YongChao Lan , ZhengYao Ma , YongPing Shen , ChengFang La ,Jie Song , XingLin Hu 1,, HongWei Din

1. Key Laboratory of Ecohydrology in Inland River Basin, Chinese Academy of Sciences, Lanzhou, Gansu 730000, China

2. Cold and Arid Regions Environmental and Engineering Research Institute, Chinese Academy of Sciences, Lanzhou, Gansu 730000, China

3. Hydrology and Water Resources Bureau of Gansu Province, Lanzhou, Gansu 730000, China

4. Hydrology and Water Resources Bureau of the Upper Yellow River Headwater of the Yellow River Conservancy Committee,Lanzhou, Gansu 730030, China

5. Institute of Hydrogeology and Engineering Geology, Gansu Bureau of Geology and Mineral Exploration and Development,Zhangye, Gansu 734000, China

Sensitivity of mountain runoff to climate change for Urumqi and Kaidu rivers originating from the Tianshan Mountains

YongChao Lan1,2*, ZhengYao Ma3, YongPing Shen2, ChengFang La4,Jie Song4, XingLin Hu1,3, HongWei Din1,5

1. Key Laboratory of Ecohydrology in Inland River Basin, Chinese Academy of Sciences, Lanzhou, Gansu 730000, China

2. Cold and Arid Regions Environmental and Engineering Research Institute, Chinese Academy of Sciences, Lanzhou, Gansu 730000, China

3. Hydrology and Water Resources Bureau of Gansu Province, Lanzhou, Gansu 730000, China

4. Hydrology and Water Resources Bureau of the Upper Yellow River Headwater of the Yellow River Conservancy Committee,Lanzhou, Gansu 730030, China

5. Institute of Hydrogeology and Engineering Geology, Gansu Bureau of Geology and Mineral Exploration and Development,Zhangye, Gansu 734000, China

The mountain watersheds of Kaidu River and Urumqi River, which separately originate from the south and north-side of the Tianshan Mountains in Xinjiang, are selected as the study area. The characteristics and trends on variation of temperature, precipitation and runoff, and the correlativity between temperature, precipitation, and runoff were analyzed based on the past 40 years of observational data from the correlative hydrological and weather stations in the study areas. Various weather scene combinations are assumed and the response models of runoff to climate change are established in order to evaluate the sensitivity of runoff to climate change in the study areas based on the foregoing analysis. Results show that all variations of temperature, precipitation,and runoff overall present an oscillating and increasing trend since the 1960s and this increase are quite evident after 1990. There is a markedly positive correlation between mountain runoff, temperature, and precipitation while there are obvious regional differences of responding degree to precipitation and temperature between mountain runoff of Urumqi River and Kaidu River Basins.Also, mountain runoff of Urumqi River Basin is more sensitive to precipitation change than that of Kaidu River Basin, and mountain runoff of Kaidu River Basin is more sensitive to temperature change than that of Urumqi River Basin.

south slope; north slope; Tianshan Mountains; Kaidu River; Urumqi River; climate change; sensitivity

1. Introduction

It is expected that the earth’s surface average temperature will increase by 1.5–4.5 °C in the 21st century, according to the fourth science evaluation report presented by IPCC (2007).The future of Northwestern China and other regions such as Tibet, Yunnan, Sichuan and so on, may include higher summer temperatures and unpredictable change of precipitation(Qin, 2002; Qinet al., 2005; Xuet al., 2005).

The Xinjiang region is located in the western portion of Northwestern China, in the Eurasian heartland, far from any oceans. The average annual precipitation is less than 200 mm in most areas of Xinjiang, but precipitation in the Tianshan Mountains is relatively abundant because the east-west trend of the mountains blocks the uplifting air currents from the Atlantic and Arctic. Precipitation in the north-side of the Tianshan Mountains may reach 500–700 mm, to 1,000 mm in some windward slopes at west portion of the north side.Therefore, the Tianshan Mountains are called "Wet Island"within the hungriness areas of Northwestern China. Sixty five percent of the rivers in Xinjiang originate from the mountains whose runoff accounts for 54% of the total runoff (Xinjiang Geographic Society, 1993; He and Guo, 1996; Water Conservancy Bureau of Xinjiang and Water Conservancy Society of Xijiang, 1999; Yuanet al., 2003). Also, the mountains have abundant extant ice, glaciers, firn in alpine belts, and precipitation (Huet al., 2003; Denget al., 2005). Thus, the north and south side of the Tianshan Mountains have become one of the most successful and developed in economy in arid regions of China, and the world. Melt water from glaciers and snow is an important component of mountain runoff to rivers originating from the north and south-side of the Tianshan Mountains and global warming has produced a significant influence on the water cycle process since the 1980s. Once a large fluctuation of the runoff from the Tianshan Mountains occurs, the social and economic development of the Xinjiang region will be unusual strongly influenced. Therefore,analyzing the impact of global warming on mountain runoff of rivers originating from the north and south-side of the Tianshan Mountains will benefit water science development,and provide a framework when putting forward countermeasures of adapting and mitigating the impacts of climate change. Currently, numerous research achievements have occurred in relation to the influence of climate change on water resources of the Xinjiang region. Shiet al.(2003)noted that in 2003, under the influence of west wind circulation, weather transformed from warm-dry to warm-wet in western China, in particular in the Xinjiang region. Evidence of this transformation can be found in an increase in precipitation and glacier melt water which exceeds evaporation quantity in mountainous areas. This leads to increase mountain runoff and a significant rise in inland lake levels. Chen and Xu (2004) discussed the correlations between water resource change in the Tarim River Basin and global climate change based on hydrology and weather data of 77 meteorological stations over the past 50 years. The long-term trend of the hydrology time sequences including temperature, precipitation, and runoff is realized by the mean of parameter examination and non-parameter examination. Wanget al.(2003)analyzed changes of temperature and precipitation over the past 40 years in the mountainous area of the Tarim River Basin. They noted that the 1990s was the warmest decade, where the extent of humidity increase was the largest in the middle and western mountainous areas of the Tianshan Mountains and the Pamirs. Wanget al.(2005) analyzed the variation characteristics of air temperature, precipitation, and runoff in the Aksu region. They concluded that the oasis climate became warmer and wetter, while the gobi and desert climate became slightly drier, and mountainous runoff increased slightly due to air temperature rise and mountainous precipitation increase that lead to glacier melt increased. Jianget al.(2005, 2007) and Jiang and Xia (2007) analyzed the secular variation characteristics and annual distribution law of river flow that have different supply sources in the Asku River Basin. Results show that mountainous runoff has been increasing since 1990 because air temperature and precipitation are in an upward trend, snow and glacier melt water has been increasing, and the influence of rising air temperature is more important than increasing precipitation. Wuet al.(2006) studied climate change and the relation between climate factors and glacier melt water and rainfall runoff in the Urumqi River Basin. They pointed out that melt water is chiefly influenced by air temperature changing in the glacier zone of high mountainous areas. Most of the aforementioned research mainly involved changing air temperature, precipitation and mountain runoff in the southern and northern slopes of the Tianshan Mountains and limited qualitative analysis relating to the influence of climate change on water resources. Unfortunately, quantificational analysis relating to the influence of climate change to water resources in the region is lacking. Thus, we selected the Urumqi River and the Kaidu River as representative rivers from the southern and northern slopes of the Tianshan Mountains in Xinjiang,and analyzed the variation characteristics and trends of mountain runoff, precipitation and air temperature in the two river basins based on hydrological and meteorological observation data from stations in the mountain areas. We present a response model of mountainous runoff to climate change according to the correlation between mountain runoff, air temperature, and precipitation. The influence of climate change on mountain runoff of the representative rivers is quantified by combining the response model and the sensitivity analysis formula that is used to reflect the response’s degree of runoff to air temperature and precipitation changes,assuming various future climate change scenes. The results can be used to reveal the differences between response of mountainous runoff to climate change for the rivers in the southern and northern slopes.

2. General situation of the study areas

2.1. Kaidu River Basin

The Kaidu River Basin is located in the north fringe of the Karasahr Basin, between 82o52′E–86o55′E and 41o47′N–43o21′N. Headwaters of the Kaidu River are from the Hargat and Zhasit trench of Mount Sarbin, middle of Tianshan Mountain, elevation of 5,000 m. The catchment area above the mountain exit (Dashankou station) is 19,022 km2, mean elevation of 3,100 m, and total extant glacial area of 473.97 km2. The annual glacial melt water volume is 5.14×10 m3accounting for 15.2% of the annual mountain runoff of the Kaidu River (Lanzhou Institute of Glaciology and Geocryology, 1987). The mountain runoff of Kaidu River is recharged by rainfall, groundwater and glacial-snow melt water from mountain ranges such as the Large and Small Yultuz basins, and valley wetlands. The Kaidu River has a total length of 530 km, its headwaters connect with the Bayinbuluk Wetlands, and ends at Bosten Lake. The total catchment area is 22,314 km2, and its annual mountain run-off volume is 34.2×108m3, which accounts for 85% of annual inflow of Bosten Lake (Wanget al., 2006). The flood season of the Kaidu River begins from April, and its peak flood season is from June to August.

2.2. Urumqi River Basin

The Urumqi River Basin is located in the middle north-side of Tianshan Mountain between 86°45'E–87°56'E,43°00'N–44°07'N. The basin borders the Tuen River Basin to the west and the Banfanggou Basin to the east whose total watershed area is 4,684 km2, and the mountain area above Hero Bridge is 924 km2(average elevation of more than 3,083 m). Urumqi River originates from Kalawucheng Mountain located at the north-side of Tianshan Mountain and flows northeast. Through the mountain outlet, the river flows to Urabo Lake, then turns north through Urumqi city,and finally disappears at Miquan County. The highest point of the basin is the peak of Tianshan Mountain, Tiangel II whose elevation is 4,479 m, and the lowest point is at the river mouth of the West White Polar Ditch whose elevation is 1,670 m. The largest difference between the highest and lowest point of the mountain basin is 2,890 m. There are 155 modern glaciers in the alpine zone of the basin, whose total area is 48.04 km2. Across mountain outlet, the river flows into the Urabo, then it turns north through Urumqi city, and finally disappears at Miquan County located at south of the Junggar Basin. The rivers’ total length is 214.3 km and the length above the mountain outlet is 62.6 km.After flowing out of the mountain outlet, the runoff of the river is introduced into reservoirs. There are four reservoirs in Urumqi area, whose total water storage is 1.8×108m3.The Urumqi River Basin has an obvious vertical zonation.Landscape distribution of the basin is as follows: (1) alpine gibben desert zone (above 3,600 m); (2) alpine meadow and steppe vegetation zone (3,600–2,600 m); (3) low mountain forested zone (2,600–1,600 m). Climate sub-regions of the basin is as follows: (1) alpine glacier and snow zone, also called living glacier zone, with 124 glacier and mean snowline altitude is 4,050 m, and the area above the snowline is 102.2 km2. There the elevation of the glacier tongue end is 3,440–4,050 m, the annual mean temperature is -6.0°C, and snowfall accounts for more than 75% of annual precipitation; (2) subalpine permafrost region, where the annual mean temperature ranges from -2.5 to -1.22 °C and snowfall accounts for 50% of annual precipitation; (3)middle-high-mountain cold temperature zone, where the annual mean temperature is 0–4 °C and snowfall accounts for 20%–30% of annual precipitation, generally 400–500 mm, and the zone has a greatest annual precipitation in three climate sub-region. The long-term mean runoff is 2.40×108m3in the mountain basin above the Hero Bridge hydrological station (Wuet al., 2006).

3. Basic data and analysis methods

The Dashankou and Hero Bridge hydrological stations are respectively control stations of mountain runoff, and the Daxigou and Bayinbuluk weather stations are respectively regional representative weather stations of the mountainous areas of Kaidu River and Urumqi River. Runoff observational data at Dashankou and Hero Bridge hydrological stations, and precipitation and temperature observational data at Daxigou and Bayinbuluk weather stations, from 1959 to 2005, was selected for analyzing climate and runoff changes in the mountain areas of Kaidu River and Urumqi River. All chosen data have been examined by the Xinjiang Water Hydrology and Resources Bureau and Xinjiang Meteorological Bureau. The hydrological and meteorological stations and observational data series in the study areas are listed in Table 1. Analysis methods such as the linear trend analysis, and multiple linear and nonlinear discriminate analysis, are used to analyze the change trends and characteristics of hydrological and meteorological elements in the study areas, and the sensibility of runoff to climate change (Wanget al., 2003;Liet al., 2004).

Table 1 Hydrological and weather stations, and data sequences in the study areas

4. Changes of climate and runoff in the study area over the past 40 years

It can be observed in Table 2 that temperature change in the study areas corresponds with global warming over the past 40 years, showing an overall oscillating and rising trend, with a remarkable increase of mean temperature in the 1990s. The change of precipitation has an obvious space and region diversity. In respect to interdecadal change of precipitation, each decadal mean precipitation before 1990 is basically less than the precipitation mean in the 1960s, and precipitation after 1990 is larger than the precipitation mean in the 1960s in the source region of Kaidu River. Precipitation at the mountain exit of Kaidu River has been increasing since the 1960s, and the increase rate of precipitation after the 1990s is larger than 50% of mean precipitation in the 1960s. The precipitation in the mountain watershed of Urumqi River had been continually increasing and reached a maximum in the 1990s. Corresponding with temperature and precipitation increasing,overall mountain runoff for Urumqi and Kaidu rivers also have increased, but mountain runoff of Urumqi River after 2000 decrease rapidly, falling below the mean of the 1990s.

Table 2 Variations of each hydrological and climatic factor at various ages in the study areas related to the mean

4.1. Temperature change

Mean temperature in the study areas have been obviously rising since the middle or later period of the 1980s. The annual lowest, highest and mean temperature presented a rising trend over the past 40 years, where climatic inclination rates are 0.21 (°C/10yr), 0.26 (°C/10yr) and 0.18 (°C/10yr), respectively. The rising rate of lowest temperature is larger than that of highest temperature, which is an asymmetrical change trend. Rising temperatures in winter and nighttime is the main reason for rising temperatures in the study areas. Our analysis shows that temperatures from the 1960s to the 1970s are lower, and the 1980s is a turning period of temperature change,while temperatures in the middle mountain zone started to rise,except for the alpine region. Temperature increase from the 1990s to the beginning of 2000s is quite remarkable, and this period is a relatively high temperature period in the past 40 years.

4.2. Precipitation change

Precipitation from the 1960s to 1980s in the study areas is less than the long term mean, and the 1980s is a turning point for precipitation change, where precipitation in the earlier stage is less, and obviously increases in the later stage.The 1990s is a period with the largest increase of precipitation except in the mountain exit (Hero Bridge station). The increasing degree of mean precipitation in the mountain basin exit of the Kaidu River (Dashankou station) is larger from the 1970s to the beginning of 2000s where the increasing degree is 56.5%. The mean precipitation in the source region of the Kaidu River is normal in the 1960s,then decreased during the period from 1970s to 1980s, and continuously increased after 1980s.

4.3. Mountain runoff change

It can be observed in Figure 1, the change process of the annual departure of mountain runoff of Urumqi River can be roughly divided into two stages, with 1986 as the pivotal point. The first stage is from the mid 1950s to 1986 where runoff is usually less than normal. Runoff continues to decrease from the mid 1950s to mid 1970s, and then bottoms out after the mid 1970s to mid 1980s. 1986 is a turning point in which runoff changes from less than to greater than normal. The second stage is from 1986 to 2004 where runoff is usually larger than normal. Runoff increases from 1986 to the end of the 1990s, then falls below normal after 2000.Mountain runoff of Kaidu River is different from that of Urumqi River. Here, runoff is larger than normal from 1955–1959, then falls below normal from 1959 to the late 1960s, then rises above normal from the late 1960s to early 1970s. Then for a period of 20 years, from mid 1970s to early 1990s, runoff is below normal. Runoff then increases rapidly in the mid 1990s, reaching a maximum in 2001, then falls rapidly after 2001 but is still above normal. Analyzing the linear trend of the yearly change process, the linear change rate of mountain runoff for Urumqi and Kaidu rivers are 3.5%/(10yr) and 2.2%/(10yr), respectively, which shows that overall, mountain runoff in the study areas presents a fluctuating and increasing trend.

From the inter-decadal change characteristics in Table 2,mountain runoff of Urumqi River begins to increase from the 1960s, reaching a maximum in the 1990s, and then falls rapidly after 2000. For Kaidu River, runoff begins to decrease from the 1960s, reaching a minimum in the 1980s.Runoff then begins to increase rapidly after the 1980s,reaching a maximum during the period of 2000 to 2005.

5. Sensibility analysis of mountain runoff to climate change

5.1. Definition of runoff sensitivity to climate change

The sensitivity of runoff to climate change means the degree of response of runoff to climate change (IPCC, 1996).To illustrate the sensitivity of runoff to climate change, the various climate change scene combinations were assumed and they were composed of the various precipitation changing scenes (increasing 20%,10%, 0%, -10%, -20% and so on) and temperature rising scenes (rising 0 °C, 0.1 °C, 1 °C and so on). The response of runoff to the various climate change scenes can be expressed by the following mathematical equation:

where,RP,Tis the current runoff,RP+ΔP,T+ΔTis the runoff under the combination of the certain precipitation variable quantity ΔP(%) and the certain temperature variable quantity ΔT( °C); ΔRΔP,ΔT(%) is increment of runoff fromRP,TtoRP+ΔP,T+ΔT, or the relative variable quantity of runoff when precipitation and temperature change from sceneP(%) andT(°C) toP+ΔP(%) andT+ΔT(°C).

But what should be emphasized in the sensitivity analysis is that the assuming climate change scenario does not change the temporal and spatial distribution of historical climate, but only reproduces the scaling series of the hydrological and meteorological factors in the future. Under the same climate change scenario, the larger the responsive degree is, the more sensitive the response of hydrological element to climate change is. Conversely, the smaller the responsive degree is, the less sensitive the response of hydrological element to climate change is. Sensitivity analysis can provide important information about climate change and has a certain effect for finding out the mechanism and difference of responsive degree of hydrological element in the various river basins. Sensitivity analysis of the runoff to climate change can determine the main factors and secondary factors affecting runoff change, but it cannot forecast the runoff change under climate change in the future (Wang and Shi,2000; Chen and Wang, 2004).

Figure 1 Yearly variation processes of annual departure of mountain runoff for Kaidu and Urumqi rivers

5.2. Response model of runoff to climate change

Analysis shows that there is a close relationship between mountain runoff, precipitation, and temperature.Therefore, we can approximately simulate an archetype by establishing a statistic equation (Fu and Liu, 1991; Liu and Tian, 1991; Wanget al., 2003), and the continuous multiplication form of the Power function is used to describe the relationship between mountain runoff, precipitation, and temperature in the study areas and it can be described as follows:

whereR,P, andTis separately average runoff depth, precipitation, and temperature,α,βandκare pending coefficients, andeis a natural logarithm which is set to easy for fitting equation. The above equation displays a nonlinear relationship between water resources system and climate change and it has definite physical significance (Chen and Xu, 2004). Based on observations of annual precipitation,temperature, and runoff depth, the values ofκ,α, andβin the equation can be obtained by means of a statistical software,and the response models for the runoff to climate change in the study areas are established for analyzing the sensitivity of mountain runoff to climate change (Table 2).

The response degree of runoff to climate change is different under different precipitation and temperature condi-tions. The response degree of runoff to different climate change scenes can be obtained by the response model of the runoff in the study areas. Significance test results show the multiple correlation coefficients between annual runoff depth, precipitation, and temperature in the mountain basins of Urumqi and Kaidu rivers in the fitted equation separately are 0.753 and 0.671 and have passed the significant test(Rα=0.011=0.3721), which shows there is a significant positive correlation relationship between mountain runoff, precipitation, and temperature. The responding models of mountain runoff to climate change for Urumqi and Kaidu rivers are listed in Table 3.

Table 3 Responding models of mountain runoff to climate change, for representative rivers located at thesouth and north-side of the Tianshan Mountains in Xinjiang

5.3. Climate change scenarios

It is expected that climate warming will continue in the future and global annual average temperature of China will raise about 1.3–2.1 °C by 2020, 1.5–2.8 °C by 2030, and 2.3–3.3 °C by 2050. The average annual precipitation will increase about 2%–3% by 2020, and 5%–7% by 2050 according to the Fourth Science Assessments Report about climate change issued by IPCC in February 2007, and The National Assessments Report about global warming issued by China Meteorological Administration in January 2006 (Qinet al., 2005; Xuet al., 2005; IPCC, 2007). Based on characteristics and trends of global and regional climatic change, the suppositional climate change scenarios schemes are that the precipitation change scenarios are 0%, ±1%, ±5%, ±10%,+15% and +20% and the temperature change scenarios are 0 °C, +0.5 °C, +1.0 °C, simultaneously, the scenes of temperature change separately are 0 °C, +0.5 °C, +1.0 °C, +1.5°C, +2.0 °C, +2.5 °C in the mountain basin of Urumqi River in the future. Likewise, the precipitation change scenarios are 0, ±1%, ±5%, ±10%, +15% and +20%, and the scenes of temperature change scenarios separately are 0°C, +0.5 °C, +1.0 °C, +1.5 °C in the mountain basin of Kaidu River in the future.

5.4. Sensitivity analysis of mountain runoff to climate change

According to the model in Table 3 and the assuming climate-scenes scheme, the sensitivity analysis results of mountain runoff to climate change in the study areas can be calculated and they are listed in Tables 4 and 5.

Table 4 Sensibility analysis results of mountain runoff to climate change in Urumqi River

Table 5 Sensibility analysis results of mountain runoff to climate change in Kaidu River

It can be observed from Tables 4 and 5 that mountain runoff of Urumqi River will decrease 0.69%, 6.99% or increase 0.69%, 6.77% and mountain runoff of Kaidu River will decrease 0.33%, 3.41% or increase 0.32%, 3.19% while mountain precipitation decreases or increases 1% and 10%under the scene of unchanging temperature, and mountain runoff of Urumqi and Kaidu rivers will increase 0.24%,1.90% and 0.56%, 5.47% while the mountain temperature rises 0.1 °C and 1 °C under the scene of unchanging precipitation. This indicates that the response of mountain runoff of rivers originates from the south and the north slope to climate change both have the consistency and otherness. The consistency is the proportional change of mountain runoff of both rivers to temperature and precipitation. The otherness(differences) is the response degree of mountain runoff of Urumqi River to precipitation change is larger than that of Kaidu River, and the response degree of mountain runoff of Kaidu River to temperature change is larger than that of Urumqi River. Different geographical positions such as elevation and area covered by ice and snow may be the driving factor causing response at different degrees of runoff to climate change. These factors require future discussion.

6. Conclusions

(1) Overall, temperature and precipitation in the study areas have increased over the past 40 years, especially after the 1980s. Also, the change processes and characteristics of precipitation and temperature both have an obviously regional difference.

(2) Changes of mountain runoff in the study areas both have a positive correlation with mountain temperature and precipitation changes, and overall they present an increasing trend under the influence of increasing mountain temperature and precipitation. The increase rate is especially remarkable after the 1980s, and mountain runoff of Urumqi and Kaidu rivers both reaches a maximum in the 1990s.However, mountain runoff of Urumqi River obviously falls after 2000.

(3) Mountain runoff of Urumqi and Kaidu rivers are both quite sensitive to climate change. Comparatively, the sensibility of mountain runoff of Urumqi River to precipitation is larger than that of Kaidu River, and the sensibility of mountain runoff of Kaidu River to precipitation change is larger than that of Urumqi River.

This research is supported by the funding of the Key Laboratory of Eco-hydrology Open Fund, Chinese Academy of Sciences and Knowledge Innovation Program of the Chinese Academy of Sciences, No.KZCX2-YW-328.

Chen LF, Wang HY, 2004. Sensitivity of runoff to climate change in small drainage basins in China. Resources Science, 26(6): 62–68.

Chen YL, Xu ZX, 2004. Possibility influence of global climate change to water resources in the Tarim River basin. China Sciences ser D, 34(11):1047–1053.

Deng MJ, Wang SJ, Dong XG, Yu HM, 2005. The Water Resource in Xinjiang and its Sustainable Utilization. China Water Conservancy and Hydropower Press, 150–167.

Fu GB, Liu CM, 1991. Computation and analysis on influence of global warming to area water resource. Acta Geographica Sinica, 46(3): 277–288.

He XL, Guo SL, 1996. The mechanism of abration a material anti-abrasion.Water Resources and Water Engineering, 7(3): 8–15.

Hu RJ, Jiang FQ, Wang YJ, 2003. Assessment on the glacial water resources in Xinjiang, China. Arid Zone Research, 20(3): 187–188.

IPCC (Intergovernmental Panel on Climate Change), 1996. Climate Change 1995. Impacts, Adaptation, and Mitigation. Cambridge University Press,Cambridge.

IPCC (Intergovernmental Panel on Climate Change), 2007. Summary for Policymakers of Climate Change 2007: The Physical Science Basis.Contribution of Working Group I to the Fourth Assessment Report of the Intergovemental Panel on Climate Change. Cambridge University Press,Cambridge.

Jiang Y, Xia J, 2007. The hydrological characteristics of runoff and its response to climatic change in Tarim. Resources Science, 29(3): 45–52.

Jiang Y, Zhou CH, Cheng WM, 2005. Analysis on runoff supply and variation characteristics of Aksu Drainage Basin. Journal of Natural Resources, 20(1): 27–34.

Jiang Y, Zhou CH, Cheng WM, 2007. Stream flow trends and hydrological response to climatic change in Tarim headwater basin. Journal of Geographical Sciences, 17(4): 51–61.

Lanzhou Institute of Glaciology and Geocryology, 1987. Glacier Inventory of China. III, Tianshan Mountain (Inter Drainage Area of Scattered Flow in East. Sciences Press, Beijing. 61–84.

Li L, Wang QC, Zhang GS, Fu Y, Yan LD, 2004. Impact of climate change to surface runoff in the upper Yellow River. Acta Geographical Sinica,59(5): 716–722.

Liu XZ, Tian YY, 1991. Researching influence of climate change to runoff by ameliorating nonlinear water balance model. Advances in Water Science, 2(2): 120–126.

Qin DH, 2002. An Assessment of Western China Environment Evolution.Science Press, Beijing. 32–57.

Qin DH, Ding YH, Su JL, 2005. Assessment of Climate and Environment Changes in China (I): Climate and environment changes in China and their projection. Advances in Climate Change Research, 1(1): 4–9.

Shi YF, Shen YP, Li DL, Ding YJ, Hu RJ, 2003. An Assessment of the Issues of Climatic Shift from Warm-dry to Warm-wet in Northwest China.China Meteorological Press, Beijing. 17–83.

Wang GQ, Shi ZH, 2000. The influence of climate change to the hydrological process of the upper Yellow River. Meteorological and Environmental Sciences, (4): 20–22.

Wang J, Ding YJ, Liu SY, Ye BS, Xie CW, Wu JK, 2005. Analysis on the change characteristics of climate and hydrological conditions in Aksu prefecture, Xinjiang since recent 40 years. Arid Zone Research, 22(3):336–340.

Wang MH, Xie Q, Wang HY, 2003. Impact of future climate change on runoff depth of the Huaihe drainage basin. Geographical Research, 22(1):79–88.

Wang R, Sun ZD, Gao QZ, 2006. Water level change in Bosten Lake under the climatic variation background of central Asian around 2002. Journal of Glaciology and Geocryology, 28(3): 324–329.

Water Conservancy Bureau of Xinjiang and Water Conservancy Society of Xinjiang, 1999. River Hydrology and Water Resources in Xinjiang. Xinjiang Technology and Health Press, Urumqi, China. 333–429.

Wu SF, Liu ZH, Han P, 2006. Impact of climate change on water resources of the Urumqi River Basin. Journal of Glaciology and Geocryology,28(5): 702–706.

Xinjiang Geographic Society, 1993. Xinjiang Geographic Manual. Urumqi:Xinjiang People Press, 84–94.

Xu YL, Huang XY, Zhang Y, Lin WT, Lin ED, 2005. Statistical analyses of climate change scenarios over China in the 21stcentury. Advances in Climate Change Research, 1(2): 80–83.

Yuan YJ, He J, Mu GJ, 2003. Summer precipitation change for recent 40 years in the Tianshan Mountains and compared with southern and northern Xinjiang Regions. Journal of Glaciology and Geocryology, 25(3):331–336.

10.3724/SP.J.1226.2011.00274

*Correspondence to: Prof. YongChao Lan, Cold and Arid Regions Environmental and Engineering Research Institute, Chinese Academy of Sciences. 320 West Donggang Road, Lanzhou 730000, Gansu, China. Tel: +86-0931-4967161; Email: lyc@lzb.ac.cn

24 April 2010 Accepted: 16 December 2010