Dynamic evaluation of groundwater resources in Zhangye Basin

LiNa Mi， HongLang Xiao， ZhengLiang Yin， ShengChun Xiao

1. Key Laboratory of Ecohydrology of Inland River Basin， Cold and Arid Regions Environmental and Engineering Research Institute， Chinese Academy of Sciences / Heihe Key Laboratory of Ecohydrology and Integrated River Basin Science， Lanzhou， Gansu 730000， China

2. University of Chinese Academy of Sciences， Beijing 100049， China

?

Dynamic evaluation of groundwater resources in Zhangye Basin

LiNa Mi1，2*， HongLang Xiao1， ZhengLiang Yin1， ShengChun Xiao1

1. Key Laboratory of Ecohydrology of Inland River Basin， Cold and Arid Regions Environmental and Engineering Research Institute， Chinese Academy of Sciences / Heihe Key Laboratory of Ecohydrology and Integrated River Basin Science， Lanzhou， Gansu 730000， China

2. University of Chinese Academy of Sciences， Beijing 100049， China

ABSTRACT

Groundwater resource is vital to the sustainable development of socio-economics in arid and semi-arid regions of Northwest China. An estimation of the groundwater resources variation in Zhangye Basin was made during 1985-2013 based on long-term groundwater observation data and geostatistical method. The results show that from 1985 to 2013，groundwater storage exhibited tremendous dissimilarity on temporal and spatial scale for the whole Zhangye Basin，especially before and after implementation of the water diversion policy. Trend of groundwater storage varied from quick to slow decline or increase. The accumulative groundwater storage decreased nearly 47.52×108m3， and annual average depletion rate reached 1.64×108m3/a. Among which， the accumulative groundwater storage of the river and well water mixed irrigation district decreased by 37.48×108m3， accounting for about 78.87% of the total groundwater depletion of the Zhangye Basin. Accumulative depletion of groundwater storage varied in respective irrigation districts. Though groundwater resources depletion rate slowed down from 2005， the overall storage in the whole basin and respective districts during 1985-2013 was still in a severe deficit such that， the groundwater resource was in a rather negative balance， which could threaten the local aquifer. This is the joint effect of climate change and human activities，however human activities， such as water diversion policy and groundwater exploitation， became increasingly intense. Our research results could provide a reasonable estimation for the groundwater balance in Zhangye Basin， providing a scientific basis for water resources unified planning and， this method can provide a relatively reliable way of estimation for large scale groundwater resources.

groundwater resources； dynamic evaluation； groundwater storage； Zhangye Basin

1　Introduction

At present the world is facing a severe water shortage， where water resources have become a worldwide major issue concerning national economic levels （Xiao and Cheng， 2006）. It is reported that in some inland river basins， including the Tarim River and the Hexi Inland Rivers， ecological water shortage reached about 47 billion m3， accounting for nearly 85.0% of the total water shortage in China （Water Resources Strategy Research Group of the Chinese Academy of Sciences， 2009）. Poorly planned water use combined with the unbalanced temporal and spatial distribution of water resources and water-landassembly have made water scarcity a critical challenge to the economic development and ecological protection in inland river basins of northwestern China （Cheng et al.， 2006； Hu et al.， 2009； Xiao et al.，2015）. According to figures released from the Water Resources Gazette of Gansu Province （1999）， water use efficiency in the midstream of the Heihe River Basin （HRB） is 119.9%， accounting for 68.1% of the total water resources of the HRB （Lan et al.， 2002）. This way of water resources use may inevitably cause water shortages and ecological degradation in the downstream of Ejina Oasis， and water use contradictions between the middle and lower reaches are intensified. Due to the growing shortage of water resources in northwestern China， the National Ministry of Water Resources implemented the policy-based Ecological Water Transfer Project （EWTP） in the Yellow River，Heihe River and Tarim River Basins in 2001 （Zhang et al.， 2003）. The EWTP in the HRB requires that in a normal year the downstream surface water transfer should be at least 9.5×108m3. Implementation of the EWTP has alleviated the trend of land degradation and ecological deterioration in the downstream of the HRB （Zhang et al.， 2011）. However， the amount of surface water entering the midstream has decreased from 10.4×108m3at the end of the 1990s to 8.0×108m3in 2010 （Nian et al.， 2014， Mi et al.， 2015）. Meanwhile， with agriculture development， cultivated land in the midstream increased from 20.90×104ha at the end of 1980s to 24.95×104ha in 2010， requiring more irrigation water and， policy-based water scarcity appeared. To cope with water scarcity， drastic groundwater drilling from unconfined and confined aquifers appeared in sporadic areas， such that groundwater abstraction increased drastically from 0.56×108m3/a at the end of the 1980s to 4.97×108m3/a in 2010 in Zhangye Basin. Aquifer depletion and groundwater levels decline， local springs dried out，and groundwater quality deterioration occurred in the midstream of HRB （Li， 1983； Chen， 1984； Chen and Qu， 1992； Wang et al.， 1999； Tang and Zhang， 2001；Ding and Zhang， 2002； Liu et al.， 2002； Wen et al.，2007）. Problems and conflicts on water resources utilization and ecological environment security between the midstream and downstream became more complicated and， water resources evaluation in the HRB became more difficult （Lan et al.， 2000）. Unwise use of surface water and groundwater will unbalance the water and minerals dissolved in the system， and may damage the ecological environment （Cui and Shao，2005）. Observations show that the groundwater level generally declined 4.92-11.49 m， reaching a maximum of 17.44 m in nearly 30 years over the midstream of the HRB because of climate change and human activities （Mi et al.， 2015）. According to the risk assessment of groundwater mining， Zhangye Basin belongs to the moderate risk region， and the exploitation potential severely insufficient in four irrigation districts of Luocheng， Pingchuan， Youlian and Liuba because of excessive mining （Xiang， 2011）. Long-term overexploitation of groundwater can not only cause groundwater aquifer depletion， but affect the whole river basin ecosystem stability and healthy development. There have been numerous related studies in the HRB centered on water resources development， management and its relationship with economic development （Zhang et al.， 2003； Fang et al.， 2007； Zhang et al.， 2011； Cai et al.， 2014 ）， water environment change （Qi and Luo， 2005， 2007； Ji et al.， 2006； Zhu et al.， 2008； Zhou et al.， 2013）， land use change and its environmental effects （Lu et al.，2003； Wang et al.， 2007； Feng et al.， 2013； Xiao et al.， 2015）. On groundwater dynamic aspects， some researchers studied the groundwater storage changes in the short or medium term. Yang and Wang （2005）used conventional statistical methods to estimate the temporal and spatial variation of groundwater level and storage of different partitions of Zhangye Basin during 1981-2001. They concluded that groundwater storage in the alluvial fan and the fine soil plain declined， while increasing in the valley plain to some extent. Wei et al. （2009） calculated the change of groundwater storage before and after the water diversion， and showed that the total exploitable-storage in Ganzhou， Linze and Gaotai counties during 2004-2005 decreased to about 8.23×108m3in contrast to the 1990s before water diversion. Cao and Nan（2011） used GRACE （the Gravity Recovery and Climate Experiment） gravity satellite data to estimate groundwater level and storage respectively in the HRB. GRACE is more suitable for estimating groundwater resources at a large spatial scale. However， Yang and Wang （2005） estimated groundwater dynamics based on single well's groundwater level variation through traditional statistics during 1981-2001， while Wei et al. （2009） estimate was over a relatively short term. Continuous research of groundwater resources at an irrigation district scale from 1985 to 2013 based on the geostatistical method in the Zhangye Basin has not been reported， especially in recent 10 years. Furthermore， the geostatistical method is popular in hydrogeological research. Therefore， the study area of this paper is the core irrigation area of the Zhangye Basin which is significantly affected by human activity. We estimated the groundwater level drawdown （GLD） using Ordinary Kriging and， evaluated the groundwater resources variation over the past 30 years. The main purpose is to provide an understanding of the behavior of the groundwater body and its long-term evolution trends in Zhangye Basin， and insights into groundwater system changes. This evaluation is very critical for thelarge scaled river basin's water resources sustainable management. The results presented herein will serve as a valuable scientific basis for the river basin's ecosystem stability and healthy development.

2　Study area and hydrogeological settings

Situated in the midstream of the HRB， Zhangye Basin belongs to the central part of the Hexi Corridor， adjoined to Yonggu hills to the east， Yumu Mountain to the west， bounded by the Qilian Mountains to the south， Heli and Longshou mountains to the north， with the basement composed of new tertiary or cretaceous bedrock （Figure 1）. The Zhangye Basin is characterized by a desert-artificial oasis composite ecosystem with five main counties： Ganzhou， Linze， Gaotai， Minle， and Shandan. Altitude in the Zhangye Basin ranges from 1，309 m to 1，459 m，annual average temperature is 2.8-7.7 °C， annual precipitation is about 90-160 mm and evaporation is 2，000-2，500 mm. The basin receives little precipitation and is subjected to strong evaporation. Surface water in this region is important for economic development and ecological balance of the middle and lower reaches. The total surface runoff volume is about 25.01×108m3/a， primarily derived from precipitation and glacial melt-water in Qilian Mountains via 29 sub-rivers from west to east. The main sub-rivers are the Liyuan and Heihe rivers and its tributaries，with a total volume of 22.26×108m3/a （Figure 1）. Heihe River originates from Qilian Mountains in the upstream， travels into Zhangye Basin through Yingluoxia gauge station to the south， and exits through Zhengyixia gauge station to the north， and ultimately drained into terminal lakes of the downstream in Ejina Basin. Heihe is the largest river with a total length of 821 km and average runoff volume of 16.0×108m3in the HRB.

Figure 1　Study area and distribution of observation wells

From the perspective of geomorphology， Zhangye basin is composed of two parts： the southern piedmont alluvial-diluvial Gobi plain and central diluvial fine soil plain， and the terrain slopes from southeast to northwest by 25‰-4‰. From south to north， sediments change from coarse-grained gravel to medium and fine-grained sand and silt. These sediments form the main aquifers with thickness ranging from 300-500 m in the south to 100-200 m in the northern edge of the alluvial fan. The aquifer transitions from a single-layered shallow aquifer to a multi-layered confined aquifer to the northern basin center accordingly， and the groundwater depth becomes shallow and gradually transitions to the exposed spring area （Gao and Li， 1990）. Modern groundwater is mainly recharged from the limited surface water via river infiltration， canal system seepage and infiltration of irrigation returned flow， accounting for more than 70% of the total surface water infiltration into the aquifer under natural conditions （Wen et al.，2008）. Spring water exposure and artificial exploitation are the main discharge. There is a close interconnection between surface water and groundwater through river infiltration and groundwater discharge（Feng et al.， 2000）.

Irrigation agriculture occupies about 82.6% of the core artificial oasis in the Zhangye Basin， which is the main water consumption region with high degree of water resources utilization. Irrigation history in the Zhangye Basin can be dated back to over 2000 yearsago （Zhong et al.， 2011）. According to statistics of the Zhangye Annual Water Resource Management Reports， mean annual total water use in the Zhangye Basin reached 23.3×108m3/a， of which over 85.0% was used for agricultural irrigation. Meanwhile，groundwater exploitation and use continuously intensifies where the volume increased from 0.56×108m3/a in the 1980s to 4.97×108m3/a in 2010.

3　Materials and methods

3.1Data sources

Data used in this study consisted of 1：500，000 hydrogeological data， 30 m Digital Elevation Model（DEM）， and irrigation districts distribution data were obtained from the Cold and Arid Regions Environmental and Engineering Research Institute（CAREERI）， Chinese Academy of Science（http：//westdc.westgis.ac.cn/）. Runoff observation data and groundwater observation data were obtained from the Gansu Provincial Bureau of Hydrology. Fifty-four groundwater observation wells with normal， continuous annual and monthly average observation data （1985-2013） were selected （Figure 1）. Water diversion data， irrigation and groundwater exploitation data for each irrigation district was collected from the Annual Water Resource Management Reports（1980-2010） published by the Zhangye Municipal Bureau of Water Conservancy.

3.2Spatial partitions

There are 27 small irrigation districts in the Zhangye Basin， with river， spring and groundwater as the main source of irrigation water. Irrigation water in Ganzhou irrigation districts mainly depend on river water， river and well water， and the proportion of groundwater abstraction is relatively large. With an increasing proportion of well water， spring water overflow becomes increasingly smaller. Irrigation water in Linze and Gaotai shifts from river and spring water to river and well water， especially the Luotuocheng district， mainly owns to well water irrigation. Irrigation water in Minle irrigation districts mainly depend on river water， and river and well water. Irrigation water types and their sources in the Zhangye Basin are presented in Table 1. According to 1：500，000 hydrogeological map combined with irrigation district boundary， the hydrogeological boundary of the Zhangye Basin was determined with a total area of 7，936.04 km2， with five partitions of river and well water mixed irrigation district（RWMID）， river water irrigation district （RID）， well water irrigation district （WID）， spring water irrigation district （SID） and river and spring water mixed irrigation district （RSMID） were made based on the type of irrigation water of respective small irrigation districts （Figure 1）.

3.3Method and estimate precision

Three steps were needed to provide groundwater storage： （1） Annual groundwater level drawdown（GLD） distributions from 1985 to 2013 were achieved by Ordinary Kriging through ArcGIS Geostatistical Analyst module， evaluated the estimate precision and then， converted to 730m×730m grid respectively（Figure 2）； （2） Using the ArcGIS Spatial Analyst Tools to extract the annual average GLDs and their area in every partition by the respective spatial partition boundary obtained earlier in 3.2； （3） Equation （2）was used to calculate the annual average and accumulative groundwater storage variations.

3.3.1Annual groundwater level drawdown (GLD) interpolation and the estimate precision

The annual GLD is the mean groundwater level difference between the base year and the year before for each observation well. For example， the GLD of well Z08 in 1986 is the mean groundwater level difference between 1986 and 1985， and so on. In this calculation， every year can be a base year. The aquifer in the Zhange Basin was considered to be a whole for the reason that the groundwater levels of the observation wells were all mixed groundwater levels， before Ordinary Kriging interpolation. This ensured that the Histogram and the QQplot was a normal distribution and meet the Ordinary Kriging interpolation condition. Spherical model or exponential model was selected to fit the semi-variance function.

If the random function Z（x） satisfies the hypothesis：（1） the mean value exists and does not depend on x，that is E［Z（x）］=m，?x； （2） for any distance h， variables［Z（x+h）-Z（x）］ with a finite variance do not depend on x， then for any x and h， the semi-variogram was：

The estimate error was assessed by cross-validation function through the ArcGIS Geostatistical Analyst module. For a model that provides accurate predictions， the Mean error and Standardized Mean error （SM） should be close to 0， the Root-Mean-Square error （RMS） and Average Standard error （AS） should be as small as possible， and the Root-Mean Square Standardized error （RMSS） should be close to 1 （Johnston et al.， 2001）. The cross-validation results show that the Mean error and the SM error ranged from -0.042 to 0.037， close to 0；the RMS ranged from 0.295 to 0.946， as small as AS error， and the RMSS error ranged from 1.028 to 1.315， almost close to 1. It was an ideal interpolation performance for such a large region. Therefore， the interpolation results show that the semi-variogram could reasonably represent the spatial structure of the GLD， thus the spatial distribution maps of the annual GLD over 29 years were built （Figure 2）. Because of long data set， only some typical year's spatial distribution maps were listed here. The spatial distribution of annual groundwater level drawdown indicated a strong dissimilarity over the whole basin.

Table 1　Irrigation water types and their source in the Zhangye Basin

3.3.2Groundwater storage estimate

The groundwater storage can be calculated using the equation outlined （Zhang and Zeng， 1994）：

In Equation （2）， ΔWkiis the groundwater storage variable for the partition k in year i， k=1， 2， 3， 4， 5，with five partitions respectively； i is from 1985 to 2013， and μkis the specific yield of an aquifer for partition k， dimensionless. According to the data from Gansu Provincial Geological Survey Bureau， the value of μkranges from 0.10-0.15. In this study，combined with the hydrogeological map， μkof RID and WID which are located in the upper and middle of the alluvial fan was given the value of 0.15，RSMID and SID in the valley plain was given the value 0.11. Since some of the RWMID are located in the alluvial fan， and some are in the valley plain and fine soil plain， μkof the RWMID was given the value of 0.12. ΔHkijis groundwater level drawdown value for grid j in partition k of year i，is the mean groundwater level drawdown of partition k calculated by the ArcGIS Spatial Analyst module and Fkis the area of partition k， which is the total area of grid j.

Figure 2　Annual groundwater level drawdown distribution maps

4　Results and discussion

Groundwater storage changes are the result of water table change， as well as the direct reflection of the relationship between groundwater recharge and discharge. In the Zhangye Basin， the temporal and spatial variations of GLD and storage show significant dissimilarity in different partitions.

In order to better explore the process of how groundwater system varied under the effects of climate change and human activities， a five-time-stage division from the temporal scale was generated. With the development of irrigated agriculture， agricultural water requirements increased， especially before and after EWTP （1998-2001， 2002-2004， 2005-2013）， and the effects of groundwater exploitation and other human activities gradually took the place of climate change. Thus， in Table 2， the five time stages were determined primarily according to the annual average surface runoff volume variations at the Yingluoxia gauge station during 1980-2012. If the "Mean runoff volume" is greater （less） than the "Annual average runoff volume"of 16.93×108m3， it is a wet （dry） year； if it approaches 16.93×108m3， it is a normal year.

4.1Annual and cumulative variation of groundwater storage

From Table 3 and Figure 4f， groundwater storage variation was given priority to decline in respective partitions and for the whole Zhangye Basin， and the process varied from fast decrease to slow increase over the five time stages. During 1985-1989，groundwater storage varied at the rate of -0.28×108to 0.09×108m3/a in respective partitions because of wet years， and experienced a steady decrease in this stage. Annual average groundwater storage decreased at the rate of 0.22×108m3in the whole basin，with a cumulative reduction of 1.10×108m3； From 1990 to 1997， with the continuous dry years，groundwater storage decreased significantly at the rate of -2.15×108to -0.03×108m3/a， with an annual average reduction of 3.74×108m3/a and a cumulative reduction of 29.88×108m3； In 1998-2001， with enhanced human activities such as cultivated land expansion and irrigation water abstraction， even in normal years， groundwater storage reduction speed in respective partitions still accelerated， which decreased at a rate of -2.68×108to 0.01×108m3/a， annual average reduction of 4.02×108m3/a， and a cumulative reduction of 16.08×108m3； 2002-2004 was the early period of water diversion between midstream and downstream， where groundwater storage variations in respective partitions tended to be complicated and tended to decline and rise at both poles，where the average range was -2.75×108to 0.47×108m3/a， annually decreased by 2.20×108m3， and cumulative reduction was 6.61×108m3； From 2005-2013， under the influence of human activities and continuous wet years， storage in respective partitions basically remained balanced or slightly increase except for the river and spring irrigation district， where average variation was -0.01×108to 0.67×108m3/a， average increase rate in the Zhangye Basin was 0.68×108m3/a， and the cumulative increase reached 6.16×108m3.

Table 2　Temporal division of runoff volume from Yingluoxia gauge station during 1985-2012

Table 3　Spatial and temporal variation of groundwater storage during 1985-2013

Over the whole Zhangye Basin， total groundwater storage decreased 47.52×108m3during 1985-2013，with an annual deficit of 1.64×108m3. Among which，water storage in the RWMID reduced to 37.48×108m3， which accounted for 78.87% of the total loss of the whole basin， and groundwater resources deficiency was seriously the heaviest； in the RID， annual groundwater storage decreased by 8.32×108m3， which accounted for 17.51% of the total reduction， and groundwater resources deficiency was heavier； in the RSMID groundwater resources deficiency was relatively less because of the smaller irrigation area.

4.2Interannual variations of GLD and groundwater storage

Due to climate change and human activities disturbance， the GLD and groundwater storage （Figures 3， 4a-4e） also show a great spatio-temporal diversities in respective partitions.

4.2.1Interannual variations of GLD and groundwater storage in the river and well water mixed irrigation district

As presented in Table 3， Figures 3 and 4a， the annual mean storage of the RWMID was basically in a negative change during 1985-2004. This implies that the water table was falling， and storage decreased. The annual mean GLD was -0.29 m/a，reached a maximum fall of -0.85 m/a in 1997， a cumulative drop of up to 5.73 m and， the corresponding storage decrease at a rate of 1.88×108m3/a， reached a maximum of 5.60×108m3/a， and a cumulative reduction of 37.55×108m3. Due to the large area of the RWMID， the storage decrease accounted for about 59.07% of the whole basin. In the period 2005-2013，the water table decline rate slowed down， and the GLD changed from negative to positive， up to a maximum of 0.26 m/a， storage rose at the rate of 0.01×108m3/a， and a cumulative increase of 0.07×108m3/a. By 2013， the annual average GLD reached 0.20 m/a， accumulated to 5.72 m， storage reduced to 37.48×108m3， and groundwater deficiency was very serious. As for the reason， the RWMID was mainly located in the south alluvial fan and northern fine soil plain， including irrigation districts of Daman， Yingke， Xigan in Ganzhou County，Pingchuan， Xinhua in Linze County and Youlian，Liuba in Gaotai County. Irrigation water use was mainly from river water and well water， with well water irrigation accounting for more than 20%-40%. From the relationship between the volume of groundwater exploitation and cumulative groundwater storage depletion in the Zhangye Basin during 1985-2013 （Figure 5）， it illustrated that groundwater storage depletion varied with the intensity of groundwater exploitation， R2reached about 0.97. It could be assumed that groundwater exploitation is the main cause of groundwater table and storage decline. The rise in water table again after water diversion may be associated with the limiting of cultivated land and exploitation. Table 4 illustrates that after 1998-2004， irrigation area quick expansion was controlled and groundwater exploitation declined in respective irrigation districts. Some irrigation districts near the Heihe River on the alluvial fan， such as Daman and Yingke， received more river water leakage in continuous wet years after 1998-2004，which may also contribute to increased groundwater level and storage （Hu et al.， 2012）.

Figure 3　Annual mean groundwater storage variations in respective partitions during 1985-2013

4.2.2Interannual variations of GLD and groundwater storage in the river water irrigation district

Table 3， Figures 3 and 4b demonstrate that during 1985-2001， annual mean GLD and storage of the RID was basically in a negative change. The water table fell and， the annual mean GLD was -0.50 m/a， which reached a maximum fall of -1.62 m/a in 1991. Groundwater storage decreased at the average rate of 1.88×108m3/a， reached a maximum of 2.87×108m3/a，and the cumulative reduction was 14.98×108m3； in the period of 2002-2013， the water table decline rate slowed down and appeared a little rise trend， the annual mean GLD was 0.30 m/a， and the maximum increaserate reached 1.43 m/a. The corresponding storage rose at the rate of 0.55×108m3/a， with a cumulative increase of 6.66×108m3/a. By 2013， annual average GLD reached 0.16 m/a， accumulated decrease to 4.70 m，storage reduced to 8.32×108m3， and groundwater deficiency was serious. Taking year 2001 as the cut-off point， the annual mean and cumulative GLD before 2001 was given priority to rapid decline in this area，which rose again at a higher rate after 2001. Because the river water irrigation district is located in the upper and middle part of the alluvial fan on the south，groundwater buried depth is deeper. The river water irrigation district is mainly composed by irrigation districts of Shangsan and Anyang in Ganzhou County，Xinba and Hongyazi in Gaotai County， and part of irrigation districts in Mingle County， river water is the main irrigation water to use. This region is the groundwater runoff belt， groundwater flow rate is fast，and groundwater recharge is mainly river water and channel water leakage. Thus， GLD variation to some extent is related to wet or dry years of the climate. In dry years， the river receives less water from the upper stream and groundwater recharge is less， in contrast，recharge is more in wet years. From 2005 to 2013， the climate exhibited continuous wet years， and the RID received more recharge from river infiltration， and the irrigation area increased very little （Table 4）. Furthermore， before water diversion， intensified groundwater exploitation on the lower part of the alluvial fan accelerated the groundwater flow rate on the upper and middle of the alluvial fan， which induced the accelerated decrease of groundwater level and storage to some extent； after water diversion， the weakened groundwater exploitation slowed down the groundwater flow rate relatively， and the groundwater level and storage decreasing rate slowed down （Table 4）.

Figure 4　Accumulative groundwater storage variations in the Zhangye Basin and respective partitions during 1985-2013

Figure 5　Relationship between the volume of groundwater exploitation and cumulative groundwater storage depletion in the Zhangye Basin during 1985-2013

Table 4　Agricultural irrigation water used in typical irrigation districts during different time stages

4.2.3Interannual variations of GLD and groundwater storage in the spring water and river and spring water mixed irrigation district

Table 3， Figures 3， 4d， and 5e show that the variation trends of GLD and storage of SID and RSMID are almost the same. During 1985-2004， GLD and storage decreased at an accelerated speed and then slowed down at a decreasing trend. The annual mean GLD of SID and RSMID was -0.14 m/a and -0.10 m/a， and the cumulative decrease was -2.73 m and -1.94 m， respectively. Corresponding groundwater storage decreased at an average rate of 0.02×108m3/a and 0.05×108m3/a， and cumulative reduction was 0.4×108m3and 1.0×108m3， respectively； in the period of 2005-2013， the annual mean GLD in SID and RSMID was -0.001 m/a and -0.02 m/a， and the cumulative decrease was -0.01 m， -0.15 m， respectively. The corresponding storage decreased at the rate of 0.00×108m3/a and 0.01×108m3/a， cumulativereduction was 0.04×108m3and 0.44×108m3， respectively. From 1985 to 2013， the annual average GLD in SID and RSMID reached -0.09 m/a and -0.07 m/a，accumulated to -2.74 m and -2.09 m， and the cumulative storage reduced by 0.40×108m3and 1.08×108m3， respectively. Groundwater depletion was not so serious. Contrast to the RSMID， the annual mean GLD and storage variation in SID was larger. This is due to the fact that the SID and RSMID are mainly located in the valley plain along the main stream of the Heihe River. The SID was mainly composed by irrigation district of Xiaotun in Linze County， and the RSMID was composed by Banqiao， Yanuan and Liaoquan irrigation districts in Linze County. Before the 1980s， spring water was the main irrigation water use in both of the partitions. With groundwater exploitation， they developed to spring and well water mixed or river and spring and well water mixed irrigation districts. The proportion of groundwater was about 30.3% of the total water use in the SID， and a relatively small proportion in the RSMID. Large amount of groundwater exploitation had diminished spring overflow and， induced groundwater exploitation continues to increase， and spring overflow decease，which formed a vicious cycle. Regional groundwater table variation and storage in this area was closely related to groundwater exploitation （Table 4）.

4.2.4Interannual variations of GLD and groundwater storage in the Well water irrigation district

As presented in Table 3， Figures 3 and 4c， the variation trend of GLD and storage in well water irrigation district was similar to river water. During 1985-2001， annual mean GLD and storage of the well water irrigation district was basically in a negative change， and showed a continuous declining trend. The annual mean GLD reached -0.13 m/a， accumulated to -2.14 m， and the mean storage volume reduced to 0.10×108m3/a， cumulative reduction was 1.78×108m3； while in 2002-2013， groundwater table quickly rose and the GLD gradually turned into a positive change. The cumulative GLD increased to 1.85 m，and cumulative storage increased to 1.54×108m3. The WID located in the middle and upper of the alluvial fan， included the Luotuocheng district in Gaotai County and part of districts in Minle County， was well water accounted for more than 90%. Due to the effect of topography and groundwater exploitation， the groundwater table and storage fluctuations were frequent and complex. The main reason for the water table rise again after 2001 may be the implementation of "Comprehensive treatment planning in the Heihe River Basin"， and the limiting of cultivated land and over exploitation after the water-saving irrigation（Table 4）.

5　Conclusions

Groundwater storage in each partition and the whole Zhangye Basin varied with the groundwater level change. Before 2001 groundwater storage and level experienced a continuously decline， then after 2001 experienced a tendency of a slow decline or slight increase. In general， the accumulative storage in each partition and the whole Zhangye Basin is still in negative variation， namely the groundwater resource is still in a serious depletion state， which threatens the aquifer severely. The accumulative groundwater storage decreased nearly 47.52×108m3， the annual average deficit reached 1.64×108m3， among which the accumulative GLD of RWMID reached 5.72 m， and groundwater storage decreased by 37.48×108m3which accounts for about 78.87% of the total groundwater depletion of the Zhangye Basin. Accumulative GLD changed from high to low order in respective partitions was： RID＞RWMID＞SID＞RSMID≥WID， and the corresponded accumulative storage variation was：RWMID＞RID＞W(wǎng)ID＞RSMID＞SID. There were no continuous estimation results on groundwater storage previously， but the result calculated by the water balance method in 1995 was very close to the Pan and Tian's research （Pan and Tian， 2001）.

It was noted that， groundwater table and storage exhibited tremendous dissimilarity on temporal and spatial scale for the respective partitions， also showed certain regularity. That is， they experienced the continuous process from decline to slow down or recovery before and after 2001. Groundwater table and storage in the WID first appeared shifting from decrease to rise again in 2001， followed by RID and SID in 2003， at last RWMID， RSMID in 2005. The main reasons were limited intensity of cultivated land expansion and groundwater exploitation， and implementation of agricultural water-saving activities. It was also due to the continuous wet years. In the partitions with stronger human activity and climate change， the earlier the groundwater level and storage increased. The effects of human activity on the groundwater resources of Zhangye Basin gradually played a dominant role. Furthermore， GLD and storage in the RID experienced the largest fluctuations over these 30 years， and the reasons may be related to hydrogeological conditions， and the wet or dry conditions on different level years may also affect river infiltration. The reasons for groundwater level rise after 2001 are hotly debated （Ding et al.， 2009； Ba et al.， 2010； Hu et al.， 2012）.

Groundwater system variation in the Zhangye Basin occurred just before and after implementation of the water diversion policy. It was reported that after the water diversion， the amount of water diverted from the Heihe River to the Zhangye Basin decreasedto 2.0×108m3/a （Nian et al.， 2014）. This induced the "policy-based" water scarcity in Zhangye Basin， plus the quick expansion of cultivated land where large amount of groundwater exploitation occurred. Figure 5 shows that groundwater exploitation is one of the main disturbance factors of groundwater system variation in the Zhangye Basin. At present， the water diversion policy can cope with the water resources allocation problem between the mid-reaches and the lower reaches， but in the long run， the problem of sustainable water development and use in the river basin will still exist. Thus， it is better to continue with the limitation of cultivated land and groundwater exploitation， or revise the water diversion curve （the proportion of water diverted to the middle reaches and the lower reaches） according to the wet or dry years in combination with agricultural irrigation water needs.

Acknowledgments:

This work was supported by the Science and Technology Service Network Plan of the Chinese Academy of Sciences （KFJ-EW-STS-005-02）， the International Science & Technology Cooperation Program of China （2013DFG70990）， and National Natural Science Foundation of China （91225301）. We also thank the editors and anonymous reviewers who provided valuable comments.

Ba JW， Ma XQ， Liu ZH， et al.， 2010. Exploration & analysis on causes of groundwater level rising in the basin of Zhangye. Ground Water， 30（1）： 34-37.

Cai XM， Landon M， Ge YC， 2014. Decision support for integrated river basin management—Scientific research challenges. Science China （Earth Sciences）， 58（1）： 16-24. DOI： 10.1007/ S11430-014-5005-2.

Cao YP， Nan ZT， 2011. Monitoring water storage variations in the Heihe River Basin by the GRACE gravity satellite. Remote Sensing Technology and Application， 26（6）： 18-26.

Chen HS， 1984. Water resources evolution and its influence on eco-environment in Minqin Basin， Gansu. Natrural Resource，20（3）： 62-71.

Chen LH， Qu YG， 1992. Water and Land Resources and Their Rational Utilization and Development in the Hexi Region. Beijing： Science Press.

Cheng GD， Xiao HL， Xu ZM， et al.， 2006. Water issue and its countermeasure in the inland river basins of Northwest China：a case study in Heihe River Basin. Journal of Glaciology and Geocryology， 28（3）： 406-413.

Cui YL， Shao JL， 2005. The role of groundwater in arid/semiarid ecosystems， Northwest China. Ground Water， 43（4）： 471-477. DOI： 10.1111/j.1745-6584.2005.0063.x.

Ding HW， Yao JL， He JH， 2009. Environmental isotope characteristics and groundwater recharge in groundwater level rise area in Zhangye City. Arid Land Geography， 32（1）： 1-8.

Ding HW， Zhang J， 2002. The problem of environmental caused by groundwater level continuous decline in the inland basins of arid area， Northwest China—an example in middle reaches of Heihe River Basin. Hydrogeology & Engineering Geology， 3：71-75.

Fang CL， Bao C， Huang JC， 2007. Management implications to water resources constraint force on socio-economic system in rapid urbanization： a case study of the Hexi Corridor， NW China. Water Resource Management， 21（9）： 1613-1633. DOI：10.1007/s11269-006-9117-0.

Feng Q， Cheng GD， Masao MK， 2000. Trends of water resource development and utilization in arid Northwest China. Environmental Geology， 39（8）： 831-838. DOI： 10.1007/ s002549900062.

Feng Q， Liu W， Xi HY， 2013. Comprehensive evaluation and indicator system of land desertification in the Heihe River Basin. Natural Hazards， 65： 1573-1588. DOI： 10.1007/ s11069-012-0429-5.

Gao QZ， Li FX， 1990. Reasonable Development and Utilization of Water Resources in Heihe River Basin. Lanzhou： Gansu Science and Technology Press.

Hu LT， Wang ZJ， Tian W， et al.， 2009. Coupled surface water-groundwater model and its application in the arid Shiyang River Basin， China. Hydrological Process， 23（14）： 2033-2044. DOI： 10.1002/hyp.7333.

Hu XL， Xiao HL， Lan YC， et al.， 2012. Experimental study of calculation method of river seepage in middle and upper reaches of the Heihe River. Journal of Glaciology and Geocryology， 34（2）： 460-468.

Ji XB， Kang ES， Chen RS， et al.， 2006. The impact of the development of water resources on environment in arid inland river basins of Hexi region， northwestern China. Environmental Geology， 50： 793-801. DOI： 10.1007/s00254-006-0251-z.

Johnston K， Jay MH， Hoef V， et al.， 2001. Using ArcGIS Geostatistical Analyst. ESRI 380 New York Street.

Lan YC， Kang ES， Zhang JS， 2000. Runoff characteristics and change trend for inland rivers in Hexi. Journal of Glaciology and Geocryology， 22（2）： 147-152.

Lan YC， Kang ES， Zhang JS， 2002. Study on the water resources and its rational development and utilization in Heihe River Basin. Journal of Lanzhou University （Natural Sciences）， 38（5）：108-114.

Li BX， 1983. The study on construction of groundwater resources in Shiyang River Basin and its rational utilization. Journal of Desert Research， 3（4）： 1-10.

Liu SY， Lu YR， Cheng XX， et al.， 2002. Groundwater system and water resources environmental effect induced water resources development in the middle and lower reaches of Heihe River. Geography and Territorial Research， 18（4）： 90-96.

Lu L， Li X， Cheng GD， 2003. Landscape evolution in the middle Heihe River Basin of Northwest China during the last decade. Journal of Arid Environment， 53： 395-408. DOI： 10.1006/ jare.2002.1032.

Mi LN， Xiao HL， Zhu WJ， et al.， 2015. Dynamic evolution of groundwater level in the middle reaches of Heihe River in 1985-2013. Journal of Glaciology and Geocryology， 37（2）：461-469.

Nian YY， Li X， Zhou J， 2014. Impact of land use change on water resource allocation in the middle reaches of the Heihe River Basin in northwestern China. Journal of Arid Land， 6（3）：273-286. DOI： 10.1007/s40333-013-0209-4.

Pan QM， Tian SL， 2001. Water Resources of Heihe River Basin. Zhengzhou： Yellow River Water Conservancy Press.

Qi SZ， Luo F， 2005. Water environmental degradation of the Heihe River Basin in arid northwestern China. Environmental Monitoring and Assessment，108：205-215.DOI：10.1007/s10661-005-3912-6.

Qi SZ， Luo F， 2007. Environmental degradation problems in the Heihe River Basin， northwest China. Water Environment Journal， 21： 142-148. DOI： 10.1111/j.1747-6593.2006. 00059.x.

Tang QC， Zhang JW， 2001. Water resources and eco-environment protection in the arid regions in northwest of China. Progress in Geography， 20（3）： 227-233.

Wang GX， Cheng GD， Xu ZM， 1999. The utilization of water resources and its influence on eco-environment in the northwest arid area of China. Journal of Natural Resources， 14（2）：109-116.

Wang GX， Liu JQ， Jumpei K， et al.， 2007. Effects of land-use changes on hydrological processes in the middle basin of the Heihe River， northwest China. Hydrological Process， 21：1370-1382. DOI： 10.1002/hyp.6308.

Water Resources Strategy Research Group of Chinese Academy of Sciences， 2009. The Science and Technology Development Roadmap of Water Resources Field to 2050 in China. Beijing：Science Press.

Wei Z， Jin HJ， Lan YC， 2009. Changes analysis of groundwater resources in the middle Heihe River using Kriging methods after water redistribution. Arid Land Geography， 32（2）： 196-203.

Wen XH， Wu YQ， Lee LJE， et al.， 2007. Groundwater flow modeling in the Zhangye Basin， northwestern China. Environmental Geology， 53： 77-84. DOI： 10.1007/s00254-006-0620-7.

Wen XH， Wu YQ， Wu J， 2008. Hydrochemical characteristics of groundwater in the Zhangye Basin， northwestern China. Environmental Geology， 55（1）： 713-724. DOI： 10.1007/ s00254-007-1122-y.

Xiang GS， 2011. Risk assessment and regulation of groundwater development in Zhangye Basin of the middle reaches of Heihe River. D.Sc. Thesis， Lanzhou： Lanzhou University.

Xiao HL， Cheng GD， 2006. Water issue and management at basin level in Heihe River， northwestern China. Journal of Desert Research， 26（1）： 1-5.

Xiao SC， Xiao HL， Peng XM， et al.， 2015. Hydroclimate-driven changes in the landscape structure of the terminal lakes and wetlands of the China's Heihe River Basin. Environmental Monitoring and Assessment， 187： 4091. DOI： 10.1007/ s10661-014-4091-0.

Yang LY， Wang GX， 2005. Variations of groundwater in Zhangye Basin of the middle reaches of the Heihe River in recent two decades. Journal of Glaciology and Geocryology， 27（2）：290-296.

Zhang JS， Kang ES， Zhao AF， 2003. The studies of water resources utilization or water resources and ecology-environments safety in middle reach of Heihe River Basin. Advances in Earth Sciences， 18（2）： 207-213.

Zhang MQ， Zeng ZZ， 1994. Water Resources Assessment. Lanzhou： Lanzhou University Press.

Zhang YC， Yu JJ， Qiao MY， 2011. Effects of eco-water transfer on changes of vegetation in the lower Heihe River Basin. Journal of Hydraulic Engineering， 42（7）： 757-765.

Zhong FL， Xu ZM， Cheng HW， 2011. The history of water resources utilization and management in the middle reaches of Heihe River. Journal of Glaciology and Geocryology， 33（3）：236-245.

Zhou W， Sun ZG， Li JL， et al.， 2013. Desertification dynamic and the relative roles of climate change and human activities in desertification in the Heihe River Basin based on NPP. Journal of Arid Land，5（4）：465-479.DOI：10.1007/ s40333-013-0181-z.

Zhu GF， Su YH， Feng Q， 2008. The hydrochemical characteristics and evolution of groundwater and surface water in the Heihe River Basin， Northwest China. Hydrogeology Journal， 16（1）：167-182. DOI： 10.1007/s10040-007-0216-7.

Mi LN， Xiao HL， Yin ZL， et al.， 2016. Dynamic evaluation of groundwater resources in Zhangye Basin. Sciences in Cold and Arid Regions， 8（3）： 0250-0262.

10.3724/SP.J.1226.2016.00250.

*Correspondence to： LiNa Mi， Cold and Arid Regions Environmental and Engineering Research Institute， Chinese Academy of Sciences. No. 320， West Donggang Road， Lanzhou， Gansu 730000， China. E-mail： xiaomizi-123@163.com

October 20， 2015Accepted： March 28， 2016