We present measurements of crystal
The greatest source of uncertainty in forecasts of sea level rise during the
21st century is the contribution from the Antarctic and Greenland ice sheets
(e.g.
Typically, ice is modelled as a very slow flowing, non-linear viscoplastic
fluid where the effective viscosity is highly temperature dependent, varying
by
Under terrestrial conditions ice exists in the hexagonal Ih phase and
individual crystals possess a high level of plastic anisotropy because their
dominant mode of deformation is slip on crystallographic basal planes (e.g.
Various numerical flow relations for ice, where the effect of
Accordingly, the ongoing value of physically motivated, yet complex flow
relations lies in their role as tools to understand intra- and
intercrystalline microdeformation, recovery and recrystallization processes
(e.g.
Here we present ice crystallographic
Location of the drilling site for the 1196 m Dome Summit
South (DSS) ice core on Law Dome, East Antarctica. The drill site is
4.7 km SSW from the dome summit. The background image is from the
Landsat Image Mosaic of Antarctica
DSS borehole site and ice core information
A primary consideration in the selection of the DSS drilling site was the
identification of a location where the rates of ice deformation, and those
upstream, were insufficient to significantly disturb the chronology of annually accumulated ice layers.
Ice-penetrating radar was used to assist site selection by identifying
locations where the regional bedrock topography was least likely to have
created flow-induced folding or other discontinuities in the ice laminae
The upper 96 m of the DSS core was obtained with a
270 mm diameter thermal drill during the 1987–1988 austral
summer and the borehole was cased to a depth of 82 m. In the
following season a drill shelter was constructed over the borehole and
thermal drilling continued to 117 m depth using a smaller
120 mm drill. Over the 1989–1990 and 1990–1991 field seasons
the electromechanical drill used to recover the main DSS ice core was
assembled and commissioned
Ice crystal
The
The crystal
As outlined by
Due to the time-consuming procedure required to manually determine
During the 1990s when the DSS ice core was drilled, the measurement of ice
crystal
The mean grain size was determined from polaroid photographs of the thin sections. A purpose-built stand was used to position the thin sections between orthogonal plane polarizing filters. This allowed individual grains to be visually distinguished by their orientation-dependent birefringence colours. A polaroid camera, mounted directly to the stand, captured high-quality colour images on a 1 : 1 scale. A transparent cover plate, etched with a 10 mm square grid, was placed over the samples to superimpose a grid of the same dimensions onto the polaroid images.
The mean grain area was calculated from the number of grains contained within
a specified area of the thin section. The region of interest was determined
by placing a sheet of low-opacity tracing paper over the polaroid image; its
transparency allowed an irregularly shaped region to be traced along grain
boundaries and the number of grains within this region to be counted. A
digitizing tablet was used to accurately determine the area of the
irregularly shaped region marked on the tracing paper. Typically these areas
for grain size analysis varied between 880 and 2400 mm
Uncertainty in the mean grain area estimates may originate from errors in
counting the number of grains within the traced region or when determining
the area of the traced region using a digitizing tablet. For digitizing
tablets, instrument-related planimetric position errors are typically
To determine an upper limit on the uncertainty in the calculated mean area we
assume the maximum error in the position of any section of the traced
boundary enclosing the counted grains to be
Some further general comments on the methods used to specify the grain size
of polycrystalline materials are warranted. While thin section analysis is
the only practical means of routinely estimating grain area, it only provides
a two dimensional estimate of a volumetric object. Furthermore, all methods
used to determine grain areas from thin sections underestimate the actual
grain dimensions because the sectioning plane almost never intersects each
grain across its plane of maximum cross-sectional area (e.g.
In the methods described by
Methods of grain size analysis that require measurement of linear intercepts
through grains (e.g.
The single mean grain size measurements per thin section included in this
data set provide a coarse representation of grain size in comparison to the
data that can be obtained using modern instruments (e.g.
Measurements of
In the numerical models used to simulate ice sheet dynamics, it is more
practical to specify a constant ice density when calculating the magnitude of
the stresses driving ice flow. The assumption of a constant density is
reasonable for ice sheets where ice thicknesses may be hundreds or thousands
of metres; however, it is necessary to convert any data sets used for
calibration or validation of the models to an ice-equivalent depth scale.
Similarly, the interpretation and application of ice core chemistry in
palaeoclimate studies is simplified by conversion to an ice-equivalent depth
scale. This is particularly convenient when synchronizing multiple ice core
records from different geographical locations. The crystal
Description of the data fields in each row of the csv formatted DSS
ice core
The combined crystal
A single comma-separated value (csv) file containing A MATLAB™ structure array containing
Variation in the eigenvalues,
The crystallographic
Variation in the DSS ice core mean grain area with actual depth. See
text for details of the mean grain area calculation. All values were
determined from horizontal thin sections (after
Variation of the standard deviation,
The eigenvalues are related as
Allowances can be made for the distribution of grain sizes encountered in
polycrystalline materials by weighting the contribution of individual
Format of the MATLAB™ R2015b
structure array,
For a population of
Sample of the thin-section figures included with the data set. From
left to right: (i) lower hemisphere
Equation (
A component of the observed variability in the DSS fabric and mean grain size
data (Figs.
Estimating the magnitude of impurity effects on the variability in derived
microstructural parameters is not possible with the DSS data set; however,
observations from
The data set made available for download at the Australian Antarctic Data
Centre also includes graphical representations of the
Observations of ice microstructures from deep drilled polar ice cores play a
vital role in the development and validation of ice flow relations for
numerical ice sheet modelling. In particular, measurements of the patterns of
ice crystal
Measurements of crystal
The Australian Antarctic Division provided funding and logistical support for drilling the DSS ice core and subsequent data analysis through projects ASAC 15, AAS 757 and AAS 4289. The authors gratefully acknowledge the contribution of all participants in the Australian National Antarctic Research Expeditions associated with retrieval of the DSS ice core. Preparation of the data for archiving was supported by the Australian Government Cooperative Research Centres Programme through the Antarctic Climate and Ecosystems Cooperative Research Centre (ACE CRC). Discussions with J. L. Roberts assisted with data management and manuscript preparation. B. Raymond assisted with data control and hosting. We are thankful for comments from Maurine Montagnat and an anonymous reviewer, who assisted in improving the manuscript. Adam Treverrow thanks R. C. Warner for stressing the importance of making these data widely available to the glaciological community. Edited by: O. Eisen