Andreas Hertl, Innsbruck

Dagrun Vikhamar, Oslo

April 1999

1.1 Background and objectives

1.2 Site description

2.1 Ice core drilling: Density investigations of Ice Cores gathered with the Motor Drill

The Motor Drill

Practical work and problems arised

Measurement of Density

2.2 Crystallography of thin cross sections

Samples

Sampling method and Analysis

2.3 Dielectric profiling (DEP) of ice cores

Impurities and dielectric behaviour of ice

Description of the DEP-system

Measurement procedure

3.1 Density values

Suggestion for improvement

3.2 Crystallography of thin cross sections

Suggestion for improvement

The glaciological investigations were carried out on Storglaciären, a glacier situated close to the Tarfala Research Station (Figure 1). Storglaciären is located in the Kebnekaise massif in Northern Sweden (in Swedish Lapland, 67° 54’N, 18° 35’E). Three major tectonic units dominate the area: the Tarfala amphibolite, the Storglaciären gneiss and the Kebne-dyke complex (Bronge 1996:163). The Kebnekaise massif is a part of the Scandinavian Caledonides.

The Tarfala Valley is a sub-arctic high alpine area with an altitude between 800 m.a.s.l. and 2114 m.a.s.l. (Grudd & Schneider 1996:115). The climate is influenced by the north-south oriented mountain range (Holmlund et al. 1996:149). The dominating wind direction is from west. The orientation of the mountain range and the wind direction results in an east-west climatic gradient, with maritime climate in the western part of the mountains, and more continental on the eastern part of the mountains, where the Tarfala Valley is located. The mean annual air temperature at Tarfala Research Station is –4.1°C (1965-1987). The mean summer and winter temperatures are respectively +5.5°C and –7.2°C. Mean annual precipitation is approximately 950 mm. About one third of the precipitation is summer rain (Hock & Holmgren 1996:122).

In addition to the PICO Hand-Drill the Motor Drill represents a further possibility to extract ice cores. Using the Motor Drill it is possible to gain ice as well as firn cores. The maximum reachable depth is restricted to one meter approximately whereas the length of the ice core taken by the PICO Drill could reach more than 10 meters.

The Motor Drill consists of a motor fixed to a tube of about one meter length and of 11 cm in diameter roughly. The inside diameter of the tube is 9,0 cm which is not exactly the diameter of the taken ice cores (see below). At the bottom of the tube there are two sharp blades which drill into the ice if the tube rotates. The velocity of drilling depends on the quality of the firn/ice (that means the hardness) as well as the sharpness of the blades and the velocity of the tube's rotation. Usually one full tank is enough to drill an ice core of more than one meter in length (see below).

During the practical field course, ice and firn cores were drilled during two days:

For reasons of the bluntness of the blades and maybe the quality of the ice, the drilling procedure drew out approximately for 6 hours. The bluntness resulted into several failures of the Motor Drill caused by an increased consumption of fuel. While refilling the tank the tube froze at the walls of the borehole every time in only some seconds. The solution of this problem was only possible using the physical principles of the lever. After a break in a depth of about 70 cm it was impossible to get the Motor Drill going once again. Taking out the Motor Drill with the aid of a steam drill the last part of the ice core (below 58 cm) was destroyed by melting. Therefore the enddepth of the drill was of 58 cm below startdepth (total depth: 250 cm).

The taken ice core had the following structure: In a depth of 2 cm below the startdepth of the Motor Drill there was a thin layer of small rocks. The maximum length of these rocks was up to 2 mm and the maximum width was up to 1 mm. This layer of rocks lead to a first break of the ice core. Other breaks resulting in different stability of the connections between different ice layers were noticed in depths of 212, 226, 232, 234 and 241 cm (figure 2). The 7 parts in all were individually weighed and stored in a cool place (table 1). At each part the top and bottom end was marked.

b) Two other ice cores were taken on March 13th in front of Stoglaciären glacier in an altitude of 1110 m a.s.l. where Lake Lillsjön is situated. It is supplied with meltwater from the glacier (figure 1). The Lake Ice Core I was taken near the outlet of Lake Lillsjön, Lake Ice Core II was lifted out from near the centre of the lake. Lake Ice Core I reached a depth of 124 cm from the surface of the virgin snow cover (total length of the ice core: 35 cm). The second ice core contained a total length of 75 cm and reached the base in 172 cm depth. In both cases a muddy lake sediment was reached at the bottom. The structure of each Lake Ice Core is shown in the following figures 3 and 4.

The density (D) values were calculated by using the mass (m) and the length (l) of each sample which were measured in the field and the known diameter of the Motor Drill which was 8.7 cm. The diameter of the tube was exactly 9,0 cm but due to the intensive and continual vibrations over long periods the diameter of the gathered ice core (d) should be reduced to 8,7 cm. The results obtained taking equation (1) are shown in figure 2, 3 and 4 and table 1.

D=(4*m)/(p *d²*L)*1000 [kg/m³] [equation 1]

With: D: density [kg/m³]

m: mass [g]

d: diameter of tube [cm]

l: length [cm]

The density data should be understood as rough approximate values because it consists of mean values only, especially in the case of very long samples. Small-scale differences of the structure within the ice core could not be recognised using this method. The density determination of very short samples being only a few cm long (like it is at the Glacier Ice Core at Regina's Place between 232 and 234 cm) is often incorrect because of uneven breaks between two samples. In this case these breaks are responsible for wrong volume determinations. This mistake occured very probably at sample 1 of the drill at Regina's Place and should be decreased by the stones enclosed in the ice (table 1).

Comparing the density values of the Motor Drill at Regina's Place and those of the digging group at the same place at a higher position one should realise that with increasing depth the values rise unexpectedly high by an order of magnitude (figure 5). The values vary in the upper 192 cm between 102 and 437 kg/m³ whereas the density of the Motor Drill part ranks from 504,65 and 889,15 kg/m³. Therefore the calculated increasing rate of the density values should be discussed. In this connection the recognised sharp change in facies in a depth of 192cm from weak snow to compressed firn and ice is to be taken into account. Probably the course of the values were falsified caused by the inaccuracy of the measurements (spring balance - kitchen scales) and / or by different sampling methods (Swedish Snow Sampler - natural breaks at the ice core), but the expected trend towards increasing density values with increasing depth is clearly shown.

The density values of the two Lake Ice Drills were partly at a very low level (in average 496,81 kg/m³). At Lake Ice Core I the section between 89 and 94 cm was described in the field as an ice layer, but the result of the calculation shows a density value of only 471,01 kg/m³. At Lake Ice Core II there are similar values between 133 and 146 cm of depth (420,54 kg/m³). Obviously there were certain irregularities in sampling and measurement. However, at the top of Lake Ice Core II there was an ice layer of 12,5 cm of vertical size that shows a density of more than 700 kg/m³ like the section between 147 and 164 cm of depth (682,77 kg/m³). This includes the transition from superimposed ice in the upper to lake ice in the lower part (figure 4) (see also chapter 3.2). Unfortunately the last-mentioned section was weighed only in its entirety and therefore the density value represents the visible transition in no way.

Taking samples was restricted to the Glacier Ice Core from Regina's Place (11.03.99) on the one hand and to the Lake Ice Core II of Lake Lillsjön (13.03.99) on the other hand.

Horizontal laying thin cross sections were taken from the Glacier Ice Core in a depth of 208 and 238 cm respectively.

The most interesting result from sampling the Lake Ice Core II was to show the transition of the size and structure of the crystals and the included air bubbles of the superimposed ice in the upper part to the lake ice in the lower part of the core sample. Horizontal thin cross sections were made in a depth of about 110 (D), 150 (B) and 165 cm (A). Additionally a vertical thin cross section (C) was done ranging from 150 up to 165 cm in depth. This thin section shows the abrupt transition from superimposed ice at the top to lake ice at the bottom represented in the structure and arrangement of the crystals.

The taken samples of 8,7 cm in diameter were stored in a cool place of maximum -5°C. The ice cores were sawed up in pieces approximately 3 cm long and afterwards they were pulled over a heated glass-table. Having done this the surface of the samples were melted and the result was a flat surface. The samples were put on small glass plates to which they froze immediately. After that the other end of the samples were melted on the heated table in the same way as long as the sample thicknesses were about 1 mm.

The frozen thin cross sections were studied by using polarizing sheets. As a result it was possible to analyse and interpret crystal sizes, shapes and orientations. Due to the polarizing sheets arranged in different ways the crystals were depicted in different colours. Ice crystals with vertical laying c-axes appeared in black; diagonal or horizontal laying crystals appeared in interference colours.

With the help of a scale carved in the glass the size of the crystals and of the air bubbles could be determined.

Impurities and dielectric behavior of ice

The dielectric behavior of ice depends on the amount and type of impurities in the ice. Impurities may incorporate into the ice crystal structure, or they lie outside the structure. Ice is almost an insulator, but defects in the ice structure (due to impurities incorporated in the ice crystal structure) may increase the thermal conductivity of ice. Energy is transferred along the structure bands. However, most impurities are soluble and are not incorporated into the ice crystal structure.

Impurities in the ice have different sources. There are natural impurities, anthropogenic impurities, isotopic impurities and dissolved gases. Some natural impurities are seawater salt, biogenic impurities (ocean plankton spores transported by the atmosphere), volcanic impurities and dust blowing from exposed areas (rocks and deserts). Anthropogenic impurities are man made pollution such as CO, N₂O, SO_4, heavy metals, DDT and nuclear contamination.

Impurities found in snow and ice cannot directly be assumed to equal the concentration in the atmosphere (Patterson 1994:398). Aerosols and gases in the atmosphere are incorporated into precipitation in various ways. Deposition processes, condensation and evaporation from the snow surface also influence the amount and type of impurities in the snow and ice.

Dielectric properties of ice have been found to be dependent on the acid and neutral-salt concentration in the ice (Moore 1993:245 and references therein).

Description of the DEP-system

The dielectric profiling (DEP) – system is a method to measure the dielectric properties of ice. The DEP-system measures the capacitance and conductivity along an ice core. Capacitance of ice is a measure of the real part (relative permittivity) of the complex permittivity of ice. Conductance depends on the imaginary part of the complex permittivity (a dielectric loss factor) (Moore 1993:245).

Figure 6 illustrates the DEP system. The DEP consists of two electrodes, a cradle for the ice core to rest on during measurements, a control unit and a computer. The measurements are directed from the computer. A 50 cm long ice core of approximately 10 cm diameter is placed longitudinally in the cradle. The ice core remains undisturbed and intact. A current (50 kHz frequency) flows through the ice core from one electrode to the other. The curved cradle constitute one electrode. The other electrode is also curved, and is manually moved along the ice core.

Measurement procedure

During the glaciology field course, one ice core from the firn area of Storglaciären (Regina's Place) and two ice cores from the Lake Lillsjön were analyzed using the DEP-system. The main objective was to localize variations in dielectric properties of the ice cores and correlate the variations to the stratigraphy, e.g. ice layers and density variations. No chemical analysis were performed on the ice cores.

The DEP-equipment was mounted on a table outside the cold room laboratory at Tarfala Research Station. The ice cores were not intact in one whole core, but consisted of several pieces of cores. The ice core pieces were placed on the cradle. Before measuring the ice cores, some input parameters had to be specified to the DEP computer program. The parameters were date, time, ice core drilling start depth and ice core length. The computer made a signal at a fixed time interval, indicating two centimeters of the ice core length. The upper electrode had to manually be moved along the ice core. Permittivity, capacitance and conductance were stored in the computer. After measurement the recorded dielectric properties could be viewed graphically on the computer and examined to evaluate if measurements had to be remade. During the measurements of the ice cores, notes were made on the existence of ice layers and impurities.

An important aspect regarding ice core analysis is the thermal conditions during transport, storage and measurements of ice cores. Ice cores should be protected from thermal variations and distortions from its original temperature where it was drilled. Depending on the type of measurements, the ice cores have to be stored at least in temperatures well below 0°C to avoid melting. Equipment to keep the original temperature is most convenient. It is also important that all the ice cores have been stored at equal conditions. Measurements should be carried out in cold room laboratories and storage should be in a freezer.

After drilling, the ice core pieces from the firn area of Storglaciären were packed in plastic bags and marked with numbers. They were transported on a snow-scooter from the glacier and down to the Tarfala Research Station. The ice core were stored outside since air temperatures were well below 0°C. The lake ice cores were treated the same way, but they were stored in the cold room laboratory. The cooling system was not put on, so unfortunatly the ice core pieces melted partially. The DEP-measurements were performed in a shadowed place, outside the cold room laboratory, where air temperatures were around -5°C. The cold room laboratory were too warm for measurements (above 0°C).

The gathered samples of the two Lake Ice Cores and of the Glacier Ice Core at Regina's Place taken on March 11th respectively March 13th show density values between 420,54 kg/m³ and 889,15 kg/m³. On this basis it should be spoken of firn and ice and not only of ice in most of the cases. Firn is represented by values between 500 and 700 kg/m³ whereas ice shows values of more than 700 kg/m³.

In comparison to the results of the digging group at the pit at Regina's Place the values obtained using the Motor Drill show an increase in order of several kg/m³. The density values increase from 384,27 kg/m³ to 676,11 kg/m³ in average. The lower section under 192 cm could be the snow layer of last summer compressed in an advanced stage.

The average density of the Lake Ice Core II is of 606,08 kg/m³.

The measurement of density could be improved by sawing up the ice core in sections of 5cm for instance in the laboratory of the research station. This would lead to a very high resolution of density values compared to the method done during the course. The expected advantage is to achieve a better understanding about the partly very different layers of soft and solid ice. To do this it would be more simple to compare the results obtained by using the Swedish snow Sampler (SSS) and the results based on the Motor Drill or the PICO Drill.

Glacier Ice Core at Regina's Place:

The thin sections in depths of 208 cm and 238 cm respectively differ in kind and size of the crystals as well as in the quantity and size of the air bubbles included. Corresponding to the metamorphosis of fallen snow (setting and compression) the quantity of the air bubbles decreases with advancing depth from about 5% in a depth of 208 cm to about 2% in 238 cm depth. The air bubbles show a size of 1mm in diameter in a depth of 208 cm whereas the diameter of the air bubbles in the lower thin section is clearly less than 1 mm.

On the other hand the crystal size increases in length in advancing depth from 6-8 cm in diameter to 10-15cm. In both thin cross sections crystals with a polygamous structure are dominant but additionally in the upper thin section (208 cm in depth) little ice sticks of a length of 1-2cm and a width of about 5 mm do still occur. A scheme of the analysed thin sections is shown in figure 7.

Lake Ice Core II at Lake Lillsjön:

In comparison to the results obtained analysing the Glacier Ice Core of Regina's Place the quantity and size of the air bubbles decrease going from the top to the bottom of the Lake Ice Core (from 20% air bubbles with 3-4 cm in diameter [D] down to 10% and 1-2 mm in width in a depth of about 150 cm [B]) (Figure 7).

This result can be considered parallel to the increasing of the ice crystals from 3-5mm [D] to 6-8mm in diameter [B]. Both, thin section B and D represent superimposed ice which is once deposited on a frozen lake as snow and which is metamorphosed later on. Therefore the superimposed ice is older than the underlying lake ice which is investigated in thin section A and C.

Thin cross section A represents the Lake Ice which top is lying in a depth of 159 cm. Taking the thin section the lake ice is characterised by many vertical standing ice columns which had grown downwards in direction of the lake bottom. Typical for lake ice is the totally or nearly complete absence of air bubbles in the ice. These facts are shown once again in thin cross section C (figure 8) which depicts a vertical cut through the ice core around the depth of 159 cm.

To receive unambiguous and scientifically sound results concerning structures of ice crystals and of their changes within the same ice core it isn't sufficient to prepare only 2-4 thin sections per ice core and to analyse these samples. It would be much better to make thin sections in regular distances. Doing so the differences and changes within the ice core could be analysed with a sufficient high resolution.

Impurities such as sea salt may increase the conductivity of ice. Therefore, impurities may be recognized as peaks when displaying the conductivity values recorded along an ice core. Other sources leading to increased conductivity are noise or errors, such as broken ice cores and space (air) between ice core pieces.

Measured conductivity and permittivity for the glacier ice core and one of the lake ice cores are shown in Figure 9and Figure 10. Some information are included in Table 1 regarding the weight, density and comments on each of the ice core pieces.

The glacier ice core has peaks in conductivity values around 0.07-0.10 cm, 0.31-0.35 cm, 0.37 cm and 0.45 cm. An examination of the ice core characteristics in Table 1 reveals that the peaks are not directly correlated to ice layers or density variations. The peak around 0.31-0.35 cm may be due to air between two ice core samples, as well as the peak around 0.37 cm. Other sources of error may be a misplacement of the electrode, which was moved on the ice core during measurement. This means that the curve may be shifted a little, if the electrode was moved too fast or too slowly.

DEP-measurements of the lake ice core are uncertain due to the melting during storing.

DEP-measurements should be done in controlled thermal conditions on whole ice cores. More experience is needed to drill whole ice cores of up to 1 m length. I would give more interesting results to compare DEP-measurements with high resolution density values. It is also necessary to increase the number of ice cores for drawing conclusions about the difference between lake ice cores and glacier ice cores.

The main objective of the glaciological field course at Tarfala Research Station was to gain experience with numerous field techniques and methods and the equipment used for work in arctic glacier regions.

In the field detailed introductions and descriptions of every working equipment were given which are used for the following practical work both on the Stoglaciären glacier and in the front of that glacier. In this context the authors attended the motor drill which also included the calculation and interpretation of density values, DEP-measurements and the preparation and discussion of thin cross sections.

During the practical field course three ice cores were taken in all using the motor drill. Arising problems were removed immediately in most of the cases. However in one case the problems led to a breaking-off of the drill procedure (Chapter 2.1). Afterwards the gathered ice cores were investigated for their dielectrical properties (Chapter 2.3). Density values were calculated and compared with the data obtained by the digging group (Chapter 2.1). Additionally thin cross sections of chosen parts of the ice cores were analysed in the cool storage laboratory at the research station (Chapter 2.2). The results obtained are given in the above noted chapters.

The main objective of the glaciological field course was not to receive only data and scientifically sound results. On the contrary the most important goal was to understand the various functions of the equipment and to make out possible problems of the practical work. In this context possible solutions were discussed.

The aspired objectives were clearly reached. Certainly all participants in the Tarfala field course will mark the research results received in arctic regions in a more realistic way in the future.