.
Sinha and Zhan
[1996]
estimated the spacing through two analytical methods of
thin sections. The first was in the NRC cold room in the
