Glacial body measurements

RTK GPS measurements were used twice during the same Antarctic expedition (at the beginning and end of the campaign) to quantify the changes that had taken place on the glacier surface and to lay the foundations for future surveys (Reference Rentsch, Welsch, Heipke and MillerRentsch and others, 1990). This method is ideal for accurate estimates of planoaltimetric variations (Reference Vassena, Diolaiuti, Motta and SmiragliaVassena and others, 2001). It also allows repeated measurements at the same point positions without using stakes to mark them on the glacier surface, thus making it possible to increase the number of measurement points. The first survey of Strandline Glacier was carried out at the beginning of December 2000. A total of 1020 planoaltimetric points (with an accuracy of 1–2 cm) were acquired along the entire surface of the glacier (each measurement point required a survey time of 15 s). A DEM (Fig. 2) was obtained after GPS data processing on the basis of a regular 10 m grid and by applying a kriging method to the survey data.

Fig. 2.

DEM of Strandline Glacier. Both the positions of the stakes used for the mass-balance evaluation and the positions of the stakes located close to the ice-cliff boundary (used to calculate its erosion) are reported. In the lower part of the figure the two DEMs of the ice-cliff terminus processed by using data surveyed in December 2000 and in January 2001 are reported.

A second GPS survey was carried out near the end of the summer, and 100 points were selected from those surveyed the first time. Comparison of the two sets of points (equal to 10% of the first GPS survey data) revealed a quasi-stable situation, with variations of about the same magnitude as the instrumental accuracy (±0.05 m). The results do not show a clear division between the accumulation and ablation areas on the glacial body (Reference Vassena, Diolaiuti, Motta and SmiragliaVassena and others, 2001). The highest ablation was measured in the lower part of the glacier, where the surface indicates convexity shaped by eolic erosion.

Terminus measurements

To evaluate ice-cliff variations, taking into account the morphological features and the need to be informed of the changes in separate sectors of glacier terminus during the time interval, 13 vertical profiles were planned for two different periods. They were surveyed starting from 13 measurement points (identified 801–813) marked on the ground by cairns and mapped by GPS RTK techniques. Vertical profiles were surveyed using a total station GPS unit; the data obtained were processed and transformed into Universal Transverse Mercator (UTM) system coordinates. Totals of 252 points in the first survey, conducted on 22 December 2000, and 340 points in the second, conducted on 9 January 2001, were acquired. The choice of such a short survey period was based on the qualitative observations of the ice-cliff behavior during the previous expeditions. These observations indicated (Reference Motta, Diolaiuti, Vassena and SmiragliaMotta and others, 2003) substantial retreat of the ice cliff when meltwater is available and that this meltwater is present during a short period characterized by higher air temperatures.

For each survey, a three-dimensional model representing the ice-cliff surface and morphology was processed using Surfer 7 software. The model was obtained on the basis of a regular 1 m grid and calculated by applying a kriging method to the survey data. Comparison of the two DEMs permitted calculation of the volumetric change, which on a comparison surface of 4051 μ² was equal to –351.53 m³. This value corresponds to a mean thickness loss of –0.087m over the entire vertical surface, with a mean daily value of –0.0048 m. These last two values represent a gross simplification, for it is assumed that the measured change was uniformly distributed spatially and temporally (Fig. 2, bottom). The ice-cliff surface used for the comparison was identified and blanked on the basis of the upper surveyed points (in the lower part of the profiles) and the lower points (at the top of the profiles). A blank of the data above 25m in altitude was used in order to exclude the apron area (which was surveyed by total station and was appreciable on the profiles obtained) from the processed data. The apron area does not permit a real three-dimensional reconstruction of the ice-cliff surface geometry. Detailed comparison of the 13 profiles surveyed at the beginning and the 13 surveyed at the end of the campaign led to the identification of various trends for the different ice-cliff sectors (Fig. 3). As can be noted in the planimetric view of the ice-cliff boundaries (Fig. 3, bottom), which were mapped using the data collected in December and January, the lateral zones, where the front morphology changes from an ice-cliff shape to a dome and a ramp shape, showed the smallest variations. The central sector showed the greatest variations, amounting to about 2 planimetric meters of retreat (Fig. 3, bottom). It is thus clear that the mean reduction (–0.087 m) calculated over the entire terminus surface is a result of the –2m change in the central sector (ice-cliff shaped) and of the variations amounting to a few centimeters or less in the lateral sectors (ramp- and dome-shaped).

Fig. 3.

Profiles comparison. On the y axis the altitude values (in m) and on the x axis the distance (the north value in m) are reported. In the bottom panel, the ice-cliff boundaries for a planimetric comparison of changes between December and January are reported. The numbers (801–813) on the map indicate the position of the 13 points from which we measured the ice-cliff profiles; also marked on the map is the position of the fixed vertex 100 (used to calculate all the coordinates of the profiles’ data).

Glacier body

A network of ten glaciological stakes was used to evaluate accumulation and ablation on the glacier body surface. Some pits were also dug in the upper zone of the glacier to study snow stratification. The surveys revealed that Strandline Glacier is characterized by accumulation zones where wind-borne snow accumulates and summer melting gives rise to very limited seepage. The presence of meltwater was demonstrated by stratigraphical surveys. Water accumulation and refreezing was observed in crevasses in the upper part of the glacier, along with the formation of veins of congelation ice (ρ ≈ 910 kg m^–3). Ice of a different density (ρ ≈ 893 kg m^–3) was observed enclosing the veins. This ice may be linked to metamorphic processes involving viscoplastic flow, in accordance with the rheologic zones of dry-based local Antarctic glaciers (Reference ChinnChinn, 1985a Reference Chinn, b). This ice deteriorates more rapidly when exposed to sunlight, particularly on the frontal ice cliff, and is characterized by a different albedo and density (ρ ≈ 706 kg m^–3), similar to those of firn.

The thickness variations measured by field glaciological campaigns on the glacier body surface confirmed and validated the results calculated by the topographical GPS RTK surveys. The mean value obtained by stake measurements was –0.04mw.e.

Glacier terminus

Although in some zones (corresponding to vertices 806– 807–808 and to the sector identified by vertex 811 on the map in Figure 3), the field surveys made it possible to identify mass loss (in a range of meters) through calving, the limited variations in the lateral sectors were not linked to the same ablation processes. Therefore, further glaciological analyses (at the same time as the topographic survey) were carried out to identify the other processes responsible for the short-term variations on Strandline Glacier and to quantify energy transfers at the glacier–air interface.

During the 2000/01 and 2002/03 field surveys, 13 vertical stakes (Fig. 2) were located on the glacier surface close to the ice-cliff upper limit in order to calculate the daily erosion of the boundary. The data collected during the two campaigns showed that the erosion was not distributed homogeneously, and the main boundary changes (–2m of linear retreat of the boundary in the 2000/01 campaign and –4m in 2002/03) were concentrated in the zone around stakes 10–12 (also corresponding to vertices 806–807 and 808). The observations indicated that the ice losses were due mainly to falls of ice through dry calving processes. Thus, the glaciological surveys confirmed the data obtained from the topographical measurements at the ice cliff.

Not only dry calving but also melting and sublimation are ablation processes at work on the ice cliff and on the glacier surface. The evaluation of the energy and mass surface balance was carried out during the 2000/01 field campaign. During the 2002/03 campaign some experimental parcels were studied to define and quantify these processes. Parcels are glacier areas (parcel mean surface extension: 5 μ²) chosen from the glacial body or the lateral zones of the ice cliff at a location where calving is not present and the other processes have to be identified and quantified. The parcel boundaries were marked by expressly dug channels that intercept and divert the waters originating up-glacier, and by other channels that collect the meltwater produced inside the parcel. The first step was to quantify the mass loss for each parcel and to identify the driving processes.

Mass loss by evaporation and/or melting was measured by removing a block of material (ice and/or snow) from each parcel, weighing it and replacing it in its original position on the glacier in a plastic container of the same shape (Reference Motta, Diolaiuti, Vassena and SmiragliaMotta and others, 2003) without compacting the material. After a time interval (of hours and/or days), the block inside the plastic box was measured again in order to quantify the difference in weight brought about by ablation processes (melting and/or evaporation). To better understand which processes were responsible for the mass loss, we prepared two different sample boxes for two different blocks of the same material (ice and/or snow) taken from glacier areas close to each other in the same parcel. In one case, the block was inserted in a plastic box with small holes on the bottom so that the meltwater produced by the melting process would be lost, and we measured the overall mass loss (by evaporation and by melting). In the second case, the plastic box had no holes and all the meltwater remained inside. Therefore, only the mass loss due to evaporation processes was measured. The two sample boxes were positioned in the same parcel at the same time, so we could obtain the exact amount of melting and evaporation on those glacier areas

The next step was to quantify energy transfers at the glacier–air interface and the consequences for the mass loss. Particular attention was paid to the snout–ramp zone, where evidence of melting has been observed during previous field campaigns (corresponding to section 813 on the planimetric map in Fig. 3). For this purpose, the parcels were equipped with the following instruments:

1. Testostor 175 mini data logger programmed for executing synchronized temperature surveys at depths of 0.12 and 0 m every 2 min. A white cover shaded the probe at 0 m. The other probe was shaded by the box.

2. Box A, made of PET, was transparent, 0.12 m high, with an absorption coefficient of irradiance of 24.0 ± 0.5%. It was immersed in the ice and/or snow until its upper rim was flush with the glacier surface. It was filled with a block of material.

3. Box B, similar to box A, but the bottom was equipped with 1 mm diameter holes to allow meltwater drainage.

4. HD9071 data logger with LP9021RAD radiometric probes (spectral response 450–950 nm; calibration uncertainty <5%; mean cosine error <6%; non-linearity <1%) inside a horizontal tunnel protected from light penetration by means of a light-grey PVC pipe. The probes measured the downward and upward flux (Reference Liston, Winther, Bruland, Elvehøy and SandListon and others, 1999; Reference Biancotti, Motta and MottaBiancotti and others, 2000) at 0.12 m depth.

5. HD9021 data logger with LP9021UVA radiometric probes (spectral response 315–400 nm, error and set-up of the probes as with LP9021RAD).

6. Ablatometric stake made of light-grey PVC; albedo similar to the ice surface. Measurements were taken upstream, downstream and sideways, and then averaged.

The material in the boxes was weighed at intervals in order to quantify the losses and identify the processes at work. At the same time, albedo and the specific weight of the material in the box were also measured using the standard International Classification for Seasonal Snow on the Ground method, and then the block of material was repositioned in the box.

The main drawback of calculating the energy balance by box measurements is the impossibility of placing exactly the same starting mass inside the two boxes (due to the different specific weights). We therefore calculated the changes in mass taking into account the different starting weight of the material in the two boxes. The procedure is explained in the following example (using data collected on 5 December 2003). In box A (located on the lateral side of the terminus in the snout–ramp zone), we measured 20 g of mass loss. This was due to evaporation only, as box A (without holes) retains the meltwater produced. Applying a simple ratio, i.e. (weight of box A) : (loss by evaporation in A) = (weight of box B) : (loss by evaporation in B) (all measurements in g) for box B, we calculated that a mass of 17 g should have evaporated. In box B (with holes), the mass loss measured was instead 175 g (by both evaporation and melting). Considering the calculated 17 g of mass lost by the evaporation process, the meltwater produced must be equal to 158 g. This water value was used to calculate the amount of water produced in box A, which in proportion must be equal to 183 g. As the total quantity of water measured in box A was 160 g, it followed that 23 g must be trapped in the porosity of the material.

For the study period the mass balance obtained through the survey of the boxes was:

Melting of box A (with ρ = 0.64): 28.6 kg m^–2

Melting of box B (with ρ = 0.55): 24.7 kg m^–2

Evaporation of box A (with ρ = 0.64): 3.1 kg m^–2

Evaporation of box B (with ρ = 0.55): 2.7 kg m^–2

The energy balance was:

Net energy used for melting in box A: 334 kJ kg^–1. 28.6 kg m^–2 = 9552.4 kJm^–2

Net energy used for melting in box B: 334 kJ kg^–1 × 24.7 kg m^–2 = 8250 kJ m^–2

Net energy used for evaporation in box A: 2500 kJ kg^–1 × 3.1 kg m^–2 = 7750 kJ m^–2

Net energy used for evaporation in box B: 2500 kJ kg^–1 × 2.7 kg m^–2 = 6750 kJ m^–2

The balance (average of A and B) was 8901 kJ m^–2 due to melting and 7250 kJ m^–2 due to evaporation. The total energy exchanged by melting is a larger amount because the water inside the ice porosity freezes during the night, returning the melt energy absorbed.

Another way of evaluating the energy balance is by an indirect approach (we can adopt it if the quantity of irradiance absorbed by the glacial surface is known). With fair approximation (density slightly affects absorbed radiation; Reference Motta and MottaMotta and Motta, 2003), on a surface large enough to disregard the boundary conditions, the radiation absorbed by the boxes is equal to:

where R _ass is the absorbed radiation, R _i and R _e are the incident and emitted radiation at the glacier surface, respectively, and D ₁₂ and U ₁₂ the downward and upward flux at 0.12 m depth (depth of box A). For a material like that contained in the two previously described boxes, having a constant thickness and which is large enough, the formula can be approximated to:

where I (irradiance) is the power absorbed by a square meter. D ₁₂ and U ₁₂ can be disregarded in first approximation; in fact, from the field data collected, D ₁₂ is 1.5% of R _i and U ₁₂ is less than one-seventh of D₁₂ (Fig. 4). This result means that the process developing in the boxes represents at least 98% of the melting process.

Fig. 4.

Comparison of U ₁₂ and D ₁₂ data collected during the 2002/03 field campaign (in W m^–2). U ₁₂ is less than one-seventh of D _12; the line interpolates the measurement points.

While the boxes were left to work, the radiometers measured I. The energy absorbed in the same period as that of the previous example was as follows: visible and infrared 8350–18 140 kJ, ultraviolet 259–340 kJ according to the coalbedo used. Albedo ranged between 0.28 for the snow and 0.40 for the ice. So the energy balance was 8.6–18.5 MJ. These results can be compared with those yielded by the boxes, and the co-albedo increased in a way approximately proportional to density. Whichever measurement method was used, the absorption capacity of solar energy increases with density.

Finally we calculated melting and evaporation in relation to the superficial ablation, using ablation stakes. While the boxes were at work, the surface lowered by 0.025 m. According to the density of the shallow snow, this value corresponds to 11.2 kg m^–2. Applying a ratio with the quantity evaporated and melted in box B, 1.09 kg m^–2 proved to have evaporated and 9.48 kg m^–2 melted. This value corresponds to an energy income of 5723 kJ m^–2, which is definitely lower than the values calculated above through the box data. The explanation is that on the glacier surface, evaporation (and therefore energy absorption) measured by stakes was proportionally higher than in the top few centimeters of depth as measured by the boxes, and surface melting was lower because of the surface temperature conditions. In the field, we observed that on a warm day the glacier surface stayed below zero for <9 hours and that the length of this time interval decreased sharply with increasing depth, so that the time-span of temperatures below zero for 0.12m depth was <4 hours. The rate of temperature variations was calculated in the range of ±1˚Ch^–1. The surface ice temperature measurements collected in the same field campaign showed strong daily fluctuations. A negative peak in the temperature values was measured at 0.12m depth when the surface temperature dropped below –6˚C. Otherwise at 0.12 m depth, the ice remained constant at the melting point. Melting is therefore a very important ablation factor. Within the same period, it varied from 13.3 ×10^–6 to 181.4×10^–6 kg m^–2 s^–1. These variations could be caused by wind (greater evaporation) and cloud cover. In Figure 5, a diagram of the energy balance of a parcel in the ramp zone is reported. The air temperature remained slightly below 0˚C for the entire measurement period (1 week during summer 2000/01).

Fig. 5.

Energy balance of the parcel in the ramp zone at the terminus of Strandline Glacier during a week of measurements in summer 2000/01. Time (daily hours) is reported on the x axis, and energy (Wm^–2) on the y axis.