当前位置:首页 >> 电力/水利 >>

Huang Pang 2010 Journal of Hydrology:塔里木河河岸带地下水对生态输水的响应


Journal of Hydrology 387 (2010) 188–201

Contents lists available at ScienceDirect

Journal of Hydrology
journal homepage: www.elsevier.com/locate/jhydrol

Changes in groundwater induced by water diversion in the Lower Tarim River, Xinjiang Uygur, NW China: Evidence from environmental isotopes and water chemistry
Tianming Huang a,b, Zhonghe Pang a,*
a b

Key Laboratory of Engineering Geomechanics, Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing 100029, China Graduate School, Chinese Academy of Sciences, Beijing 100039, China

a r t i c l e

i n f o

s u m m a r y
The Lower Tarim River in NW China is under severe ecosystem degradation due to stopped stream ow and diminished groundwater recharge. Since year 2000, eight water diversions from the upper stream and from the neighboring Kaidu–Kongque River have been implemented to alleviate the ecosystem disaster. In order to assess the effectiveness of the water diversion project and to identify proper tracers of groundwater dynamics, we sampled the riparian groundwater system in 2007 and 2008 along the 350 km-long river channel through the 40 monitoring wells situated along nine transects perpendicular to the river and three soil proles. Measurements on the samples have included environmental isotopes (18O, 2H, 3H) and water chemistry. The results show that remarkable changes have been induced by the water diversions. The observed response of riparian groundwater system includes general decrease in total dissolved solid (TDS) and rise of water table. Scope with greater than 1 m rise in water table is within $700 m from the riverbank in the upper segments and $300 m in the lower ones. Greater rise of water table occurs near the river bank. Tritium data show that the extent of modern recharge (since 1960s), including that from the diverted water, is limited to 600 m from the riverbank at the upper segments and 200 m at the lower ones. Stable isotopes show that groundwaters, regardless of modern or pre-modern, are enriched in heavy isotopes and are plotted in parallel to the meteoric water line in the d–d plot, attributed to evaporation during recharge. Groundwater is generally of Na–Mg–Cl–SO4 type and is formed by dissolution of minerals, such as halite, sulfate, and carbonates, based on component correlation matrices analysis. The salinity of groundwater is mainly affected by that of the diverted water and of the local antecedent groundwater, salts in the unsaturated zone, evapotranspiration during recharge. As the zone of smaller groundwater depth (less than 5 m) suitable for the most existing Populus euphratica and Tamarix ramosissima, the main species targeted by the rescue effort, restricts to 200 m from the riverbank, and narrows down towards downstream, long-term stability of the ecosystem cannot be achieved by the current water diversion scheme and regulating/saving water in source-streams and the Upper/Middle Tarim River is crucial for continuing water diversion. 2010 Elsevier B.V. All rights reserved.

Article history: Received 11 June 2009 Received in revised form 21 March 2010 Accepted 5 April 2010 This manuscript was handled by L. Charlet, Editor-in-Chief, with the assistance of S. Yue, Associate Editor Keywords: Water diversion Environmental isotopes Water chemistry Salinization mechanism Tarim River Arid environment

1. Introduction Under the dual impacts of anthropogenic activities and climate change, a common scenario in arid and semiarid catchments, particularly in the lower reaches of them, is severe ecological degradation, such as death of vegetation, intensied groundwater salinization, soil salinization and desertication, etc. (Feng et al., 2005; Gremmen et al., 1990; Ma et al., 2005; Richardson et al., 2007; Stromberg et al., 1996; Wang and Cheng, 2000). In groundwater-dependent ecosystems, dynamics of soil moisture relative to water table uctuations control the overall ecosystem dynamics
* Corresponding author. Tel.: +86 10 82998613. E-mail address: z.pang@mail.iggcas.ac.cn (Z. Pang). 0022-1694/$ - see front matter 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.jhydrol.2010.04.007

(Tamea et al., 2009), and when the riparian ecosystems are threatened by insufcient supply of the river, all the services provided by the ecosystems may be threatened (Stromberg et al., 2007). Therefore, ecological allocations of water should be considered for groundwater-dependent ecosystems (Eamus et al., 2006). Interest and investment in river restoration projects are growing worldwide (Klein et al., 2007 and references within). Better understanding the essential groundwater–surface water interaction and water chemistry evolution is of great signicance to any restoration planning and implementation (Baillie et al., 2007; Loheide and Gorelick, 2007). With a relatively stable inow from headwater streams, Tarim River, Xinjiang Uygur, NW China as a typical catchment of its kind, water discharge to the lower reaches has decreased rapidly as a

T. Huang, Z. Pang / Journal of Hydrology 387 (2010) 188–201

189

result of large scale agriculture development and irrational water resources utilization in the upper and middle reaches of the river (Feng et al., 2005; Song et al., 2000). The runoff has ceased to ow into the 350-km-long Lower Tarim River since the construction of the Daxihaizi Water Reservoir in 1972, causing severe damages to the riparian forest dominated by Populus euphratica (Song et al., 2000). The so-called ‘‘Green Corridor” and the high way along the Lower Tarim River (Kuala to Ruoqiang, part of National High Way G218) are endangered by degrading riparian vegetation and desertication. Since 2000, the water diversion project (cost CNY 10.7 billion) has been implemented to alleviate the vegetation degradation. After the eight impulsive water diversions, the water tables have risen at differing degrees from the river bed and the composition, types, distribution and growth status of the riparian vegetation have changed correspondingly (Chen et al., 2008a,b; Deng, 2009; Hou et al., 2007a,b; Tao et al., 2008). Due to long-time runoff cutoff, the ecosystem is not sustainable as survival plants are dominated by the older ones. Although a total volume of 2.27 billion m3 of water has been diverted to the lower reaches by eight water diversions, it is still at the stage of compensation for groundwater recession. The present impulsive linear water diversion through the river bed cannot radically preserve the local eco-environment, though ecological degradation is slowed down to some extent. Sustainable water utilization and ecosystem protection scheme are imperative for the degraded ecosystem restoration and prevention of the Taklimakan Desert and the Kuluk Desert from merging in the Lower Tarim River. The study of water cycle evolution to solve water resources and relative environmental issues has become a signicant task of the hydrological science (Zhang et al., 2000). Since climate, landforms, lithology and anthropogenic activities inuence the crustal weathering and groundwater chemistry (Shen et al., 1993), an understanding of the chemical evolution of the groundwater can provide insight into the interaction between water and environment and can contribute to rational water resources management (Adams et al., 2001; Edmunds, 2009; Ma et al., 2009), especially in arid regions with fragile ecosystem and intense anthropogenic interferences (Bennetts et al., 2006). Groundwater chemistry is important for ecosystem restoration since water quality controls vegetation growth status and soil characteristics, though the most present studies focus on responses of water table and vegetation changes in the Lower Tarim River. One of the disadvantages of high salinity water to vegetations growth is to prevent vegetation from absorbing moisture and to plague soil fertility (He et al., 2006; Manchanda and Garg, 2008). Plant features, including species richness, species diversity and species composition are signicantly related to salinity and deteriorate with increasing salinity (Lymbery et al., 2003). As groundwater depth, ood events and salt concentration are the key factors controlling the growth of the arbor and shrub (Thevs, 2007; Williams et al., 2006), the variations of water table and quality and their impacts on vegetation growth under water diversion circumstances need further study. To increase the chance of success in riparian ecosystem restoration, it is important to gain knowledge of the hydrological and hydrogeochemical processes involved. The Xinjiang Water Conservancy Bureau and Ministry of Water Resources of China had investigated environmental background in the Lower Tarim River before water diversion (Deng, 2009). Based on monitoring data of water table and total dissolved solid (TDS) in groundwater since 2000 and two systematic sampling campaigns carried out in August 2007 and May to June, 2008, this paper attempts to assess the effectiveness of the water diversion project through understanding the mechanism of groundwater system changes, including groundwater residence times, the extent of groundwater recharge under water diversion, the changes of salinity and salini-

zation mechanism so as to provide a basis for an optimal water diversion schemes using environmental isotopes (18O, 2H, 3H), water chemistry and water table monitoring data. It is also a purpose of current study to identify suitable environmental tracers for the eco-hydrological processes. 2. Background 2.1. General setting The Tarim River Basin is located in the south of Xinjiang, NW China. It has an area of 1.04 106 km2 and is anked by the Tianshan Mountains to the north and by the Kunlun Mountains to the south (Fig. 1). The Taklimakan Desert, the largest desert in China, is located in the center of the basin, occupying an area of 3.37 105 km2 (Zhu et al., 1981). Aksu River, Yarkant River and Hotan River are three large rivers in the west of the basin, which feed the Tarim River at Aral. Due to river regulation, the later two presently recharge the Tarim River only during large oods. The Tarim River starts from Aral to Taitema Lake with a length of 1321 km. The upper stream of the Tarim River is from Aral to Yingbazha (495 km), while the middle reaches is from Yingbazha to the Qiala Water Reservoir (398 km), and the lower reaches is from the Qiala Water Reservoir to Taitema Lake (428 km) (Fig. 1). The lower reaches can be divided into three segments: the upper from the Qiala Water Reservoir to the Daxihaizi Water Reservoir, the middle from the Daxihaizi Water Reservoir to Aragan, and the lower from Aragan to Taitema Lake. The river is divided into two branches starting from the Daxihaizi Water Reservoir. The west is the old Tarim River (142 km) and the east is the Tarim River (205 km). The two branches combine at Argan. Further downstream from the Daxihaizi Water Reservoir, there is little irrigation and agricultural activities. The Kaidu River ows into Boston Lake, then is pumped into Kongque River at southwest of the lake, and nally reaches Lop Nur. There are two eolian deserts situated in both sides of the Lower Tarim River, the Taklimakan Desert on the west and the Kuluk Desert on the east, with moving and semi-moving sand dunes. 2.2. Hydrogeology The Tarim River Basin is a Mesozoic–Cenozoic basin surrounded by folded mountains. Outcrop of the strata from late Paleozoic to Cenozoic is described in Fig. 1 and the Paleozoic is widely distributed in mountain areas. Archaeozoic and Proterozoic schist and gneiss, Paleozoic and Mesozoic sand stones, conglomerates and magmatic rocks occur in the sources area of the Tarim River (XETCAS, 1965; Zhu et al., 1981). The large fault between basins and mountains controls the tectonic evolution and forms a series of major depressions in the Tarim River Basin, such as Kuche Depression, North Depression, Southeast Depression and Southwest Depression (Cai et al., 1997). The major sediment is tertiary in those depressions, for instance, the Kuche Depression with depth of thousands of meters has a tertiary deposit with a maximum depth of 4500 m (Li et al., 2000). The quaternary has extensive distribution in those depressions (Fig. 1). The occurrence of groundwater is similar between the Southern Tianshan watershed and Northern Kunlun watershed. The sink of the two groundwaters systems is centered in the south of the Tarim River. The diluvial aquifer from the northern mountains is composed of sand deposits some 100–300 m thick forming an unconned aquifer in which the present day water table ranges between 20 m and 200 m below surface (Fig. 2). This allows a certain amount of surface runoff in the piedmont fan to inltrate and recharge the aquifer. At the southern edge of this diluvial fan, the

190

T. Huang, Z. Pang / Journal of Hydrology 387 (2010) 188–201

Fig. 1. Sketch map of the Tarim River Basin (modied from Li et al. (2000)) 1 – Proterozoic; 2 – Paleozoic; 3 – Mesozoic; 4 – tertiary; 5 – granite; 6 – quaternary; 7 – river; 8 – surface water in the Upper Tarim River and the Aksu River; 9 – regional groundwater ow.

Fig. 2. Hydrogeological cross section in the Middle Tarim River, modied from Li et al. (2000), see Fig. 1 for the location.

aquifer comprised of alluvial loam becomes conned or semi-conned and the thickness is about 200 m. The regional ow is from north to south (Li et al., 2000). In the middle and lower reaches of the Tarim River, local groundwater system gets recharge from surface water through bank inltration. The Lower Tarim River is located at a at alluvial plain and the hydrogeological condition is simple. The multi-layered aquifer with homogenous lithology is dominated by uvial and lacustrine facies of ne sand and silty ne sand, and locally by eolian sand (Fig. 3). The typical porous aquifer can be divided into phreatic and conned ones. The shallow groundwater is closely related to surface water, serving as the main water sources for the riparian vegetation. The phreatic aquifer mainly consists of ne sand and silty ne sand and has a relatively low permeability and poor water yield with the thickness, hydraulic conductivity and specic yield of 30–40 m, 1.2–4.8 m d1 and less than 150 m3 d1 m1, respec-

tively. The aquifer is underlain by a clay bed, which extends continuously and horizontally and almost has little hydraulic connection to the conned aquifer underneath (Alim and Xu, 2003; Dong and Deng, 2005). The minerals contained in the granites and metamorphic rocks are difcult to be dissolved. In the alpine areas, the soluble rocks are mainly carbonate (Zhu and Yang, 2007). The dissolution begins with eroding host rocks with CO2 from precipitation and soil zone and generally forms low mineralization type water of HCO3–Ca in mountainous area. The originating water for the Tarim River has low TDS (less than 0.5 g/L) at river closure mouths. In the tertiary sediments at the low elevations, more soluble deposits are found (Zhu et al., 1981). Under inland dry climate conditions, those deposits make considerable contribution to the high TDS in river and groundwater in middle and lower reaches of the river. At the Tarim River sections, Fan et al. (2002) showed that the average

T. Huang, Z. Pang / Journal of Hydrology 387 (2010) 188–201 Table 1 Runoff changes at each station on the Tarim River (108 m3). 1950s Aral Qiala Tikanlik Argan Lop Village 49.4 13.5 8–9 Persist 5–4 1960s 51.7 11.4 2.9 Discontinue 0.2 1970s 44.4 6.7 0.5 Nil Nil 1980s 44.8 3.9 0.4 Nil Nil

191

1990s 42.0 2.8 0.1 Nil Nil

ers. The Lower Tarim River, as the terminal of water and salt of the Tarim Basin, has high salt concentration due to long-term evaporation (Cheng, 1993). The background of the regional TDS distribution in groundwater in the Lower Tarim River before the water diversions is that, the scale of low-mineral groundwater region with TDS of 1–3 g/L distributes in riparian area with the width from 1000 to 2000 m to the riverbank due to river dilution, however, beyond this zone, TDS in groundwater is greater than 5 g/L and more than 10 g/L in the lower segments of the lower reaches (Cheng, 1993; Deng, 2009). 2.3. Climate The Lower Tarim River is dominated by typical continental temperate arid climate. According to Tikanlik meteorological station in the Lower Tarim River, the average precipitation is about 40 mm/yr; the potential evaporation is 2590 mm/yr; the annual average temperature is 10.5 °C and average sandstorm days are 8.2 days per year, causing severe wind-sand hazards. 2.4. Ecosystem degradation Under the impact of anthropogenic activities, runoff of three source streams (Aksu River, Hotan River and Yarkant River) to the Tarim River has decreased gradually in the last 50 years due

Fig. 3. Hydrogeological cross sections in the Lower Tarim River, modied from Deng (2009), see Fig. 1 for the location; also showing the location of the three soil proles.

monthly mineralization since 1958 was greater than 1 g/L over the entire year except for the wet season (August), due to runoff decreasing and salt drainage of the main canals for irrigation. Zhang et al. (1995) studied water chemistry of source-rivers of the Tarim River using conventional hydrochemistry and strontium isotopes (87Sr/86Sr) and pointed out that weathering of silicates, carbonates and evaporite contributed most of the TDS in those riv-

Fig. 4. Sampling locations.

192

T. Huang, Z. Pang / Journal of Hydrology 387 (2010) 188–201

Table 2 Statistics of eight water diversions to the Lower Tarim River. No. First Second Third Fourth Fifth Sixth Seventh Eighth Total Delivering duration 2000.5.14-7.12 2000.11.3-2001.2.5 2001.4.1-7.6 2001.9.12-11.18 2002.7.20-11.10 2003.3.3-7.11 2003.8.4-11.3 2004.4.23-6.22 2005.4.18-6.7 2005.8.30-11.2 2006.9.25-11.21 13 (57%) Boston Lake (108 m3) 0.98 2.25 1.84 1.63 2.45 2.4 0 0.74 0.52 0 0.26 9.7 (43%) The Tarim River (108 m3) 0 0 0 0.35 0.86 0.97 2.85 0.29 0 2.28 2.07 22.7 (100%) Total (108 m3) 0.98 2.25 1.84 1.98 3.31 3.37 2.85 1.03 0.52 2.28 2.33 Transection for water reach Karday (E) Kargen (I) Taitema Lake Taitema Lake Taitema Lake Taitema Lake Taitema Lake Taitema Lake Yikanbujima (H) Taitema Lake Taitema Lake

to extensive oasis agriculture with increasing water utilization (Feng et al., 2005). With gradually decreased inow to the Tarim River, proportion of water consumption in the upper and middle reaches increased gradually from 1970s to 1990s, while ow to the lower reaches reduced signicantly (Table 1). The groundwater depth has increased to 8–12 m. As a result, serious ecosystem degradation in the Lower Tarim River has been observed and the major manifestation is described as follow (Deng, 2009): (1) The area of the groundwater-dependent P. euphratica forests has decreased from 54,000 hm2 in 1950s to 7.3000 hm2 at present. The Green Corridor of great strategic signicance for the Lower Tarim River between Taklimakan and Kuluk Deserts has been brought to the verge of destruction, as the two deserts are taking on a trend of merging and some owing sand has already intruded into the river bed and the national highway G218. (2) The desertication has intensied. According to the aerial photographs in 1959–1983, desertication has developed most seriously in the Lower Tarim River. The area of the desertied land has increased by 22% during 24 years. Particularly, desertication has developed dramatically since 1972 as the Lower Tarim River dried out. As a result, the vegetation coverage has been declining, and many roads, farmlands and villages have been buried by sand, causing a serious threat to the survival and development of the oasis. 2.5. Water diversion To protect the Green Corridor, the riparian vegetation restoration is imperative. Taking advantage of the wet period of the Kaidu River, the Ku–Ta Channel (Yuli to Qiala, Fig. 4) was constructed for diverting water from Kongque River to the Lower Tarim River. The source water for the rst ve water diversions is mainly Boston Lake (Table 2) while that for the later water diversions is mainly the Tarim River, which was all stored in the Daxihaizi Water Reservoir before being released to the lower reaches. Altogether 2.27 billion m3 of water (1.30 billion m3 from the Boston Lake and 0.97 billion m3 from the Tarim River) has been diverted to the Lower Tarim River from the Daxihaizi Water Reservoir from year 2000 to the end of 2006. Six impulsive water diversions out of the eight reached the Taitema Lake, the terminal lake of the Tarim River. Groundwater system has been changed by the water diversions, resulting in redistribution of water salinity, water table and soil moisture content, which affect the evolution of the ecosystem. 3. Sampling and analyses The rst sampling campaign was carried out during the wet season on 26–31, August, 2007, after the eighth water diversion.

Surface waters from Kaidu River, Boston Lake, Kongque River, the middle reaches of the Tarim River and residual waters of water diversion in dry riverbed in the Lower Tarim River were collected. Two rainfall samples were also collected during a storm event. Groundwater samples were collected from boreholes at varying distances from the riverbed along nine groundwater monitoring transects, namely: Akdun (A), Yahopumarhan (B), Yengsu (C), Abudali (D), Karday (E), Tugmailai (F), Aragan (G), Yikanbujima (H), and Kargan (I) (Fig. 4), respectively. Along each transect, monitoring wells with depth of 8–17 m were dug at intervals of 100– 200 m. In total, 40 monitoring wells were measured and sampled. Water table and TDS of groundwater have been monitored three times a month during the water diversions and once a month in between the water diversion events. A further sampling campaign was implemented in May 28 to June 3 in 2008 in Aksu River, the Upper Tarim River, the Qiala
Table 3 Site measurements and isotopic composition of surface waters. Sample TDS (g/L) EC (ms/cm) Temp (°C) pH d18O (‰) d2H (‰) 0.28 0.31 0.45 1.24 1.35 1.30 1.28 1.24 12.7 10.5 30.7 31.1 23.0 24.8 27.7 24.8 19.6 27.4 24.5 24.6 24.3 23.6 23.8 23.8 24.3 23.9 23.5 23.2 8.0 11.1 8.6 10.8 7.8 10.0 7.6 8.8 8.0 8.5 8.1 8.4 8.6 8.6 8.6 8.6 8.7 8.6 8.6 8.6 8.6 8.5 8.1 9.9 8.4 8.3 7.8 8.5 7.8 1.0 78.0 77.6 74.9 73.7 58.7 57.9 54.9 59.4 54.0 17.1
3

H (TU)

Aksu River K1 0.132 TL 0.149 A1 0.217 Tarim River T1 0.608 R8 0.666 R9 0.642 R7 0.631 R6 0.617

20.0

18.4 20.3 18.8 27.7 28.3

Kaidu–Kongque Basin R11 0.181 0.38 L1 1.053 2.09 L2 0.981 1.49 L3 0.949 1.89 L4 0.840 1.68 L5 0.844 1.69 L6 0.838 1.68 L7 0.838 1.67 L8 0.769 1.55 L9 0.792 1.59 L10 0.844 1.70 R10 0.661 1.34 Water reservoir Qiala 0.754 Daxihaizi 0.894 1.49 1.79

2.7

26.9

2.4 6.7 4.6 3.6 3.5 2.0 2.4 2.5 2.2 1.1

23.4 46.7 48.3 42.4 5.8 17.1 13.1 13.8 12.7 6.6

26.9 28.1

24.0

9.2 9.0 8.8 8.2 9.8

Residual water in the Lower Tarim River R1 3.20 6.06 23.3 R2 2.64 5.00 25.5 R4 3.82 7.19 26.0 R3 2.93 5.56 27.3 Rainfall Rn1 Rn2

27.0 32.6 22.7

29.8

T. Huang, Z. Pang / Journal of Hydrology 387 (2010) 188–201

193

Water Reservoir and the Daxihaizi Water Reservoir. The data of these surface water bodies will serve as source water composition and major modications through evaporation. Soil samples were collected from three sampling proles in the Lower Tarim River. The soil samples were obtained using a hollow-stem hand auger with interchangeable 1.5 m aluminum rod from three proles (5.8–7.7 m depth, SP1 and SP2 in section C and SP4 in section G, Figs. 3 and 4). Bulk soil samples of $400 g were collected at intervals of 0.25 m. Samples were homogenized over the sampled interval and immediately sealed in polyethylene bags. Gravimetric moisture content was determined by drying a minimum of 80 g of soil at 110 °C for 12 h. To determine chloride content, doubledeionised water (40 mL) was added to the oven-dried soil sample (40 g) (Scanlon, 1991). Samples were agitated on a reciprocal shaker table for 8 h. The supernatant was ltered through 0.45 lm lters. Chloride was then analyzed by ion chromatography. The chloride concentration of the soil solution is then calculated by dividing the measured concentration by gravimetric moisture content and by multiplying the mass ratio of solution to ovendry soil. Depth of boreholes, water table, location (GPS), water temperature, pH, TDS and electrical conductivity (EC) were measured on the site. Fifty milliliters of water was collected for stable isotopes and was measured in the Stable Isotopes Laboratory, Institute of
Table 4 Site measurements and isotopic composition for groundwaters at each section. Sample Akdun (A) w1 w2 w3 Local well A2 A3 A4 Distance to the river (m) 50 (east) 150 (east) 250 (east) 50 50 (east) 150 (east) 250 (east) 250 (east) 50 150 250 350 450 750 50 150 250 150 250 550 850 50 150 250 50 500 800 100 300 500 300 500 Welldepth (m) 2.97 4.40 5.09 3.92 4.42 5.93 5.96 2.17 4.62 5.66 5.46 5.50 6.03 6.17 4.87 4.86 4.50 5.92 6.01 5.57 9.54 4.52 4.60 4.98 4.80 9.22 6.57 5.60 5.88 6.56 10.22 12.26

Geology and Geophysics, Chinese Academy of Sciences. 2H/H and O/16O were measured on isotope ratio mass spectrometry (MAT253) by chrome reduction and equilibrium with CO2, respectively. Results are reported as d2H and d18O (d = (Rsample/ Rstandard 1) 1000) using the Vienna Standard Mean Ocean Water (VSMOW) as standard. The analytical precision is ±1‰ for d2H and ±0.1‰ for d18O. Five hundred milliliters of water samples were collected for tritium measurement in the Groundwater Tracing Laboratory, Institute of Geology and Geophysics, Chinese Academy of Sciences through electrolytic enrichment with a tritium enrichment factor of about 20 and the liquid scintillation counting (Quantulus 1220) method with a detection limit of 0.3 TU (Tritium Unit). Water chemistry was measured using ion chromatography (Dionex-500) at the Beijing Research Institute of Uranium Geology. The methods for cation measurements are taken from the National Analysis Standard DZ/T0064.28-93 while for anion are from DZ/T0064.51-93. Alkalinity was measured on automatic titrator (785 DMP). Analytical precision was 3% of concentration based on reproducibility of samples and standards and detection limit was 0.1 mg/L. The charge balance error for 24 samples with low TDS ranges from 4% to 4% and that for 11 samples with high TDS ranges from 13% to 8%, which may be caused by large dilution ratio and disturbance of organic matters. The results are shown in Tables 3–5.
18

TDS (g/L) 2.33 2.43 2.07 1.284 1.170 1.383 2.04 3.72 0.884 1.588 1.737 1.631 1.986 15.32 1.116 0.973 2.25 1.160 0.750 1.192 5.90 1.605 1.363 3.11 1.693 0.878 1.039 1.355 1.279 0.742 >55 55

SEC (ms/cm) 4.47 4.62 3.99 2.53 2.31 2.71 3.94 7.02 1.77 3.11 3.38 3.18 3.84 26.20 2.21 1.94 4.32 2.29 1.51 2.36 10.77 3.13 2.68 5.89 3.32 1.75 2.05 2.65 2.51 1.49

Temp (°C) 15.3 16.9 16.7 17.9 19.0 17.5 19.0 18.2 17.6 15.8 16.7 17.9 17.9 18.0 20.8 18.6 19.6 18.0 18.3 19.7 19.2 21.4 18.6 19.5 18.0 21.1 19.2 20.3 22.2 18.9

pH 7.2 7.3 7.3 8.1 7.8 7.8 7.9 6.4 9.6 8.3 7.9 7.7 7.5 7.2 8.1 7.7 7.7 7.8 7.9 7.4 7.4 8.0 7.9 7.9 9.2 7.9 7.9 7.6 7.5 7.6

d18O (‰)

d2H (‰)

3

H (TU)

Yahopumarhan (B) w4 B1 w5 B2 w6 B3 w7 B4 Yengsu (C) w14 C1 w13 C2 w12 C3 w11 C4 w10 C5 w9 C6 w8 C7 Abudali (D) w15 D1 w16 D2 w17 D3 Karday (E) w21 E2 w20 E3 w19 E4 w18 E5 Tugmailai (F) w22 F1 w23 F2 w24 F3 Aragan (G) w27 G2 w26 G4 w25 G5 Yikanbujima (H) w32 H1 H2 w31 w30 H3 Kargan (I) w29 w28 I2 I3

5.7 6.3 6.3 6.3 7.4 5.2 6.3 5.5 7.4 6.7

44.6 49.2 48.9 49.9 56.3 41.6 48.8 44.1 56.7 53.7

22.2 21.4 21.8 37.0 18.9 25.6 43.8 46.3 22.5 1.1

5.2 7.4 8.2 8.1

41.1 57.1 62.9 61.4

22.7 19.1 <0.3 0.8

7.0 7.9 7.5 6.3 6.2 7.6

54.1 60.1 57.4 49.2 48.7 58.9

19.6 29.2 2.4 29.0 30.0 4.1

107.60

22.2

6.8

6.6

51.3

<0.3

194

T. Huang, Z. Pang / Journal of Hydrology 387 (2010) 188–201

Table 5 Chemical composition of the surface water and groundwater sections (mg/L). Sample Aksu River K1 TL A1 Tarim River T1 R8 R9 R7 R6 Ca2+ Mg2+ Na+ K+ HCO 3 92 98 120 153 132 128 115 130 194 255 246 261 251 94 101 660 222 324 398 436 192 445 321 362 754 323 304 302 353 487 314 336 253 260 282 Cl SO2 4 66 52 94 226 247 233 265 234 47 547 438 285 331 404 1609 1477 1191 503 553 875 222 430 662 687 5912 433 235 396 980 563 320 340 420 530 291 F NO 3 2.0 2.3 3.1 3.4 2.6 2.5 0.8 2.7 2.0 1.1 1.6 1.1 1.7 2.2 3.1 <0.1 4.1 6.1 11.5 15.1 1.6 3.8 2.8 10.7 11.2 6.0 5.8 4.2 3.7 <0.1 4.8 4.1 4.9 5.5 6.2 Fig. 6. Post map of Tritium content (TU) for groundwaters from the Lower Tarim River with a solid line showing the scope of modern recharge.

39 39 44 82 68 67 73 65

11 7.6 20 45 33 33 36 31

3.7 13 17 127 153 160 150 150 12 250 195 138 204 239

2.3 2.0 2.8 7.6 9.2 10 14 10 3.9 20 16 8.2 12 14

4.0 16 25 193 287 272 238 264 7.7 339 274 181 235 353 1705 1721 1511 525 573 906 451 665 770 728 9291 457 277 548 3748 682 304 415 734 581 247

0.4 1.0 0.2 0.4 1.0 1.1 0.9 0.7 0.1 0.5 0.4 0.3 0.5 0.9 1.9 6.5 1.6 1.3 1.0 1.0 1.0 0.6 1.0 1.4 1.9 0.7 0.3 0.7 3.7 1.0 1.8 1.4 0.8 0.6 0.7

Kaidu–Kongque Basin R11 64 16 L1 60 91 L10 57 76 R10 73 50 Water Reservoir Qiala 77 Daxihaizi 67 71 67

Fig. 5. The precipitation tritium input from 1952 to 2007 and the decayed value for 2007.

Residual water in the Lower R1 142 214 R4 150 249 R3 115 177 Yahopumarhan (B) w4 126 84 w6 126 105 w7 151 159 Yengsu (C) w13 47 w12 92 w11 145 w10 136 w8 571 Karday (E) w21 107 w20 94 w19 96 w18 183 Aragan (G) w27 32 w26 80 w25 103 Yikanbujima (H) w32 85 w31 160 w30 94 64 103 142 135 1095 62 59 80 204 110 105 70 98 109 66

Tarim River 887 40 1125 54 799 37 329 350 485 270 454 422 415 4550 338 174 321 1949 547 167 268 373 249 151 23 27 31 21 23 46 29 83 17 26 18 53 23 18 21 24 20 16

4. Results and discussion 4.1. Tritium Tritium with half-life of 12.32 years carries the information of water itself and has severed as an important tracer in determining modern groundwater recharge, movement and ages (Michel, 2005). Tritium is used to distinguish pre-modern recharge from modern recharge in this study. The approximate tritium input sequence of precipitation in the study area is based on the tritium sequence for the neighboring Lop Nur from 1952 to 1996 reconstructed by Jiao et al. (2004), Urumqi meteorologic observation station from 1986 to 2001 (IAEA and WMO, 2006), river samples collected in 2001 by Li et al. (2006), and river samples plus one rainfall sample collected by this study. The data for the overlapping period (1986–1996) is similar (Fig. 5). In 2001, the content of tritium in surface water from the Aksu River and the Tarim River ranges from 20 to 56 TU, while in 2007, from 18 to 21 TU, and that in Boston Lake and residual water is more concentrated ranging

from 22 to 33 TU. Although glacial melt water contributes about 40% to river runoff in the Tarim River Basin (Shen and Wang, 2002), tritium in precipitation in Tarim Basin is still higher than other areas. The content of tritium for the rainfall sample is 29.8 TU. This high tritium content may relate to nuclear tests, atmospheric circulation in high latitude (Liu, 2001). The results show a tritium content decrease from 2586 TU (1963) to 20$30 TU (2007). Using an exponential decay equation, the decayed tritium contents for 2007 in precipitation, which would represent tritium concentrations in groundwater that had inltrated between 1952 and 2007, ranges from 50 TU to 225 TU from 1962 to 1966, and ranges from 10 to 35 TU from 1967 to 2007 (Fig. 5). Therefore, groundwater with tritium content less than 10 TU is regarded as pre-modern water, or at least most part is pre-modern water when mixing with modern water is considered. Fig. 6 shows the tritium contents in the riparian groundwater from the Lower Tarim River. The solid line in the gure is the boundary between groundwater that is recharged by modern water (19.1–46.3 TU) and that by pre-modern water (less than 4.1 TU). The extent of modern recharge (including the diverted water recharge) is limited to 600 m in the upper segments and to 200 m in the lower segments with a decreasing trend. There is a rising extent at section G (the conuence area of two tributaries) due to ush recharge. According to the groundwater samples collected in the vicinity of the river bed of the Lower Tarim River in 1989 by Liu et al. (1997), 30% of them have the tritium contents more than 67 TU after the 18 years’ decay by 2007, thus the groundwater in those regions is a mixing of the diverted water from year 2000 and the antecedent groundwater.

T. Huang, Z. Pang / Journal of Hydrology 387 (2010) 188–201

195

4.2. Stable isotopes 4.2.1. Stable isotopes in surface waters Only two stations (Urumqi and Hotan, Fig. 1) are involved in the Global Network of Isotopes in Precipitation (GNIP) and the precipitation data are close to the Global Meteoric Water Line (GMWL, Craig, 1961). Therefore, for simplicity, the GMWL has been used as the reference meteoric water line for this study. In the eld sampling in the Lower Tarim River, two rainfall samples were also collected during a rainfall event. The isotopic composition is shown with a slope of 5.5 (Fig. 7), indicating evaporation during precipitation. The Kaidu River at the entrance to the Boston Lake has isotopic composition of 7.8‰ for d18O and 54.0‰ for d2H. Isotopic enrichment occurs in the Boston Lake and the Kongque River as a result of evaporation. However, the Boston Lake water is not mixed completely with the deviation of 0.7‰ in d18O of the three water samples collected in the south part of the lake (Fig. 7). Those ve samples form an evaporation line:

r2 H 5:4 r18 O 11:3; R2 0:998; n 5

1

Daxihaizi Water Reservoir in 1972, the river ceased to ow. The pre-modern waters with tritium less than 4.1 TU recharged by the Tarim River show d18O value of 8.2‰ to 6.6‰ with an average of 7.5‰ and the d2H values of 62.9‰ to 51.3‰ with an average of 57.6‰ (Fig. 7).These groundwater samples are located relatively far away from the river bed at each section as compared to the modern water samples (Fig. 6). The modern groundwaters with tritium contents ranging from 18.9 TU to 46.3 TU have been recharged by the Tarim River since the 1960s and some of them have mixed with the diverted water. The modern groundwaters show d18O values of 7.9‰ to 5.2‰ with an average of 6.4‰ and d2H values of 41.1‰ to 60.1‰ with an average of 50.0‰, heavier than pre-modern waters. The modern and pre-modern groundwaters exhibit a similar behavior to fall in slight parallel to the meteoric water line, but enriched relative to the recharging river water. The phenomenon is commonly observed in dry climate (Chapman et al., 2008; Prasanna et al., 2009) and attributed to evaporation during river recharge to the riparian groundwater system in a rather uniform manner.

with the slop of 5.4, indicating humidity of $75% during evaporation (Gonantini, 1986). Four water samples were collected in 2007 from the middle reaches of the Tarim River. The isotopic composition ranges from 7.8‰ to 8.5‰ for d18O and from 59.4‰ to 54.9‰ for d2H, respectively. These data points are closely plotted with an average deuterium excess (d2H-8 d18O, Dansgaard, 1964) of 8.2. However, the isotopic composition of the Tarim River has considerable seasonal and annual variations. According to Liu et al. (1997) and Li et al. (2006), oxygen isotope of the river water from the Upper Tarim River is as depleted as 10.5‰. In the 2008 sampling campaign, the oxygen isotope for the Upper Tarim River (Aral, T1) is 9.9‰. Aksu River, located at a higher altitude, heavy stable isotopes are more depleted with respect to the Kaidu River. The samples from Aksu River (K1, TL, and A1, Fig. 1) show d18O between 10.0‰ and 11.1‰ with the average of 10.6‰. In all, the river water recharging the Lower Tarim River has the isotopic composition ranging from 11.1‰ to 7.8‰ with respect to d18O. The stable isotopes in the water reservoirs in the lower reaches are enriched in heavy isotopes due to evaporation: the Qiala Water Reservoir has the d18O of 4.6‰ and the Daxihaizi Water Reservoir has the d18O of 3.6‰ due to extended evaporation. The residual waters collected from the river bed at the Lower Tarim River show d18O from 2.0‰ to 3.5‰ and d2H from 17.1‰ to 5.8‰, as a consequence of intensive evaporation. 4.2.2. Stable isotopes in groundwaters Riparian groundwater along the Lower Tarim River used to be fed by the river before the cutoff. After the construction of the

4.3. Hydrochemical characteristics 4.3.1. Surface water and groundwater chemistry Hydrochemical data for surface water samples can be divided into four groups: source streams (Aksu and Kaidu River: K1, TL, A1 and R11), rivers (Tarim and Kongque River: T1, R6-9, R10), lakes and water reservoirs (Boston Lake, Qiala and Daxihaizi Water Reservoir: L1, L10, Qiala and Daxihaizi) and residual waters in the Lower Tarim River (R1, R3 and R4) and is all plotted in piper trilinear diagram (Fig. 8). The source streams have very low TDS ranging from 0.132 to 0.217 g/L and are mainly characterized by Ca– Mg–HCO3–SO4 type of chemistry. Hydrochemical type of the Tarim River and Kongque River has changed to Na–Ca–(Mg)–Cl–SO4 facies, and the TDS changes slightly from 0.609 to 0.666 g/L. Boston Lake, Qiala and Daxihaizi Water Reservoir have Na–Mg–Cl–SO4 facies with TDS ranging from 0.754 to 1.053 g/L. The residual waters in the river bed of the Lower Tarim River has a high TDS between 2.64 and 3.82 g/L, with hydrochemical type of Na–Mg–Cl–SO4 and high pH of 9.0 ± 0.2, due to long-time intensive evaporation and dissolution of the salt from the near surface layer of the river bed. The TDS in groundwater samples in the eastern river bed (sections A and B) and in the western (sections C, D, E, F, G, H and I) are all plotted in Fig. 9. The TDS in groundwaters has a large variation ranging from 0.7 g/L to more than 55 g/L (exceeding the detect limit of the eld device). However, TDS in the groundwater within 600 m to the river bed ranges from 0.8 g/L (E3) to 3.7 g/L

Fig. 7. Stable isotopic composition for surface waters and groundwaters.

Fig. 8. Piper trigraph for surface waters.

196

T. Huang, Z. Pang / Journal of Hydrology 387 (2010) 188–201

Fig. 9. TDS distribution (g/L) in the Lower Tarim River.

(C1) with an average of 1.6 g/L, except for section I, which is located in the terminal of the water ow and salt with very high TDS. Beyond the distance, TDS in groundwater is also high except for sample G5. The piper plot for the groundwaters (Fig. 10) shows that the dominant anions are Cl and SO4 while the main cations are Na and Mg, forming geochemical facies of Na–Mg–Cl–SO4 type. The uoride concentrations are high in the Tarim River Basin. In Aksu River, the uoride concentrations range from 0.2 to 1.0 mg/L; in the Tarim River those become higher, ranging from 0.4 to 1.1 mg/L and groundwaters in the Lower Tarim River have the higher concentrations ranging from 0.3 (w20) to 3.7 mg/L (w18) with an average of 1.2 mg/L. The high uoride concentrations in metasomatic deposit with intrusive contact in carbonate and acid igneous, Jurassic coal layers in some cases, and the high frequency of surface water–groundwater interactions in mountainous areas respond for high uoride concentration in the Aksu River (Chen, 2008). The ratios of F to Cl for source stream (the Aksu River) range from 0.008 to 0.100 and that for the groundwaters in the Lower Tarim River range from 0.001 to 0.003 except for w26 of 0.006. Therefore, the high uoride concentration in groundwater is mainly caused by evapoconcentration of river with high uoride concentration rather than local water–rock interaction. The water type conversion from Ca–Mg–HCO3–SO4 with low TDS in source stream to Na–Mg–Cl–SO4 with high TDS in the lower reaches of the basin and uoride concentration evolution show the one of the most typical water evolution of arid water system, going along with evaporation and dissolution.

4.3.2. Groundwater salinization mechanism The groundwater salinity in mainly controlled by dissolution, evapoconcentration and evapotranspiration in arid inland plain regions. Large quantities of salt in the unsaturated zone and aquifers in some cases can cause high salinity, even more than 55 g/L. The soluble salt in the Lower Tarim River is mainly evaporite (NaCl, CaXMg(1X)SO4) and carbonate minerals (CaXMg(1X)CO3) due to long-time concentration and accumulation (Zhu et al., 1981; Zhang et al., 1995). Correlation matrices for TDS and the main ions (Table 6) are used to nd relationship between every two of the variables. There are strongly correlations between the TDS and ions (r > 0.73) except for NO3 and F. Strong correlations exist between TDS and Cl, SO4 and Na (r > 0.95), which only precipitate at very high salinity. The strong correlations between Na and Cl (r = 1.00) and an approximately 1:1 trend (Fig. 11a) suggest Na and Cl mainly comes from halite. Ca and Mg have strong correlations with SO4 (0.97, 1.00) and relatively weak with HCO3 (0.77, 0.84). The equivalent ratios for (Ca + Mg) against HCO3 are more than 1 (Fig. 11b), suggesting that, besides dissolution of carbonates, dissolution of

Table 6 Correlation matrices for showing marked correlation at a signication level of 0.05 (n = 18) for groundwaters. TDS TDS Ca2+ Mg2+ Na+ K+ HCO 3 Cl SO2 4 F NO 3 1 0.95 0.98 1.00 0.92 0.81 1.00 0.97 0.50 0.31 Ca2+ Mg2+ Na+ K+ HCO 3 Cl SO2 4 F NO 3

1 0.97 0.93 0.89 0.77 0.95 0.97 0.37 0.48

1 0.96 0.88 0.84 0.97 1.00 0.35 0.37

1 0.92 0.80 1.00 0.96 0.54 0.28

1 0.73 0.92 0.87 0.57 0.29

1 0.79 0.84 0.26 0.41

1 0.96 0.53 0.29

1 0.33 0.39 1 0.02 1

Fig. 10. Piper trigraph for groundwaters.

Fig. 11. Ionic ratio for the groundwaters.

T. Huang, Z. Pang / Journal of Hydrology 387 (2010) 188–201

197

sulfate (CaXMg (1X) SO4) contributes the additional (Ca + Mg) when concentration of (Ca + Mg) increases with that of SO4. The equivalent ratios for SO4 against HCO3 are all higher than 1, suggesting that quantity of (Ca + Mg) from dissolution of sulfate is more than that from carbonates. There is no further systematic increase in HCO3 and Ca concentration in groundwaters with higher TDS, although the TDS increases signicantly. The saturation indices for the groundwaters are calculated with PHREEQC software (Parkhurst and Appelo, 1999). All groundwaters are undersaturated with respect to gypsum and supersaturated with respect to calcite and dolomite in the Lower Tarim River. Such a pattern indicates that saturation of a relevant salt controls the concentration of the respective ions.

4.3.3. Soil moisture and salt content Near surface soil moisture is low (0.05–2.16%), but generally increases with depth until coming into contact with silt layer (Fig. 3 and Fig. 12), with the depth of 0.55 m for SP1, 1.80 m for SP2 and 3.95 m for SP4. Below the silt layer, the moisture content decreases rapidly before reaching the water table (i.e. SP1 and SP2) or meeting the other silt layer (6.7 m, SP4). It can be concluded that the soil type controls its moisture content. Since chloride is a conservative tracer in water cycle process that can only precipitate in very high concentration and is the major ion in the groundwaters in the Lower Tarim River and has strong correlation with the TDS (Table 6), it is a good tracer to check the salt content in the unsaturated zone. The chloride concentrations in capillary samples for SP1 and SP2 are 487 mg/L and 679 mg/L, less than that in groundwater samples, which are 665–770 mg/L and $728 mg/L, respectively. There is an increasing trend in chloride concentration from the bottom to the top of the proles. The presence of chloride peak in the three proles, reaching 6.1 g/L for SP1, 87.5 g/L for SP2 and 9938 g/L for SP4 (actually most chloride in solid phase), respectively (Fig. 12), is widespread in (semi) arid condition and shown in numerous locations worldwidely (Gates et al., 2008; Phillips, 1994; Scanlon, 1991). Persistently removing the moisture near surface by evapo-transpiration contributes the accumulation of solutes. By comparison with similar environments, such as the Badain Jaran Desert, the diffuse recharge is expected to be very little and less than 1 mm/yr (Gates et al., 2008; Ma and Edmunds, 2006) under such arid condition where annual precipitation is only 40 mm/yr. If the volume weighted average chloride concentration in rainfall of 1.7 mg/L (from 2005 to 2007) for rural monitoring station in the vicinity

of Xi’an, China (EANET, 2008) and the value of 1.5 mg/L measured by Ma and Edmunds (2006) in Badain Jaran Desert and 1.7 mg/L measured by Xu et al. (2009) are adopted, and the aerosol ux deposition is negligible, which is acceptable provided the long-term aerosol ux is near steady state (Goni et al., 2001), the annual chloride input from atmosphere is expected to be 70–100 mg m2 yr1, contributing little to the salinity of the region of high salt content. When considering the potential impact on water diversion and ecological restoration, soil salt and solution give better idea of salt evolution. Fig. 12 shows the chloride concentration in dry soil, appearing to be highly variable. The SP1 has less chloride in the whole prole, ranging from 24 to 862 mg/kg, compared to the SP2 and SP4, which show maximum chloride concentration of 4416 mg/kg (2.05–2.30 m depth) and 11,925 mg/kg (0.35–1.00 m depth), respectively. However, at the bottom of the proles, chloride concentration is rather low. When groundwater table rises within the depth, the chloride concentration in groundwater table would be acceptable for the Tugai vegetation (Song et al., 2000; Thevs, 2007). Taking prole SP4 for an example and assuming groundwater chloride concentration of 400 mg/L, soil bulk density of 1.5 g/cm3, sand porosity of 0.35 and silt porosity of 0.41, the chloride concentration would be 780 mg/L when groundwater table rises to the depth of 5.0 m. However, if ooding the SP4, it would be another story. The ood water would ush the salt into groundwater and the chloride concentration would exceed 6000 mg/L.

4.4. Water table and salinity changes The rst (from 14/05/2000 to 12/07/2000) and the second water diversion (from 03/11/2000 to 05/02/2001) got to Karday (E) and Kargan (I), respectively, and that of the later six water diversions all got to Taitema Lake except for the seventh, which only made it to Yikanbujima (H) (Table 2). The four sections have been used to discuss the changes of the water table and salinity, which are sections B and C, representing the upper segments and G and H, representing the lower segments. Fig. 13 illustrates the changes of the water table and salinity (TDS) at sections B, C, G, and H after the eight water diversions and the data from May, 2000 represents the inventories of pre-diversion. The variations of water tables have the following characteristics: (1) before the water diversion, the water tables were similar at each section and the ow eld was stable; (2) after the water

Fig. 12. Moisture and chloride content for the three soil proles.

198

T. Huang, Z. Pang / Journal of Hydrology 387 (2010) 188–201

Fig. 13. Water tables and TDS changes at certain groundwater monitoring section in the Lower Tarim River under water diversions.

diversion, the water tables rose to different extent. The closer to the river bed, the faster the groundwater table rose. The unstable inltration takes place at river bed (Deng, 2009), as a result, the high groundwater hydraulic head moves towards both river-sides. At the beginning of diversion, the water table close to the river bed rose quickly and after the diversion it decreased due to recharge the adjacent aquifer. Compared with the groundwaters far away from the river, the groundwaters close to river have larger changes in water table. Therefore, from the end of one diversion to the beginning of the next one, redistribution of ow eld and evapotranspiration result in water table drops. Taking July 2002 (before the fourth diversion) and August 2007 (after the eighth diversion) as an example, water tables at sections B and C all decreased. Changes of TDS in groundwaters show the following characteristics: (1) TDS in groundwater from sections B and C has increased after the rst diversion; (2) after the second and third water diversions, TDS in groundwaters from all sections has decreased. (3) The TDS for samples C7 (w8, 850 m away from the river) tempestuously increased from 4.57 g/L to 22.37 g/L after the fourth water diversion, to 15.32 g/L in August, 2007. The TDS in groundwater at certain location is affected by TDS in the diverted water and of the antecedent groundwater, salt content of aquifer and the unsaturated zone as well as groundwater move-

ment in the Lower Tarim River under impulsive water diversions. The TDS in water from the Daxihaizi Water Reservoir was as high as 4.2 g/L and 3.7 g/L before water diversion and during the rst diversion, respectively (Fig. 14), due to dry and salty river bed for long time, as a consequence of the increased TDS close to the river at sections B and C after the rst water diversion. As diverted water continually reached to the Daxihaizi Water Reservoir, the TDS decreased to 1–2 g/L (Fig. 14). The recharging diverted water with relative low TDS diluted the groundwater during the later water

Fig. 14. TDS changes for the Daxihaizi Water Reservoir during water diversions.

T. Huang, Z. Pang / Journal of Hydrology 387 (2010) 188–201

199

diversions. The varying TDS in groundwater at sections B and C had a general decreasing trend except for sample C7 (w8). The tritium contents are between 18.9 and 46.3 TU for B2, B3, B4 at sections B and C2, C4 and C6 at section C, which belong to modern water and suggest some of mixing between the diverted water and the antecedent groundwater. The water table for sample C7 has increased caused by water head pressure transfer rather than diverted water molecules recharge, since groundwater at C7 is premodern water with tritium content of 1.1 TU. From July 2002 to June 2003, the water table had increased about 1 m while the TDS increased from 4.57 to 22.37 g/L for samples C7 dramatically (Fig. 13). The increasing TDS was very likely caused by dissolving salt in unsaturated zone as water table rose, something discussed in detail in Section 4.3.3. For section G, tritium content for G2 (50 m away from the river bed) is 19.6 TU and G5 (800 m away from the river bed) is 2.4 TU as pre-modern water. The water tables of both boreholes have risen. The TDS for G2 decreased rapidly at the earlier stages, and stabilizing at 1 g/L. The rising water table of G5 is also caused by water head pressure transfer. The decreasing TDS for G5 suggests that the closer to the river bed, the lower TDS in groundwater is. The groundwater was recharged by the Tarim River with low TDS in the past, when groundwater depth was deeper (more than 8 m) and evapotranspiration was less intensive before the water diversions. Section G is the conuence reaches of the old Tarim River and the Tarim River (Fig. 4), therefore relative uent runoff condition and deep groundwater depth of about 11 m before diversion with little evaporation mostly respond for low TDS in groundwater. As a result, the decreasing TDS at section G occurs under water diversion. The same process occurs at section H. Boreholes H1 and H2, 100 m and 300 m away from the river, with tritium content of 29.0 TU and 30.0 TU, respectively, have a variation from 0.94 to 1.75 g/L in TDS. The water table for sample H3 with tritium content of 4.1 TU has risen about 2 m. Meanwhile, TDS has a little decreasing trend, as well as for sample G5. 4.5. Groundwater table changes relative to vegetation The variation of groundwater table between that before water diversion (05/2000) and after the eighth the water diversion (08/ 2007) and the groundwater depth (08/2007) are plotted in Fig. 15 using the Inverse Distance Weighting Method. Groundwater depth in three boreholes (E5, I2, I3), located far way from the river and in the lower segments, has unusually increased, most probably being high water table observation (the perched water) before the

diversion since the water table were persistently decreasing during water diversion. The left groundwater depths are all decreasing via the implementation of eight intermittent water diversions, water tables rising by more than 1 m within 700 m from the river in the upper segments to 300 m in the lower (Fig. 15a). Greater water tables rising occurs nearby the river channel, which can reach 6.7 m (G2). The scopes of water tables rise are wider than modern recharge limit, due to water head pressure transfer. Since the vegetation coverage exhibits a continued declining trend with dropping water table (Hao et al., 2010) and modern water recharge scope (Pang et al., 2010) in the area, Niu and Li (2008) concluded there was 7345 hm2 increased vegetation area from 2000 to 2006 after water diversion based on satellite images of the main area where vegetation distributed in the Lower Tarim River. However, the groundwater depth suitable for the growth of P. euphratica is less than 5 m in the Lower Tarim River, according to the relationship among groundwater quality (Chen et al., 2008b; Xu et al., 2007) and proline accumulation (Chen et al., 2003). If the target of the ecological restoration is to maintain the vegetation dominated by P. euphratica, the optimal groundwater depth would be less than 5 m. However, the territory with the suitable groundwater depth is narrow, within 200 m from the river bed and becomes even narrower towards downstream (Fig. 15b), far way from most of the existing ecosystem at 1000–1500 m. The desert riparian ecological forest formed in the historical period under larger river recharge is still diminishing. According to the Program of ‘Recent Tarim River Basin Comprehensive Management’ announced by the Xinjiang Government and Ministry of Water Resources of China in 2001 (XGMWRC, 2002), the aims of ecosystem recovery are to convert 22,000 ha agriculture led to natural ecosystem in the Upper/Middle Tarim River, to reduce the area of high water consuming crop (e.g. paddy), to ensure that the runoff to the Daxihaizi Water Reservoir increases to 3.5 108 m3/yr and the diverted water ows into the Taitema Lake, so as to improve the eco-environment in the Lower Tarim River. Although groundwater depth, water salinity, status of the riparian vegetation has measurable positive changes, it is still far away from meeting the requirements of the suitable groundwater depth (less than 5 m within riparian zone of 1000–1500 m from the river bed) for the survival of the existing P. euphratica and Tamarix ramosissima, the main species targeted by the rescue effort in the Lower Tarim River. As the groundwater depth less than 5 m is spatially limited to within 200 m from the river bank, and it narrows down towards downstream, long-term stability of the ecosystem cannot be achieved by the current water diversion scheme. Due to climate variation (e.g. the runoff of Aral in the Tarim River deceased by

Fig. 15. The variation of groundwater table (m) before water diversion and after the eighth water diversion (a) and groundwater depth (m) in the Lower Tarim River (August, 2007).

200

T. Huang, Z. Pang / Journal of Hydrology 387 (2010) 188–201 Bennetts, D.A., Webb, J.A., Stone, D.J.M., Hill, D.M., 2006. Understanding the salinisation process for groundwater in an area of south-eastern Australia, using hydrochemical and isotopic evidence. Journal of Hydrology 323, 178–192. Cai, C., Bo, M., Ma, T., Chen, C., Li, W., Liu, C., 1997. Approach to Fluid–Rock Interaction in the Tarim Basin. Geological Publishing House, Beijing. 155pp. Chapman, J.B., Lewis, B., Litus, G., 2008. Chemical and isotopic evaluation of water sources to the fens of South Park, Colorado. Environmental Geology 43, 533– 545. Chen, J., 2008. The formation mechanism of high uoride concentration in groundwater in alluvial/diluvial fan of Tailan River, Wensu. Henan Water Resources & South-to-North Water Diversion 53 (2), 59–61. Chen, Y.N., Chen, Y.P., Li, W., Zhang, H., 2003. Response of the accumulation of proline in the bodies of Populus euphratica to the change of ground water level at the lower reaches of Tarim River. Chinese Science Bulletin 48 (18), 1995– 1999. Chen, Y.J., Zhou, K., Chen, Y.N., Li, W., Liu, J., Wang, T., 2008a. Response of groundwater chemistry to water deliveries in the lower reaches of Tarim River, Northwest China. Environmental Geology 53, 1365–1373. Chen, Y.N., Pang, Z., Chen, Y.P., Li, W., Xu, C., Hao, X., Huang, X., Huang, T., Ye, Z., 2008b. Response of riparian vegetation to water-table changes in the lower reaches of Tarim River, Xinjiang Uygur, China. Hydrogeology Journal 16, 1371– 1379. Cheng, Q., 1993. Tarim River Research. Hohai University Press, Nanjing. 246pp. Craig, H., 1961. Isotopic variations in meteoric waters. Science 133, 1702–1703. Dansgaard, W., 1964. Stable isotope in precipitation. Tellus 16, 436–468. Deng, M., 2009. Theory and Practice of Water Management in the Tarim River, China. Science Press, Beijing. 564pp. Dong, X., Deng, M., 2005. Xinjiang Groundwater Resources. Xinjiang Science and Technique Press, Urumqi, China. 193pp. Eamus, D., Froend, R., Loomes, R., Hose, G., Murray, B., 2006. A functional methodology for determining the groundwater regime needed to maintain the health of groundwater-dependent vegetation. Australian Journal of Botany 54, 97–114. EANET-Acid Deposition Monitoring Network in East Asia, 2008. EANET data on the acid deposition in the East Asian region. <http://www.eanet.cc> (accessed December 2008). Edmunds, W.M., 2009. Geochemistry’s vital contribution to solving water resource problems. Applied Geochemistry 24, 1058–1073. Fan, Z., Ma, Y., Zhang, H., Du, L., 2002. Salinization and improvement ways of water quality of Tarim River, Xinjiang, China. Advances in Water Science 13 (6), 719– 725. Feng, Q., Liu, W., Si, J., Su, Y., Zhang, Y., Cang, Z., Xi, H., 2005. Environmental effects of water resources development and use in the Tarim River basin of northwestern China. Environmental Geology 48, 202–210. Gates, J.B., Edmunds, W.M., Ma, J., Scanlon, B.R., 2008. Estimating groundwater recharge in a cold desert environment in northern China using chloride. Hydrogeology Journal 16, 893–910. Gonantini, R., 1986. Environmental isotopes in lake studies. In: Fritz, P., Fontes, J.Ch. (Eds.), Handbook of Environmental Isotope Geochemistry, vol. 2. Elsevier, Amsterdam, pp. 113–168. Goni, I.B., Fellman, E., Edmunds, W.M., 2001. Rainfall geochemistry in the Sahel region of northern Nigeria. Atmospheric Environment 35, 4331–4339. Gremmen, N.J.M., Reijnen, M.J.S.M., Wiertz, J., van Wirdum, G., 1990. A model to predict and assess the effects of groundwater withdrawal on the vegetation in the Pleistocene areas of The Netherlands. Journal of Environmental Management 31 (2), 143–155. Hao, X., Li, W., Huang, X., Zhu, C., Ma, J., 2010. Assessment of the groundwater threshold of desert riparian forest vegetation along the middle and lower reaches of the Tarim River, China. Hydrological Processes 24, 178–186. He, X., Shao, D., Liu, W., Dai, T., 2006. Review of the researches on utilization of farmland drainage as resources. Transactions of the CSAE 22 (3), 176–179. Hou, P., Beeton, R.J., Carter, S., Dong, X., Li, X., 2007a. Response to environmental ows in the lower Tarim River, Xinjiang, China: ground water. Journal of Environmental Management 83, 371–382. Hou, P., Beeton, R.J., Carter, S., Dong, X., Li, X., 2007b. Response to environmental ows in the Lower Tarim River, Xinjiang, China: an ecological interpretation of water-table dynamics. Journal of Environmental Management 83, 383–391. IAEA and WMO, 2006. Global Network of Isotopes in Precipitation. The GNIP Database. <http://isohis.iaea.org>. Jiao, P., Wang, E., Liu, C., 2004. Characteristics and origin of tritium in the potassium-rich brine in Lop Nur, Xinjiang. Nuclear Techniques 27 (9), 710–715. Klein, L.R., Clayton, S.R., Alldredge, J.R., Goodwin, P., 2007. Long-term monitoring and evaluation of the Lower Red River Meadow Restoration Project, Idaho, USA. Restoration Ecology 15 (2), 223–239. Li, W., Hao, A., Liu, Z., Wan, L., Guo, J., 2000. Perspective Areas for Groundwater Development in Tarim Basin. Geological Publishing House, Beijing. 194pp. Li, W., Hao, A., Zheng, Y., Liu, B., Yu, D., 2006. Regional environmental isotopic features of groundwater and their hydrogeological explanation in the Tarim Basin. Earth Science Frontiers 13 (1), 191–198. Liu, J., 2001. Fluorine concentration changing tendency study of china atmospheric precipitation in the recent 10 years. Site Investigation Science and Technology 4, 11–19. Liu, D., Liu, S., Xu, Z., 1997. Environmental isotope studies on shallow groundwater in the lower Tarim River, Xinjiang. Journal of Chengdu University of Technology 24 (3), 89–95.

$50% from October 2007 to September 2008), not decreased agricultural area and not saving water in source-stream and the Upper/Middle Tarim River, the water diversion project to the Lowe Tarim River has been interrupted, especially since 2008. The embarrassment for the water diversion is mainly caused by imbalance water utilization between upper streams and lower streams, unfair water allocation between eco-environmental water requirement and agriculture water use. This is not only a scientic, but also a socioeconomic issue in the Tarim River, since the Program has given more specic investment and relevant measurements. 5. Conclusions Stable isotopes, tritium and water chemistry in the Lower Tarim River provide insight into basic hydrological and geochemical processes. The isotopic composition of shallow groundwater forms a trend line that is almost in parallel to the GMWL but is enriched in heavy isotopes compared with the recharging river water. This can be attributed to evaporation during the river recharge to the riparian groundwater system in a rather uniform manner. Tritium data show that the extent of modern recharge (since 1960) is limited to 600–200 m from the river bank with a descending trend towards downstream. The Lower Tarim River is the terminal for both water and salt of the basin, and the dissolution of salts, such as evaporites and carbonate minerals, is the main geochemical process controlling groundwater salinity. Water table has risen after the eight water diversions except for three unusual boreholes where perched water probably exists. The scope with more than 1 m of water table rise is from 700 m from the riverbank in the upper segments to 300 m in the lower segments. Groundwater salinity at certain locations is mainly affected by the salinity of the diverted river water and of the local antecedent groundwater, salts in the unsaturated zone, evapotranspiration during the diverted water relocation. The TDS of groundwater has generally decreased after the water diversions. The thirsty Lower Tarim River needs more water. Regulating and saving water in the Upper/Middle Tarim River is crucial for continuing water diversion. Furthermore, monitoring of groundwater should be continued to assess how fast the 2.27 billion m3 of water diverted will evaporate and how the hydrological regime will change. Extend investigations, especially the soil moisture, soil salt content, stable isotopes and tritium should be deployed to study the hydrological, geochemical and biological processes to evaluate water and soil conditions simultaneously and to test various plant communities for anti-salt and anti-aridity species. Acknowledgments The research is supported by the Knowledge Innovation Program of the Chinese Academy of Sciences (kzcx2-yw-127) and National Science Foundation of China (No. 40672171). The authors wish to thank Yaning Chen, Xiaoming Zhou, Chenggang Zhu, Xinhe Jiang and Zhongfeng Duan for their assistance in the eld work, Bing Xu for tritium analyses and Yiman Li for part of the water chemistry analyses. References
Adams, S., Titus, R., Pietersen, K., Tredoux, G., Harris, C., 2001. Hydrochemical characteristics of aquifers near Sutherland in the Western Karoo, South Africa. Journal of Hydrology 241, 91–103. Alim, T., Xu, W., 2003. Study on the groundwater movement in the vicinities along the river channel of stream water transportation to the lower reaches of Tarim River, Xinjiang. Arid Land Geography 26 (2), 129–135. Baillie, M.N., Hogan, J.F., Ekwurzel, B., Wahi, A.K., Eastoe, C.J., 2007. Quantifying water sources to a semiarid riparian ecosystem, San Pedro River, Arizona. Journal of Geophysical Research, G03S02. doi:10.1029/2006JG000263.

T. Huang, Z. Pang / Journal of Hydrology 387 (2010) 188–201 Loheide II, S.P., Gorelick, S.M., 2007. Riparian hydroecology: a coupled model of the observed interactions between groundwater ow and meadow vegetation patterning. Water Resources Research 43, W07414. doi:10.1029/ 2006WR005233. Lymbery, A.J., Doup, R.G., Pettit, N.E., 2003. Effects of salinisation on riparian plant communities in experimental catchments on the Collie River, Western Australia. Australian Journal of Botany 51, 667–672. Ma, J., Edmunds, W.M., 2006. Groundwater and lake evolution in the Badain Jaran Desert ecosystem, Inner Mongolia. Hydrogeology Journal 14, 1231–1243. Ma, J., Wang, X., Edmunds, W.M., 2005. The characteristics of groundwater resources and their changes under the impacts of human activity in the arid North-West China – a case study of the Shiyang River Basin. Journal of Arid Environments 61, 277–295. Ma, J., Ding, Z., Edmunds, W.M., Gates, J.B., Huang, T., 2009. Limits to recharge of groundwater from Tibetan Plateau to the Gobi desert, implications for water management in the Mountain front. Journal of Hydrology 364, 128–141. Manchanda, G., Garg, N., 2008. Salinity and its effects on the functional biology of legumes. Acta Physiologiae Plantarum 30, 595–618. Michel, R.I., 2005. Tritium in the hydrologic cycle. In: Aggarwal, P.K., Gat, J.R., Froehlich, K.F.O. (Eds.), Isotopes in the Water Cycle: Past, Present and Future of a Developing Science. Springerlink, Netherlands, pp. 53–66. Niu, T., Li, X., 2008. Study on landscape change and vegetation restoration in the lower reaches of Tarim River. Remote Sensing Information 5, 68–73. Pang, Z., Huang, T., Chen, Y., 2010. Diminished groundwater recharge and circulation relative to degrading riparian vegetation in the middle Tarim River, Xinjiang Uygur, Western China. Hydrological Processes 24, 147–159. Parkhurst, D.L., Appelo, C.A.J., 1999. User’s Guide to PHREEQC (Version 2) – A Compute Program for Speciation, Batch-reaction, One-dimensional Transport, and Inverse Geochemical Calculations. USGS, 312pp. Phillips, F.M., 1994. Environmental tracers for water movement in desert soils of the American Southwest. Soil Science Society of America Journal 58, 14–24. Prasanna, M.V., Chidambaram, S., Hameed, A.S., Srinivasamoorthy, K., 2009. Study of evaluation of groundwater in Gadilam basin using hydrogeochemical and isotope data. Environmental Monitoring and Assessment. doi:10.1007/s10661009-1092-5. Richardson, D.M., Holmes, P.M., Esler, K.J., Galatowitsch, S.M., Stromberg, J.C., Kirkman, S.P., Pysek, P., Hobbs, R.J., 2007. Riparian vegetation: degradation, alien plant invasions, and restoration prospects. Diversity and Distributions 13, 126–139. Scanlon, B.R., 1991. Evaluation of moisture ux from chloride data in desert soils. Journal of Hydrology 128, 137–156. Shen, Y., Wang, S., 2002. New progress in glacier and water resources changes in Tarim Basin, Xinjiang. Journal of Glaciology and Geocryology 24 (6), p819. Shen, Z., Zhu, Y., Zhong, Y., 1993. Hydrogeochemistry. Geological Publishing House, Beijing, China. 189pp.

201

Song, Y., Fan, Z., Lei, Z., 2000. Research on Water Resources and Ecology of Tarim River, China. Xinjiang People’s Press, Urumqi, China. 481pp. Stromberg, J.C., Tiller, R., Richter, B., 1996. Effects of groundwater decline on riparian vegetation of semiarid region: the San Pedro, Arizona. Ecological Applications 6 (1), 113–131. Stromberg, J.C., Beauchamp, V.B., Dixon, M.D., Lite, S.J., Paradzick, C., 2007. Importance of low-ow and high-ow characteristics to restoration of riparian vegetation along rivers in arid south-western United States. Freshwater Biology 52, 651–679. Tamea, S., Laio, F., Ridol, L., D’Odorico, P., Rodriguez-Iturbe, I., 2009. Ecohydrology of groundwater-dependent ecosystems: 2. Stochastic soil moisture dynamics. Water Resources Research 45, W05420. doi:10.1029/2008WR007293. Tao, H., Gemmer, M., Song, Y., Jiang, T., 2008. Ecohydrological responses on water diversion in the lower reaches of the Tarim River, China. Water Resources Research 44, W08422. doi:10.1029/2007WR006186. Thevs, N., 2007. Ecology, Spatial Distribution, and Utilization of the Tugai Vegetation at the Middle Reaches of the Tarim River, Xinjiang/China. Cuvillier Verlag Goettingen. 180pp. Wang, G., Cheng, G., 2000. The characteristics of water resources and the changes of the hydrological process and environment in the arid zone of northwest China. Environmental Geology 39, 783–790. Williams, D.G., Scott, R.L., Huxman, T.E., Goodrich, D.C., Lin, G., 2006. Sensitivity of riparian ecosystems in arid and semiarid environments to moisture pulses. Hydrological Processes 20, 3191–3205. XETCAS-Xinjiang Expedition Team of the Chinese Academy of Sciences, 1965. Groundwater in Xinjiang. Science Press, Beijing. pp. 25–32. Xinjiang Government and Ministry of Water Resources of China (XGMWRC), 2002. The Program of Recent Tarim River Basin Comprehensive Management. China Water & Hydropower Press, Beijing. 93pp. Xu, H., Ye, M., Song, Y., Chen, Y., 2007. The natural vegetation responses to the groundwater change resulting from ecological water conveyances to the Lower Tarim River. Environmental Monitoring and Assessment 131, 37–48. Xu, Z., Li, Y., Tang, Y., Han, G., 2009. Chemical and strontium isotope characterization of rainwater at an urban site in Loess Plateau, Northwest China. Atmospheric Research 94, 481–490. Zhang, J., Takahashi, K., Wushiki, H., Yabuki, S., Xiong, J., Masuda, A., 1995. Water geochemistry of the rivers around the Taklimakan Desert (NW China): crustal weathering and evaporation process in arid land. Chemical Geology 119, 225– 237. Zhang, Z., Shen, Z., Xue, Y., 2000. Evolution of Groundwater Environment of the Northern China Great Plain. Geological Publishing House, Beijing. 281pp.. Zhu, B., Yang, X., 2007. The ion chemistry of surface and ground waters in the Taklimakan Desert of Tarim Basin, Western China. Chinese Science Bulletin 52, 2123–2129. Zhu, Z., Chen, Z., Wu, Z., 1981. Studies on Aeolian Landforms in the Taklimakan Desert. Science Press, Beijing. 110pp.


相关文章:
更多相关标签: