Introduction

An important role of the Asian continent in the global climatic system has been recognized. The Xizang (Tibet) Plateau has been considered to be an elevated heat source which contributes to seasonal reversal of air circulation (Luo and Yanai , Reference Luo and Yanai1983; Reference Luo and Yanai1984; He and others, 1987). Also, Eurasian snow cover was found to have a close correlation with the strength of Indian monsoon rainfall (Hahn and Shukla, Reference Hahn and Shukla1976). It was further suggested that the El Nino-Southern Oscillation phenomenon was intimately associated with the snow cover in Asia (Yasunari, Reference Yasunari1987; Barnett and others, Reference Barnett, Dümenil, Schlese and Roecker1988). The role of the plateau is of great importance to the global climate.

Meteorological data for the plateau, however, are very few. Climatic information may be obtained by analyzing ice-core samples from many glaciers and ice caps in the area (Thompson and others, Reference Thompson, Wu, E and Xie1988).

Field Activity

The West Kunlun Mountains are located on the northwestern edge of the Tibetan Plateau (Fig. I). The Chongce Ice Cap in these mountains extends over a distance of about 7 km, having two noticeable peaks 6530 m and 6374 m high (Fig. 2). The elevation of the terminus is about 5800 m. A verage thickness, measured by a pulse modulation radar, was about 70 m near the lower peak and 100-150 m around the midstream of the ice cap (Zhu, Reference Zhu1989).

Pit studies were made at seven sites, A3, AS, B4, B8, B8' (adjacent to B8), BI2 and BI3 for stratigraphic observations and for sampling (Ageta and others, Reference Ageta, Zhang and Nakawo1989). Shallow cores of 10 m, 23 m and 32 m were obtained at B13, B12, and B8 respectively (Nakawo and others, Reference Nakawo, Nakayama, Kohshima, Nishimura, Han and Zhou1989). Snow and ice samples from the pits and the cores were mostly melted in situ and shipped for further analyses.

Fig. 1. Locations of the Chongce Ice Cap (lat. 35°14' N; long. 81 °07' E) in the West Kunl un Mountains.

Results

The stratigraphy, oxygen-isotope content, liquid conductivity, and solid particle content are shown in Figure 3 for snow or ice layers near the surface deposited in two years (July 1985-August 1987), identified by snow-stake measurements, at AS during the period. Partic le contents are presented as weights of particles of a diameter large r than 5 /lm, collected through filte ring. The surface layer consisted mostly of firn layers with relatively thin ice layers in between. Three horizontal dirt bands were visibly identified as shown in the figure. Two peaks can be seen in the 6180 profile with minimum values at around depths of 0, 0.5 and 0.8 m, although the amplitude decreased with depth. Also, there are cyclic varIatIOns with a period of about 0.4 m in both conductivity and solid particle content, with an additional conductivity peak at about 0.6 m.

Fig. 2. The Chongce Ice Cap and the observation sites. ABC: Advance Base Camp (after Chen and others, Reference Chen, Wang, Liu and Gu1989).

The values of S180 for summer precipitation of newly deposited surface snow in 1989 were about -2 to -10%0, larger than average 6 values for glacier ice or values for glacier run-off water, which would roughly equal the annual average. This indicates that winter precIpItation should show a smaller 6 value than summer. The relatively low values in the 6 profile in Figure 3, at around 0.5 and 0.8 m, hence, would correspond to winter layers, since the 0.8 m thick layer covered two full years.

Fig. 3. Stratigraphy, 6180, liquid conductivity and particle content (given in weight of particles larger than 5 /Lm) in the two full year layer (July 1985 to August 1987) at A5 (6180 m a.s.1.).

Year-round precIpItation data are not available at the West Kunlun Mountains. At meteorological and hydrological stations around the mountains, it has been shown that precipitation takes place mostly in the warm season (May to September), with drought from fall to spring (Kang and Xie, Reference Kang and Xie1989). The layers with relatively large 6 values (summer) are much thicker than those with small 6 values (Fig. 3): the summer balance is larger than the winter balance on the ice cap. Relatively high conductivity seems to be associated with the layer formed in the dry season. High concentration of particles or dirt layers would also form in winter when the precipitation is poor. These are encouraging, since annual layers would possibly be identified by analyzing δ¹⁸O, conductivity and particle content.

Fig. 4. Characteristics of surface snow/ ice layer at Bl3 (6366 m a.s.l.) near the summit of the lower peak of the ice cap. (a): Stratigraphy, S¹⁸O, liquid conductivity, and particle content (given in weight of particles larger than 5/Lm and in number of microparticles (I < ϕ < 5μm)). (b): Cations ratio to chlorine. The scale at the top is for the ratio to sea water, and the bottom scale is for the ratio to lake water.

Figure 4a shows profiles of these variables at Β13, around the lower peak of the ice cap (Fig. 2). Particle content is given by the weight of large particles, and by the number of micro-particles of diameter smaller than 5 μm. The variation of δ¹⁸O becomes less intense with increasing depth, and no periodic change can be found at greater depths where continuous ice layers exist. This is probably due to the homogenization of the heavy isotope caused by the refreezing of melt water in the surface snow/firn layer.

Profiles of conductivity and large particle content are very similar to each other (Fig. 4a). Three peaks appear in each curve, indicating the depths of winter layers, as discussed above, even at the greatest depth where no noticeable peaks were found in the δ curve. It should be noted that the locations of peaks of particle content are not always compatible with the site of visible dirt bands, although Thompson and others (Reference Thompson, Wu, E and Xie1988) reported the existence of annual dust layers in the Dunde Ice Cap, Qilian Shan, China.

Figure 4b shows variations of cations at the same spot. The upper scale gives the ratio of each cation to Cl divided by the ratio to Cl in sea water. The lower scale is the ratio to Cl with the reference of lake water around the mountains (Fushimi and others, 1988). Figure 4b seems to suggest that precipitation in the mountains could be mostly of local origin, i.e. from evaporation in and around the area (Nakawo and Takahara, Reference Nakawo and H1988), since the ratio of the cations to Cl in the core samples is very close to the ratio found in lake waters.

The profiles of Mg and Ca are very similar to each other; three noticeable peaks are found at the same depths where conductivity and particle content exhibited maxima. It is interesting to note that the variation of Κ is inversely correlated with the change of Mg and Ca: peaks in Κ profile are found at depths where minimum values are obtained for Mg and Ca, and vice versa. In other words, Mg and Ca increase in winter and decrease in summer, while Κ decreases in winter and increases in summer. It is considered, therefore, that analyses of these cations would also enable us to estimate annual layers, and hence the climatic history around the area.

Acknowledgements

The analyses of cations and micro-particle content were carried out at the National Institute of Polar Research, for which the authors are grateful to Dr. Y. Fujii. Field activities were financially supported by the Monbusho International Scientific Research Programme (No. 62041043, 63043030). Laboratory analyses were supported by the Special Coordination Funds for Promoting Science and Technology.