I. Introduction

McCall Glacier lies in the eastern part of the Brooks Range, in the Romanzof Mountains, at lat. 69 18' N., long. 143 48' W. The glacier has an area of 6.22 km² and an altitude span from 1340 to 2720 m (Fig. 1). It is one of the few small glaciers which exist in the low precipitation environment of the Brooks Range at altitudes above 1000 m. Most of these glaciers have northern exposures, and only a few face south, so that the energy balance, particularly its net radiation term, is unlikely to be an important factor in the presence of these glaciers.

Fig. 1.Aerial view of the three accumulation basins of McCall Glacier looking towards the south.

To investigate this, a heat-balance study was carried out on the McCall Glacier in summer 1970 as part of a combined heat-, ice-, and water-balance investigation, which is being conducted under the auspices of the International Hydrological Decade (I.H.D.). McCall Glacier is the only Arctic glacier currently being studied in the United States of America, and is of special importance as it lies at the intersection of two glacier chains recommended for intensive study in the I.H.D.: the Arctic Circle and the American chains. It had been studied previously during the I.G.Y. (e.g. Reference KeelerKeeler, 1959; Reference OrvigOrvig, 1961; Reference Orvig and MasonOrvig and Mason, 1963), and by us from 1969-72. The results of the mass and water balance for 1969 and 1970 have been published (Reference WendlerWendler and others, 1972, in press).

It is hoped that the present study will provide a better quantitative understanding of the relationships between energy transfer, climatic and meteorological elements, and melting processes. No attempt was made to integrate the heat balance over the whole glacier; rather, it was calculated for a single point at 1730 m altitude.

2. Period of Observation and Instrumentation

The observations started in summer 1969. However, as the micrometeorological instrumentation was not in good working condition until the end of the ablation season (August), the data of 1969 were not included in this study.

The observations started again on 7 April 1970, but as the ventilation system for the temperature sensors was not working before 17 May, only the period between 17 May to 31 August was analysed. On 1 September the micrometeorological measurements were stopped. This period includes the whole ablation season for the altitude on the glacier at which the measurements were carried out.

The observations were made about 100 m from the foot of the lateral moraine, somewhat east of the middle of the glacier tongue (see Fig. 2). The glacier slopes to the north at about 7 to the horizontal at this point, and this is also the general exposure of the glacier, which has a fairly simple geometry on a northsouth axis.

Fig. 2.Locality map of the McCall Creek Basin.

The radiation was measured with a PD-₄ Davos radiometer. This instrument has four sensors, two of which are glass-shielded, the other two polyethylene-shielded, so that the incoming and reflected short-wave and the incoming and outgoing all-wave radiation can be measured independently. Long-wave incoming and outgoing radiation can then be computed as differences. The four radiative fluxes as well as the zero point and the instrument temperature were recorded continuously with Siemens six-channel galvanometric point recorders, which were located in a hut on the moraine about 140 m from the site of observations. As back-up for the incoming short-wave radiation, a Belford actinograph was utilized. The radiation measurements were calibrated in the field against a standard Linke-Feussner actinometer, which had been calibrated by its manufacturer Kipp en Zonen, Delft, Holland.

Wind speeds and temperatures were measured in a logarithmic profile (0.5, 1, 2 and 4 m) above the glacier surface. The instrumentation was adjustable in height so that it could be at constant height above the surface (Fig. 2). Raim 3-cup micrometeorological wind sensors which have low starting velocities (0.3 m s⁻¹) were used to measure the wind speed. For every revolution of the anemometers, an electrical pulse was transmitted to Sodeco digital counters located, with the rest of the recording instruments, in the hut, and the integrated number of revolutions was printed out every half hour.

The air temperatures were measured with artificially ventilated thermocouples, and the ice or snow temperatures were measured with eight thermocouples buried in the ice down to a depth of 8 m; the output was continuously recorded on the Siemens point recorders. The dew points were measured with Panasonic aluminum oxide sensors situated at two altitudes (0.5 and 4 m); these sensors change their electrical resistance with increases and decreases of the atmospheric water-vapor pressure. The ablation was measured twice daily with ten small (5 mm) ablation stakes situated near the micrometeorological equipment, and the snow density and stratigraphy were measured occasionally during the time when there was snow cover at the site. At a later date (17 July) a meteorological shelter, containing a thermohydrograph calibrated against an Assmann psychrometer, and a maximum and minimum thermometer, was also placed near the micrometeorological equipment. With the help of these latter instruments the heat fluxes could be estimated, at least for melting conditions, during times when the micrometeorological instrumentation was being calibrated or not in working condition.

3. Climatology and Weather Conditions

The mean and extreme climatological conditions can be obtained from Table I. The mean temperature for the latter part of May (17-31) is negative ( -1.0 C), and even the June temperature, at 0.5 C, is near freezing point. In July the highest mean monthly temperature 3.8 C, was observed, while in August it became colder again (1.3 C). The relative humidity like the temperature, showed an increase from 69% in May to 78% in July and then increased to a maximum value of 80% in August. The wind speed did not vary substantially during the summer months, but the cloudiness showed a steady increase from May (5.1 tenths) to August (6.7 tenths), a result which is in agreement with most other observations in the Arctic regions (e.g. Reference SearbySearby, 1968). Since no long-term climatological records exist for McCall Glacier, climatic trends can be determined from Barter Island, 90 km north of the glacier. Temperature means for the period 1 May to 31 August 1970 were 0.5 C colder than the long-term mean.

Table I.Meteorological Data, Mccall Glacier, Summer 1970

In Table II the diurnal variations of the meteorological elements are given. The highest variation in temperature was found in May, and the daily amplitude decreased steadily throughout the summer. This course was not followed by the water-vapor pressure, which showed its largest diurnal variation during the month with the highest temperature. However, this is not really surprising, as at higher temperatures the amount of water vapor the air can hold is much greater; the same temperature rangewhen assuming a constant relative humiditywould then result in a bigger diurnal variation in water vapor.

Table IIDiurnal Variationsin Meteorological Conditions, Mccall Glacier, Summer 1970

The daily mean values of the climatic parameters are given in Figure 3. Naturally, there is a great deal of variation in these values, and it can be noted that there are extended periods during which no freezing occurs at all. Nevertheless the melting season is only 2-3 months long.

4. Heat Balance at the Glacier Surface

The heat-balance equation at the glacier surface consists of the following components.

with (RB)_s the short-wave radiation balance, (RB)_L the long-wave radiation balance, S the sensible heat flux, L the latent heat flux, B the heat flux in the snow or ice, and M the snow or ice melting.

All fluxes which bring energy towards the surface were considered to be positive, while the fluxes taking energy away from the surface were considered to be negative. Other fluxes, e.g. rain, were estimated, but were so small that they could be neglected.

Fig. 3.Daily averages of meteorological elements, McCall Glacier, summer 1970.

A substantial potential error could be introduced by neglecting advective heat fluxes, which could not be estimated. The topography and nature of the surrounding terrain of the glacier makes it almost certain that on many occasions heat was advected to the measuring site, but the magnitude of this effect is unknown.

The period of measurements was subdivided into four phases:

• (a) pre-melting period 17 May-12 June

• (b) melting period, snow 13 June-17 July

• (c) melting period, ice 18 July-28 August

• (d) post-melting period 29 August-31 August.

The distinction between these four phases is not perfect, as on some warm days in spring some melting occurs, and cold spells happen during the whole melting season; however distinctive differences are still observed during these periods. Furthermore, it should be noted that the last period consists of only three days, and therefore, the values found for this period are not necessarily typical for the early post-melting phase.

(a) The radiation balance

A summary of the radiative fluxes which were obtained with the PD-₄ Davos radiometer (Fig. 4), is given in Table III. For the pre- and post-melting periods, negative values were found for the radiation balance while they were positive during the melting period. The short-wave incoming radiation decreases during the summer. This is an effect of the increasing cloudiness coupled with the decreasing day length during the latter part of the summer. The albedo is high (nearly 80%) for the fresh snow covers which exist during the pre- and post-melting periods and decreases during the melting period to 58% (snow) or 48% (ice) respectively. The value for ice (48%) agrees with values found for superimposed ice (Reference HolmgrenHolmgren, 1971), but is too high for glacier ice; Reference Wendler, Wendler, Fahl and CorbinWendler and Ishikawa (1973) found a value of 29%. The high value is explained by two snowfalls during the melting period, which raised the mean albedo for this period appreciably as shown in Figure 5. The snowfall of 1 and 2 August in particular raised the albedo substantially for about eight days. Thus new snowfalls in summer are extremely important for the mass balance, as has been pointed out before (e.g. Reference HoinkesHoinkes, 1968), both by modifying the radiation balance and by adding to the mass balance of the glacier. For example, for the total ablation period on McCall Glacier during 1970, the amount of solid precipitation added to the annual mass balance was about 10%. The energy required to melt this mass was about equal to the energy lost due to the increased albedo of the new snow cover, reducing the absorbed radiation.

Fig. 4.PD-₄ Davos radiometer. Note double glass domes (lefl) and Lupolen-shielded sensors (right)

(b) Sensible and latent heat fluxes

The eddy fluxes were calculated from profile measurements with instruments on a 4 m high tower (Fig. 6) using Prandtl's relation (Reference LettauLettau, 1939, Reference Lettau1949; Reference PrandtlPrandtl, 1956). In doing this, it was assumed that no advection takes place. Furthermore, the transfer or austausch coefficient found for momentum exchange is assumed to be identical with the transfer coefficient for sensible and latent heat exchange, which is not necessarily correct for non-adiabatic conditions (Reference WebbWebb, 1965). During adiabatic or near-adiabatic conditions, logarithmic profiles were fitted to the wind observations to obtain a mean value for the roughness parameter z₀ for ice of 0.24 cm. This value is in good agreement with those quoted by other investigators (e.g. Reference HoinkesHoinkes, 1953; Reference UntersteinerUntersteiner, 1957; Reference Streten and WendlerStreten and Wendler, 1968). For snow, a mean value for the roughness parameter of 0.09 cm was found, which again is comparable with values found by numerous other authors under similar conditions (e.g. Reference LiljequistLiljequist, 1957; Reference Wendler and StretenWendler and Streten, 1969).

Fig. 5.Daily mean albedo and new snowfalls on McCall Glacier, summer 1970.

Table III.Radiative Fluxes, Mccall Glacier, Summer 1970

Fig. 6.Micrometeorological tower on McCall Glacier. Note the ventilated temperature sensors (left), the humidity sensors (center). The sensors shown are not the Panasonic sensors mentioned in the text, but are Pernix sensors installed at a later stage (not used during 1970). Also note the wind-speed sensors (right).

For non-adiabatic conditions, correction was applied to the austausch coefficient according to Reference LettauLettau (1949):

where A is the austausch coefficient, A _a the austausch coefficient for adiabatic conditions, and x a dimensionless stability criterion similar to the Richardson number.

The mean corrections for the four periodspre-melting, melting snow, melting ice, and post-meltingwere found to be 7, 15, 14 and 31% respectively, which means, that owing to the generally stable air above the glacier surface, the eddy fluxes were somewhat supressed.

Applying this correction, the eddy fluxes listed in Table IV were computed from the temperature and water-vapor profiles measured above the glacier surface.

Table IV.Eddy Heat Fluxes, Mccall Glacier, 17 May 31 August 1970

It can be seen that the mean sensible heat fluxes were positive for all four periods. The highest values were found to be about 50 Ly d⁻¹ (2 MJ d⁻¹ when melting occurred, the result of a strong temperature gradient above the glacier surface and relatively strong winds.

The latent-heat fluxes were negative for the first three periods, which means that evaporation exceeds condensation. The highest value was found during the pre-melting period.

The amount of evaporation decreases during the summer, and in late summer (post-melting period), the condensation exceeds the evaporation, and a positive flux towards the surface is found. This again is understandable, as the relatively warm summer air is now cooled, and can therefore hold less moisture, hence condensation takes place.

(c) Heat flux in the glacier ice

The heat flux into or out of the glacier ice could be calculated from the temperature measurements which were made at eight points down to a depth of 8 m and the known values of density and specific heat of snow and ice. The mean flux values obtained are shown in Table V.

Table V.Heat Flux in the Ice, Mccall Glacier, 17 May-31 August 1970

During the first three periods, a downward flux, that is a warming of the ice or snow, was observed, while during the post-melting period, a flux towards the surface was measured. During the pre-melting period the flux is small, as snow is not a good conductor of heat. The high value for the snow-melting period is at first glance astonishing, especially as the top snow layer becomes isothermal and no heat can be transferred through it by conduction. However, the melt water percolates into the snow-pack, and part of it refreezes at a lower level, releasing its latent heat and transporting a much greater amount of heat than can be carried by conduction. In midsummer (melting ice), the heat is transported by conduction only, as ice is not permeable to water, except frequently in the uppermost layer, which fractures into a permeable honeycomb structure. In this top isothermal layer, which has a maximum thickness of about 10 cm, heat is transferred by percolating melt water and radiation only. As ice is a better -conductor than snow (ten times better than snow of density 0.3 Mg m⁻³) more heat is conducted during the melting periods. During the post-melting period the flux is towards the surface, but is small, owing to the presence of new, low-conductivity snow.

(d) Heat for snow and ice melt

To melt either snow or ice, latent heat energy of 330 kJ kg⁻¹ is required. The amount of snow or ice melt was measured carefully twice daily with ten thin (5 mm) ablation stakes. Owing to the effects of settling of the snow and density changes, it is very difficult to measure daily values of snow ablation accurately. A mean standard deviation of 24% was found in the daily variation of the surface level between the 10 stakes.

It is easier to measure the ice ablation, but the standard deviation is nearly as large for daily values (18%), since the amount of ice melt is on the average about half that of snow melt, and hence the inaccuracy in reading the ablation stakes becomes more important.

The maximum daily ablation values of the ten stakes were about 5.1 cm for snow and 6.1 cm for ice, with mean values of 3.1 cm snow and 1.9 cm ice, while during both periods cold spells occurred, in which no ablation took place at all for a 24 h period. In Table VI the mean and extreme values for the energy used for melting are given. It can be seen that energy available for melting increased by almost 50% after the snow cover had melted.

Table VI.Heat Fluxes for Snow and Ice Melt, Mccall Glacier, Summer 1970

(e) The heat balance as a whole

The heat balance is shown diagrammatically in Figure 7, and the percentage values are given in Table VII. It can be seen, not surprisingly, that the radiation balance is the largest source of heat for melting. This has been shown numerically by many previous investigations for different snow and ice terrain, e.g. on glaciers by Reference LaChapelleLaChapelle (1959) and Reference Ambach and HoinkesAmbach and Hoinkes (1963), on ice sheets by Reference AmbachAmbach (1963) and for the seasonal snow cover by Reference Gold and WilliamsGold and Williams (1961) and Reference WendlerWendler (1967); summaries are given by Reference GeigerGeiger (1965) and Reference KondratyevKondratyev (1969).

The study is subject to a number of errors,

• (a) theoretically by the two assumptions:

  • (1) no advection takes place,

  • (2) the austausch coefficients for momentum, sensible and latent heat are identical; and

• (b) by the errors in the measurements.

The assumption of no advection is likely to be inaccurate, as discussed previously, but no estimate can be made of the magnitude of the error due to this. The second assumption is also not likely to be quite valid, since neutral conditions of stability occurred rarely and the air was generally in stable equilibrium. The actual errors in measurements can be estimated better. The errors in the radiation measurements are estimated to be within 5% under normal conditions. However, owing to new snowfalls, and occasional breakdowns of the instrumentation, making interpolations necessary this accuracy probably decreased to 8%. This includes the small error introduced due to the radiative fluxes having been measured on a horizontal surface, while the glacier is inclined to the horizontal at about 7 deg. The eddy fluxes are believed to be within 10%.

Fig. 7.Mean daily values in Langleys of the components of the heal balance for four periods of SUMMST 197o, McCall Glacier (1 Langley - 41.9 kJ m⁻²).

Table VII.The Components of the Heat Balance, Mccall Glacier, 17 May-31 August 1970 (Expressed as percentages of the heat energy source)

5. Conclusions

Several features of the relationship between mass and energy balance of McCall Glacier are of interest. As elsewhere, the most important contribution to snow and ice melt is radiation. This contribution amounts to about 60% of all energy sources. On the other hand, the magnitude of the radiation balance in summer is only about half of that of the tundra north of the Brooks Range at Barrow (Reference Kelley, Kelley, Bailey and LieskeKelley and others, 1964; Reference Kelley and WeaverKelley and Weaver, 1969; Weiler and others, 1972). This means that while the winter values, which consist mostly of long-wave radiation, are probably not greatly different for both snow-covered Arctic locations, the annual radiation balance is considerably less energetic on McCall Glacier. This is so despite the presence of fairly persistent low stratus cloud decks at Barrow, which generally cover the area north of the Brooks Range, but are below the altitude of McCall Glacier. Both high surface albedo and, to a lesser degree, the screening of the sun by the mountains surrounding the glacier, which reduces the duration of sunshine on the glacier surface by a mean of 39% in summer, combine to give a low radiation balance, favorable to the continued existence of the glacier.

In the ablation process, evaporation is responsible for only about 2% of the total ablation; the rest occurs by melting. This is a fairly typical value, reproduced quite well at other Arctic locations (Reference Weller, Weller, Cubley, Parker, Trabant and BensonWeller and others, 1972). The contribution of evaporation to the ablation is more important in spring, when it accounts for 100% of the ablation, of course, before the snow begins to melt, but its importance in actual energy and percentage terms decreases as summer progresses. When melting commences at the snow surface, the melt water percolates into the snow and refreezes at lower levels in the snow pack. This is a very effective way of transporting energy into the substratum, other than by conduction. Melting also occurs frequently in the top to cm of the ice, after the snow has melted, to create an isothermal layer across which heat is transported by melt water circulating in the loose lattice of melting ice crystals, and by radiative transfer.

In recent years, the annual mass balance of McCall Glacier has been negative in i969, 1970, 1971, and 1972, the years during which its balance was studied intensively. Fahl (unpublished) has recently shown what type of large-scale synoptic pattern over northern Alaska determines the growth of the glacier mass by precipitation. The present study has attempted to show how physical processes of energy exchange at the glacier surface determine its reduction.