Introduction

Our study concerns the present condition of continental glaciers at mid-latitudes and the evaluation of snow and glacier resources. Although central Asian glaciers have received a good deal of attention (Reference Konovalov,Konovalov, 1979; Reference Krenke,Krenke, 1982; Academia Sinica, 1986–87; Reference zhenniang.Yang Zhenniang, 1988; Reference Hong,, Zongchao and Yifeng.Ma Hong and others, 1992), one of the largest, the Pobeda-Khan Tengry glacier system at the northern periphery of central Asia, has not been well explored. Studies of glacier energy and mass balance in remote alpine watersheds require detailed monitoring of the local climate. Snow accumulation and its metamorphism into firn and ice, snow-and ice-melt and runoff are controlled by the magnitude of energy available to drive these processes.

The first of our investigations of remote high-mountain watersheds at mid-latitudes was completed in the northern Tien Shan, the northern periphery of the central Asian mountain system, an area affected by western cyclonic and northern anticyclonic activity (Reference Aizen,, Aizen, and Melack,Aizen and others, 1995, Reference Aizen,, Aizen, and Melack,1996b). In 1990–92, we carried out investigations at the southern periphery of the central Asian mountain system in the Himalaya and southeast Tibet, with monsoon climatic conditions (Reference Aizen, and Aizen,Aizen and Aizen, 1994a, Reference Aizen, and Aizen,b). This paper summarizes investigations of local temporal and spatial Variations of the mass-energy components in the inner sub-continental central Asian high mountains, characterized by a precipitation deficit at low elevations and increasing moisture at high altitudes.

The Pobeda (Chinese name Tuomuer)-Khan Tengry massif, composed of more than 1200 glaciers with total area about 4320 km², is the largest sub-continental glacier system at the northern periphery of Central Asia (Fig. 1). The Pobeda-Khan Tengry glaciers are the major source of the main rivers in those regions of internal drainage in central Asia. These affect the great Tarim and Balkhash hydrographic systems, where 45–50% of total runoff is contributed by glaciers (Reference Dolgushin, and Osipova,Dolgushin and Osipova, 1980; Reference Zichu,, Benxing,, Jijun and Yafeng.Xie and others, 1982).

Fig. 1.

Pobeda-Khan Tengry glacier massif, central Tien Shan. Shading denotes glaciers; flags are observation stations.

Measurements and Data Collection

In the summers of 1989, 1990 and 1992, we conducted expeditionary observations on Inylchek glacier located at the center of Pobeda-Khan Tengry massif (Fig. 2). Inylchek glacier covers all glacial zones from 2900 to 7400 m and has two major branches stretching 60.5 km from east to west. The area of the glacier is 794 km².

We carried out the field measurements and observations at four points: the active ablation zone (near the Mercbakher glacial outburst lake) at 3400 m; at 4150 m, close to the firn line; at 5200 m, the upper level of liquid runoff formation; and at 6100 m in the accumulation zone. We used standard Russian hydrometeorological instruments and four automatic mini-meteorological stations built Grant Instruments (Cambridge) Ltd, England (Table 1). The data loggers of the stations recorded hourly measurements of net total radiation, total incoming radiation, reflected radiation, atmospheric pressure, snow and ice temperature, and discharge from the ablation plots. Measurements were recorded at two levels (0.5 and 2.0) in above the surface for air temperature, relative humidity and wind speed. Wind direction was recorded at 2.0 m. The glaciotogical observations included measurements of ablation, accumulation and snow-firn-ice stratigraphy in pits and ice cores. Ablation was measured on three slightly inclined 2 × 2 m ablation plots (Fig. 3) at 3500, 4150 and 5200 m. Measurements were made in the morning and evening at 121 points located within 20 cm cells. Discharge from these plots was measured using a current meter which automatically recorded the passage of water. Snow-density measurements were made in a snow pit located near the ablation squares. The density of melting ice was assumed to be 890 kg m⁻³ (Reference Shumskiy,Shumskiy, 1978). The discrepancy between measured ablation and discharge averaged 5 mm.

Fig. 2.

Southern Inylchek glacier.

Table 1.

Measurements of accumulation were made in the accumulation zone at the beginning and end of summer, because maximum precipitation occurs during summer in this region. Depth of snow cover was measured five times at more than 400 points located 50 m from one another. At a single point, the measurement error was on average about 5%. Snow density on the snow surveys was measured by electric balance.

We analyzed the tritium concentration (³H) in atmospheric moisture at 2 m above the surface, in water from glacial channels in different glacial zones (Reference Aizen,, Aizen,, Melack, and Martma,Aizen and others, 1996a) and in samples collected from snow and firn pits and obtained by hand drilling. The collection interval throughout the core varied from 5 to 30 cm depending on the homogeneity of the layers. The equilibration technique for the preparation of samples was applied (Reference Epstein, and Mayeda,Epstein and Mayeda, 1953). The precision of the samples data was 1–10 TU.

Fig. 3.

Ablation plot for measurements of ice- and snowmelt.

The statistical analysis and simulation of present and past meteorological conditions in this region were based on long-term data from a meteorological station located 150 km west of the glacier massif. The station operated from 1930 to 1990 (Reference Kobisheva,Kobisheva, 1990). We also used precipitation data collected from 1959 to 1972 at sites located at altitudes from 2800 to 4200 m in the Inylchek glacier basin. Topographic maps of the central Tien Shan region at 1:25000 scale and synoptic maps of surface and 500 m bar were also used.

Methods of Calculation of Heat and Mass Balance

The simplified thermal balance equation may be stated as:

where W is the melt intensity at a point (kg m⁻²); l is the latent hear of ice fusion (J kg⁻¹); B is the total radiation balance (J m⁻¹), including the shortwave radiation balance Q(l — A) and longwave radiation balance; Q is the incoming shortwave radiation; A is the surface albedo; P _t is the turbulent-heat flux from the atmosphere (J m₋₂); LE is the latent-heat flux due to evaporation or condensation (J m⁻²). Using hourly gradient measurements, turbulent heat and humidity fluxes were calculated from functions of similarity computed using Equation (2). All calculations were carried out using the theory of the surface boundary layer in a homogeneous fluid, developed by Reference Monin, and Obukhov,Monin and Obukhov (1954) and Reference Kazanskiy,Kazanskiy (1965).

where t(°C) is the air temperature, q (mbar) is air humidity, v (m s⁻¹) is wind, d is the distance found from non-linear equations, d _o is surface roughness, z is height of measurements above the surface. Evaporation was also measured by repeated weighing of decimeter cubes of snow and firn using an electric balance with an accuracy of 0.1 g.

To calculate an annual mass-balance index (I _bi). Equation (4) was used. All components were determined at the long-term mean altitude of the equilibrium-line position (H_e.l =4476 m) which was calculated using Reference Kurowski,Kurowski’s (1891) method. We assume this elevation is the level of the average characteristics of mass exchange (Reference Ahlmann,Ahlmann, 1940; Reference Krenke,Krenke, 1982).

where P _e.l is annual solid precipitation (from October to September) at the altitude of the equilibrium-line position, W ^e.l is total glacier melt. E ^e.l is the mass of evaporation or condensation and J ^e.l is refrozen meltwater.

The total precipitation was calculated as

where Р ₀ (mm) is mean precipitation at the Tien Shan meteorological station at H ₀ = 3614 m a.s.l.; γ(P). (0.52 mm/m) is mean altitudinal gradient of precipitation calculated from data at the Tien Shan meteorological station (308 mm) and long-term data from precipitation sites at 4200 m (673 mm).

To estimate the total glacier melt, the air temperature at the altitude of the equilibrium-line position was calculated as

where T _o (° C) is mean air temperature at the Tien Shan meteorological station; γ(Τ) (0.0053° C m⁻¹) is mean altitudinal gradient of air temperature calculated from data at the Tien Shan meteorological station and expeditionary observations at 4150 m on the glacier.

Circulation and Climatic Processes

The high ranges surrounding central Tien Shan prevent the entrance of moisture; hence, winter precipitation is small, especially in January and February, accounting for only 8–10% of the total in this region (Reference Kobisheva,Kobisheva, 1990). In summer, the level of condensation rises, which leads to a precipitation increase at high elevations. Development of convection and strengthening of unstable atmospheric stratification result in a summer maximum of precipitation in June and July caused by cold moist air masses from the west. At the same time, these high mountains are an obstacle to dry tropical air masses formed over the Takla Маkan, Gobi, Alashan and Tsaidam deserts and moved to the north. These processes have a substantial effect on the mass and energy exchange of the glaciers.

Fig. 4.

Synoptic (a-d) and meteorological (e) conditions at 4150 m on southern Inylchek glacier during the ablation period (July-August 1989) of the expeditionary observations. (a) and (b) are anticyclonic warm and cold weather; (c) and (ć) are cyclonic warm weather; (d) is cyclonic cold weather; (e) is diurnal mean air temperature (1); total radiation balance (2); glacier ablation (3); relative humidity (4).

According to our observations on Inylchek glacier and the analyses of synoptic maps, there are four main synoptic processes (Fig. 4; Table 2):

Results of Expeditionary Measurements

Studies in the Accumulation Area

The firn fields of southern Inylchek glacier stretch from southeast to northwest, from the slopes of the Voyennikh Topografov (6873 m) and Rapasov’s (6934 m) peaks to the base of Khan-Tengry peak. The area is more than 30 km² and the difference between the lowest and highest points is 1900 m (Fig. 7). A 700 m high icefall divides this area into two main parts with slightly inclined surfaces, one from 4400 to 5200 m and the Other from 5800 to 6300 m. The minimal wind and absence of snow-avalanche redistribution inside this area are favorable for the natural accumulation of precipitation. The steady snow regime is proved by its small spatial variability (σ = ± 88 kg m⁻²).

Table 4.

Range of tritium concentmtion (³ H, TU) in samples of atmospheric moisture (am), glacier runoff (gr), channel water (cw), water in glacial lakes (wl), new snow (ns), surface snow (ss), moisterJim (mf) and ice (i) in lnylclzek glacier injury and August, 1989

Fig. 6.

Stratigraphic profiles of snow, firn and ice; the temperature in drillholes and tritium content in samples from the snow, firn and ice cores at 6148 m (а),5200 m (b), 4400 m (с) of southern Inylchek glacier, (1) is new snow, (2) is fine-grained firn, (3) is medium-grained firn, (4) is coarse-grained firn, (5) is crust with dust, (6) is ice lens, (7) is dense-firn lens, (8) is crust dust and ice lens, (9) is trace of get-wet-through firn layer, (10) is get-wet layer coarse-grained firn, (11) is get-wet layer medium-grained firn, (12) is ice crust, (13) is monolith of ice, (14) is marks of annual layers.

Fig. 7.

Ice-formation zones on firn field of southern Inylchek glacier. (1) is recrystallization, (2) is recrystallization-regelation and cold infiltration, (3) is warm infiltration, (4) is infiltration, (5) is rocks, (6) is moraine terrace, (7) is observation site, (8) is drilling holes, (9) is snow pits, (10) is snow-survey transects, (11) is altimetric points.

According to our observations in 1989, 1990 and 1992, snowmelt and runoff in the accumulation area occurred to 5200 m, whereas on southern slopes this boundary reached 5800 m. The annual accumulation here is saturated and, during the day, air temperatures are about zero. During anticyclonic weather, a thin radiation ice crust is formed on the snow surface. Under the crusts, meltwater forms vast “firn bogs” in places with slight inclinations and depressions. Night cooling is insufficient to freeze this moist snowpack, therefore meltwater was drained from this zone during day and night.

From 5800 to 6300 m, the annual mean snow accumulation was found to be 918 g m⁻² (in 1989). Simultaneous measurements of precipitation, occurring during the observational periods at 5200 and 6100 m revealed an altitudinal increase of precipitation of 2 mm per 100 m.

To assess the annual accumulation and temperature, two cores about 30 m long were drilled at 6148 m. Figure 6 shows the structure of two averaged snow-firn cores. Twenty-one annual layers of accumulation were separated in the drilled holes by snow—firn—ice-core stratigraphic analysis and 23 in the 28 m depth section along the cleared firn-ice wall of the crevasse at the same altitude. Annual layers were identified by summer horizons of more compact firn and yellow-brown aeolian dust. According to our and previous studies (Reference Racek,Racek, 1954), dust indicating the annual layers is delivered from deserts only in the second half of summer and beginning of autumn during steady anticyclonical weather.

Verification of annual layers was also made by tritium (³H) sampled from the strata (Fig. 6). The marks were the layers accumulated during rather high tritium content in 1981–82 and 1985–86, caused by nuclear tests conducted in China and the disaster at Chernobyl. According to identification of the annual layers in the cores, the mean annual snow accumulation was found to be 900 mm (Table 5) at 6148 m. Results of stratigraphic analyses are justified by a good correlation (Equation (10)) between annual accumulation measured at 6148 m and annual precipitation measured at Tien Shan station.

where A _k is accumulation over the hydrological year (from September to August) in Inylchek glacier (6148 m), P is annual precipitation measured at Tien Shan station, and η is the correlation ratio. According to Chinese results (Reference ChengCheng Tong, 1982), the annual precipitation was about 800 mm at the same altitudes on the southern slope of Kok Shaal Too near Pobeda peak. Increase of precipitation with altitude is not a linear function (Fig. 8). The altitudinal distribution of precipitation was calculated on the basis of long-term meteorological data from the Tien Shan meteorological station, precipitation sites and snow surveys in the Inylchek glacier area.

Four ice-formation zones have been revealed in the southern Inylchek accumulation area (Fig. 7), none of them having any clearly delineated distribution area. The icefall dividing the accumulation area into two parts seems to lie in the cold infiltration-recrystallization zone where the meltwater becomes ice due to infiltration. Stratification analysis of snow-firn layers has shown that above 6000 m ice was formed by subsidence and recrystallization at depth exceeding 30 m. Therefore, the altitudinal belt over 6000 m we determined as a recrystallization zone where melt does not occur.

Table 5.

Thickness of annual snow-firn layers (h_i, cm), densityy (P_i , Mg m ^-3) and water equivalent (C_i, mm) in the recrystallization. zone of Inylclzek glacier; ]4- 75 August 1989, at 6148m

Fig. 8.

Vertical change of annual precipitation in Inylchek valley.

The value of ice formation in lower zones (F) was calculated by Zikin’s method (Reference Zikin,Zikin, 1962).

where C = 2.09 × 10³ J kg⁻¹ ° C⁻¹ is specific heat capacity of ice; l = 333.6 × 10³ J kg⁻¹ is heat of ice melt; p (kg m⁻³) is ice density; Z ₁ (m) is the boundary between firn and ice; Z(m) is the lower boundary of the active ice layer; and ΔT (° C) is ice-temperature variation. Ice temperatures in drillholes at altitudes of 6148, 5200 and 4400 m were measured at the beginning and the end of the ablation period (Fig 6a and c). The value of ice formation in the warm infiltration-recrystallization zone (4600–5200 m) calculated by Equation (11) was 140 kg m², under p = 880 kg m⁻³ and Z = 7 m. In the infiltration zone (4200–4600 m), the value of ice formation was 90 kg m⁻², under p = 890 kg m⁻³ and Z = 9 m.

In the 30 m hole drilled at 6148 m, diurnal temperature changes were observed only in the top layer of annual accumulation (Fig. 6a). At greater depths, the temperatures were constant and at the bottom the temperature was −21° C, that is, about annual mean air temperature at this elevation. Calculations of mean annual air temperature at this level were made through the Tien Shan station data using the temperature gradient of −0.53° C per 100 m. At 6100 m, in August 1989 and 1990 the mean diurnal air temperature was about −10° C. During daytime, it did not rise over −2.0° C, while at nights it could be as low as −25.0° C. At these altitudes, air temperature and moisture variations depended mostly on the inter-mass processes. Heat flux brought by föhns was not significant there. High humidity, observed especially at night (85–95%), was favorable for hoar-frost formation. The condensation overnight can be as high as 1.5–2.0 mm. At the accumulation area of southern Inylchek glacier screened from the west, the average wind speed was 1.7 m s⁻¹, and never exceeded 6 m s⁻¹. At the same time, on Kok Shaal Too and Tengry Tag slopes, the wind speed reached 40–80 m s⁻¹ during cyclonic and föhn weather.

During cloudless weather, at 55–60% relative humidity, the evaporation was 1.5–2.0 mm d⁻¹. It is less than at the same altitudes of western Kun Lun, in an extra-Continental climate where evaporation reached 5.5–9.7 mm d⁻¹, with 30% relative humidity (Reference Higuchi, and Zichu.Higuchi and Xie Zichu, 1989). During anticyclonic weather, local vertical fluxes develop cloudiness, humidity increased abruptly and from 1400 to 1500 h snow started. Precipitation from such cloudiness being typically local amounted to 5–7 mm d⁻¹. At nights, the cloudiness was dispersed or disappeared.

Mass Balance of Inylchek Glacier

To calculate annual mass-balance indices of Inylchek glacier (Fig. 9), the long-term data of total precipitation and mean summer air temperatures at the Tien Shan meteorological station were extrapolated up to the elevation of the equilibrium-line position (4476 m). Annual precipitation was calculated by Equation (5), annual ablation by Equations (6), (8) and (9). Assessment of annual melt W_e.l) has been calculated by

where m is number of days when melt occurs under the certain state of Equations (8) and (9). We assumed loss by evaporation was compensated by condensation and a consistent value of refrozen meltwater equaled 90 mm a⁻¹. There are cycles of positive and negative deviations in glacier mass-balance indices (Fig. 9). From 1940 to 1953 and from 1973 to the present, the glacier had a negative mass balance. For Inylchek glacier, the average weighted mean annual accumulation is 752 mm, the snow-ice ablation average weighted mean is 1070 mm and the index of net glacier-mass change is negative, −318 kg m⁻² a⁻¹, indicating degradation of modern Pobeda-Khan Tengry glaciers.

Fig. 9.

Indices of glacier mass balance (a) and their centrally weighted moving-average values (b) of Inylchek glacier, central Tien Shan.

Conclusions

Our analysis extends understanding of the climatic, meteorological and hydroglaciological conditions around the continental glacial Pobeda-Khan Tengry massif, one of the largest in the world.

Four major synoptic situations have been identified, with an average 4 day duration, which control regimes of temperature, radiation balance and melt. In the accumulation area, weather conditions have a smaller effect and a considerable role is determined by internal processes affecting a local water exchange. The development of internal processes is promoted by the high moisture content of air masses coming mostly from the west. At the same time, our observations indicated that dust is delivered by air fluxes from central Asian deserts during the development of föhn.

Radiation contributes more energy for snow and glacier melt than the combination of all other forms of heat transfer and amounts to more than 90% of the input in heat. Melt is most intensive during periods of anticyclonic weather with föhn development. Only at these periods did turbulent exchange and condensation occur. Föhns bringing large amounts of dust from Tarim affect the radiation balance and cause the development of local cloudiness and the formation of precipitation. Because the melt regime is determined there mainly by solar radiation, it was preferable to calculate the melt and runoff through the solar-radiation data, or to use air temperature associated with different weather patterns. Evaporation and condensation are either mutually compensatory or of negligible effect during other synoptic situations. Spatial distribution of ice-formation zones yielded the upper boundary of liquid runoff at 5200 m.

Accumulation processes were associated with cold cyclonic weather. Estimates of annual accumulation were obtained from 4150–6300 m and increase the amount of information about precipitation distribution in the accumulation areas of the Tien Shan sub-continental glaciers. Spatial distribution of ice-formation zones revealed that the recrystallization zone is above 5800 m. Four ice-formation zones have been revealed in the southern Inylchek glacier accumulation area.

The net glacier mass change is negative, −318g m⁻² a⁻¹, indicating degradation of modern Pobeda–Khan Tengry glaciers.