Knowledge on impacts of water variability on chemical river water quality at the World largest closed drainage basin is a critical aspect in improved management of water resources during a time of persistent droughts (Agrawala et al., 2001), increasing water demand and potentially decreasing water supply of central Asia. Stream water quality is affected by the load applied to the river, water temperature and streamflow volumes, affecting both concentrations and total loads. Our interest focuses on the Tien Shan and Pamiro-Alai highlands with glacier distribution, seasonally snow covered regions and with a thermal state of sporadic and discontinuous permafrost that supply of water to central Asia the home to the vast majority of the Earth's human population which is also very sensitive of aquatic ecosystems and human health to changing climates. The drying of Central Asia in the 20th century has captured a large share of the attention afforded to environmental problems in the Aralo-Caspian basin (Kira, 1995, Vaganov, 1998; Voropayev, 1997, Glantz, 1999). The decrease in accumulated snow (Aizen et al., 1997) with its concomitant decrease in snow melt contribution to river runoff, increased permafrost and changed glacier melting in the heads of river basins exhibit the significant variability in river stream flow and water temperatures.

The main objective of our research is an identification of regional chemical river water quality responses to river runoff fluctuations in the continental alpine watersheds during the 20th century. We distinguish whether associations exist, and evaluate the magnitude and the nature of the relationship. The probabilistic models of regional stream chemistry response to hydrological variability in high-altitude watersheds is implemented in this study and help to process the methods for predicting future river water quality.

Evaluation of long-term changes in stream flow, water temperature and solute content of surface water in the Tien Shan and Pamiro-Alai watershed was done based annual monthly data collected from gauging stations for over twenty years. The analysis was done for the basins with and without glacierized areas and permafrost distribution.

To pick up an anthropogenic component in long-term ionic changes, we took into account the ratio between the river runoff for two periods with human impact and without it. Besides, we calculated the ratio between ionic concentration of HCO3- and concentration other ions as criteria of human impact on stream chemistry assuming only natural variability in the hydrocarbon. Statistical analyses did detect associated changes of river runoff, their different components (glacier, snow and permafrost), water temperatures and its solute content. Separation of anthropogenic and natural variability revealed increased concentration of SO2-, Ca2+ and Na+ + K+ ions since the middle of 1970 occurred simultaneously with decreasing river runoff and intensive agricultural development in Central Asian.