Introduction
Studies of glacier water resources for agricultural water supply were first carried out in 1958 in western China. Glaciers in the Qilian mountains were investigated and a field station was established between 1958 and 1962 at
Lauhugao Glacier in the western Qilian mountains. Later, a station was established at Ürümqi Glacier No. 1 in the eastern Tien Shan mountains (1959 to 1966, 1980 to present). This is the only station at which long-term multiscientific measurements have been achieved at an alpine glacier in China. Based on hydrometeorological measurements at permanent and semi-permanent stations, and on glacier inventories, estimation of glacial water resources in China was carried out in the 1980s. This paper attempts to use the regional trend of specific discharge from glaciers to estimate such runoff where there are no direct measurements, and to describe the compensating effect of glaciers on runoff in mountain streams.
Basic characteristics of different types of glaciers in the mountainous areas of western China
Glaciers in China
Excluding the polar regions, glaciers cover 2.24 x 105 km2 of the Earth’s surface. About 26% (5.87 x 104 km2) of their area is located in China. Thousands of small glaciers occur over an extensive area that spans about 2700 km from east to west (103°45′ to 73°55′E), and about 2400km from south to north (27° to 49° N). All are found in high mountains and plateaus in western China, and most are valley glaciers. The glaciers can be classified into three types, depending on climatic, thermal and physical characteristics. Most glaciers belong to the continental type, while sub-continental and maritime types occupy only a small area. The continen-tal-type glaciers occur in arid and semi-arid climatic regions which experience lower air temperatures and receive less precipitation than the maritime-type glaciers which are located in areas with moist climate, with abundant precipitation nourishment and higher air temperatures. The flow velocity of continental-type glaciers, in general, is much lower than that of the maritime type (Reference Shi and ZichuShi and Xie, 1964; Reference Shi and JijunShi and Li, 1981; Reference Li and ZhenLi and Zhen, 1982; Reference Huang and ZuozheHuang and Sun, 1982; Reference Zichu and ZhenXie and Zhen, 1982; Reference Ren and ShiRen, 1988) and the yield ofglacier meltwater runoff from continental-type glaciers is much smaller than from maritime-type glaciers (Reference YangYang, 1991) (Table 1).
Glacier Ablation
Heat sources for glacier melt
The main heat source for glacier ablation is net radiation, accounting for 60.5 to 92.1% of the total budget. Sensible heat can contribute from 6.6 to 35%. The latent heat is usually considered to be small Reference Zeng and Guangrong(Zeng and Dong, 1966; Reference Zeng and YouguanZeng and Kou, 1975; Reference Xie, Meishang and ShiXie and Cao, 1965). The composition of heat budgets has an obvious regional pattern. The radiation heat flux increases with increasing aridity, decreasing latitude and increasing altitude. For sensible heat, the opposite applies, because the influence of moist air from southeastern and southwestern monsoons becomes more important, and the percentage contribution of sensible heat increases (Table 2). Different glacier surface types (firn, ice-snow and debris-covered)
The composition of heat balance in various glaciers during the ablation season
influence heat budgets through different albedos (Reference Zeng, Meisheng and XuezhiZeng and others, 1984).
The relationship between glacier meltwater and radiation balance can be expressed as
where Q is the discharge from an experimental plot of ice surface (m3s−1), QB is the radiation balance value (MJ m−2 d−1 ), a is an empirical coefficient and n is an empirical exponent.
The relationship between ablation and air temperature
Although Equation (1) provides a useful relationship for estimating glacier meltwater runoff, the value of QB is difficult to obtain. Air temperature is often used as an index of energy balance for establishing an empirical relationship to estimate ablation of ice. The synthetic empirical equation can be approximated as follows:
where A is glacier ablation rate (mm d−1), T is air temperature (°C) at the median elevation of the glacier, and ϸ is a coefficient reflecting the influence of climatic conditions in different regions (Reference YangYang, 1981):
in which b = QB/QT, where QB is the radiation balance and QT the total heat received during the melt season. It is similar to the method of estimating glacier ablation at the equilibrium line, proposed by Reference KhodakovKhodakov (1975). Where there is no direct measurement of air temperature on the glacier surface, air temperature gradient can be used and an approximate relationship is:
T0 is the air temperature (°C) at the meteorological station, H is the altitude difference (m) between the meteorological station and the median elevation of the glacier, dT/dH is the air temperature gradient (°C100 m−1), Jt allows for the increase of air temperature gradient with elevation (Reference KhodakovKhodakov, 1975) and L is the glacier length (km).
Some estimates of ablation shown in Equation (2) are in close agreement with the results of measurements from ablation stakes (Reference YangYang, 1991).
Glacier Melt Runoff
Characteristics of glacier melt runoff
For more then six months, air temperatures in high mountain areas remain below 0°C, until late spring or early summer, when they rise above 0°C. Snow on the glacier surface begins to melt in late April or early May on continental-type glaciers, and in early April on maritime-type glaciers. The melt is weak, and the melwater is absorbed in the snow layers. During the cold nights, the meltwater is refrozen and surface runoff is arrested. As the season advances, the melt increases to sustain continuous runoffon the glacier surface, forming a network of channels. In small glaciers of the continental type, such as in Ürümqi Glacier No. 1, the meltwater moves along the major channels which enter mountain streams directly. However, in larger glaciers, or maritime-type glaciers, much of the meltwater disappears through cracks and moulins, to emerge from under the terminus in one or more large streams. This outflow is sometimes maintained even during winter (Reference Yang and ShiYang, 1988). Quntailan Glacier in Tien Shan (subcontinental-type) and Gongga Glacier in Hcnduan Shan (maritime-type) are such examples.
Synchronous co-variation between air temperature and meltwater runoff is more significant in continental-than maritime-type glaciers. Analysis shows the relationship between discharge and air temperature when there is no precipitation can be given as follows
where Q is the glacier melt discharge (m3 s−1), T is mean air temperature (°C) on the glacier during the ice-melt period , and α, β, a, b, and c are empirical coefficients (Reference KangKang, 1983; Reference Hu and NianjieHu and Li, 1989; Reference Yang and ShiYang, 1988).
The compensating effect of melt runoff
Most mountain streams in western China receive glacier meltwater. The percentage of glacier melt in runoff increases with increasingly arid climate. For example, in the internal drainage rivers of the Heshi region, from east to west, the percentage ofglacier melt in runoff increases from 4 to 32%. During dry warm summer periods with scarce precipitation, large percentages of glacier meltwater augment low flow. When abundant precipitation with low air temperature occurs, generally it has a negative influence on glacier melt runoff. In streams dominated by glacier melt runoff, the coefficient of variation of mean annual runoff (Cv = 0.10 to 0.20) is smaller than that of rain- or snow-fed streams (Cv = 0.20 to 0.45).
Glacier Meltwater Resources
The main problem in estimating the amount of glacier meltwater runoff is the lack of hydrometeorological data. Specific runoff from glaciers Mg(l s−1 km−2) has been obtained from several glacierized areas in China. As it has an obvious regional regularity, Mg can be estimated from glacierized areas without direct measurement as follows:
where K is the modification coefficient of ablation area, K=fg/fg0 fg s the percentage of ablation area for the glacier where the specific runoff is to be determined, fg0 is that for the glacier where the runoff (Mg0 has been measured, dH0 is the difference in the elevation of the equilibrium lines between the two glaciers, dT/dH is the air temperature lapse rate, usually equal to 0.65°C per 100 m, dT′ is the air temperature modification value owing to the climatic difference between two glaciers at the same elevation. dMg/dT is the increment of Mg with air temperature, empirically found to be 5.01s−1 km−2 per °C.
For two glaciers located not only in the same climatic region, but also having similar equilibrium line altitudes, the equation reduces to
The difference in Mg is caused only by the difference in the percentage of ablation area.
Using Equation (8) the total volume of glacier meltwater (m3) produced during the ablation season is
where t is the melt period, about 150 days for continental-type glaciers, 180 days for subcontinental-type and 210 days for maritime-type glaciers. Fs is the glacier area (km2).
The mean annual amount ofglacier meltwater runoff in western China is estimated to be about 56.4km3 or 5.64 x 1010m3 (Table 3). It is about 10% of the total amount of runoff from all sources in the four provinces in western China.
Glacier areas and annual meltwater runoff
