Abstract
Intense anthropogenic activities in arid areas have great impacts on groundwater process by causing river dried-up and phreatic decline. Groundwater recharge and discharge have become hot spot in the dried-up river oases of arid regions, but are not well known, challenging water and ecological security. This study applied a stable isotope and end-member mixing analysis method to quantify shallow groundwater sources and interpret groundwater processes using data from 186 water samples in the Wei-Ku Oasis of central Asia. Results showed that shallow groundwater (well depth < 20 m) was mainly supplied by surface water and lateral groundwater flow from upstream, accounting for 88 and 12%, respectively, implying surface water was the dominant source. Stable isotopes and TDS showed obviously spatiotemporal dynamic. Shallow groundwater TDS increased from northwest to southeast, while the spatial variation trend of groundwater δ18O was not obvious. Surface water and groundwater in non-flood season had higher values of stable isotopes and TDS than those in flood season. Anthropogenic activities greatly affect groundwater dynamics, where land-cover change and groundwater overexploitation are the main driving factors. The findings would be useful for further understanding groundwater sources and cycling, and help restore groundwater level and desert ecosystem in the arid region.
HIGHLIGHTS
The sources of shallow groundwater in the dried-up river oasis of central Asia were quantified.
Surface water was the dominant source of shallow groundwater.
Anthropogenic activities greatly affect groundwater dynamic and cycle.
Graphical Abstract
INTRODUCTION
Understanding the role of anthropogenic activities in the groundwater hydrological processes in arid region is becoming increasingly important, because intense anthropogenic activities and climate change could noticeably alter the water balance and quality of aquifer system (Han et al. 2017; Wang et al. 2021). In the past decades, groundwater storage in arid areas has decreased worldwide primarily due to a growing influence of anthropogenic activities (de Graaf et al. 2019; Shakya et al. 2019). Groundwater salinization and phreatic decline could affect hydrological processes, biogeochemical cycles, and vegetation growth (Tweed et al. 2011; Fuchs et al. 2019; Jia et al. 2020). Land-cover type change could affect the regional surface water and groundwater redistribution and hydrologic processes, thus influencing groundwater dynamics in arid areas (Shukla et al. 2018; Li et al. 2019a). Groundwater exploitation plays a crucial role on the dynamics and water balance of aquifer system in arid inland areas (Asoka et al. 2017; Hu et al. 2019; Mancuso et al. 2020). Groundwater resource sustainability is vital to the socio-economic development and ecological maintenance in arid areas due to arid climate and limited surface water resources (Gleeson et al. 2010; Taylor et al. 2013; Lezzaik et al. 2018), especially in arid inland oasis (Guo et al. 2019b; Wang et al. 2020). The Tarim Basin is located in the arid inland region of central Asia and the largest inland river basin in China, and is the core area of constructing the ‘Silk Road Economic Belt’ (Chen et al. 2019). In the last 30 years, groundwater overextraction has been causing river dried-up, phreatic decline, and ecological degradation in the oasis-desert areas of the Tarim Basin due to increasing population and farmland expansion (Zhang et al. 2014; Guo et al. 2019b), challenging the local water and ecological security (Chen et al. 2019). Consequently, it is essential to understand groundwater sources and cycle to sustainably manage water resources in the arid inland oasis of the Tarim Basin (Steward & Allen 2016; Wang et al. 2021).
In recent years, groundwater dynamic and its response to climate change and human activities have become hot spots in arid inland areas (de Graaf et al. 2019). Previous studies on the groundwater dynamics in arid inland regions, using hydrological monitoring, stable isotopic and hydrochemical data, hydrological models, and remote sensing inversion, have evaluated groundwater level and storage dynamics (Wang et al. 2019), groundwater salinity and hydrochemical evolution (Hasan et al. 2011; Liu et al. 2018b), groundwater recharge and circulation (Wang et al. 2013; Cartwright et al. 2019), the interaction between surface water and groundwater (Xue et al. 2019), and its relationship with driving factors (Liu et al. 2018a; Wang et al. 2021). Wang et al. (2020) reported that shallow groundwater in the oases of northern Tarim Basin may be recharged by river and irrigated water infiltration, lateral groundwater flow, and little precipitation. Wang et al. (2021) found that human activity was the major driving factor of groundwater level decreasing in arid oasis. However, groundwater sources and cycling mechanisms were not well understood in the oases of Tarim Basin, especially quantitative evaluation due to scarce data and intense human activities (Wang et al. 2020, 2021), which affects the implementation of sustainable water resource management in this arid inland region (Chen et al. 2019).
Stable isotopic and geochemical indicators are effective tools for investigating the groundwater cycle process and hydrochemical evolution under the influence of climate change and human activities in arid regions (Duan et al. 2013, 2016; Gibson et al. 2017; Guo et al. 2019b). Stable natural tracers behave conservatively and their concentrations are not affected by hydrogeochemical reactions within aquifer systems (Richards et al. 2018). Hence, environmental tracer methods and the end-member mixing analysis method (EMMA) have been widely used to quantitatively determine groundwater recharge sources and rates (Scanlon et al. 2006; Yin et al. 2011; Guo et al. 2019a), flow paths (Huang & Pang 2010), residence times (Peng et al. 2012), and biogeochemical cycle in arid regions (Pang et al. 2010; Wang et al. 2020; Gong et al. 2021). Li et al. (2019b) investigated the water cycle in inland regions from stable isotope tracing. Therefore, for arid inland areas with limited hydrological observation, stable isotopic and geochemical indicators are particularly useful for analyzing groundwater sources and cycle (Li et al. 2019b).
In this study, we attempted to combine stable isotopic indicators and EMMA to determine shallow groundwater sources and interpret groundwater processes using data from 186 water samples in the Wei-Ku Oasis, a typical dried-up river oasis of Tarim Basin in central Asia. The main objectives of our study were (1) to examine the spatiotemporal characteristics of stable isotopes and salinity in surface water and groundwater, (2) to quantify the sources of shallow groundwater, and (3) to explore groundwater recharge and discharge under the influence of anthropogenic activities (i.e. land-cover change, groundwater extraction, agricultural irrigation, and drainage activities). The results were expected to help further understand groundwater sources and cycling under intense anthropogenic activities, and improve sustainable water management in the dried-up river oasis of arid inland areas.
STUDY AREA
Our study region is the Wei-Ku Oasis (82.09–83.47 °E, 40.90–41.85 °N), and located in the middle and lower reaches of the Weigan-Kuqa River basin, with an area of 7,104 km2 (Figure 1). The Weigan-Kuqa River basin is composed of the Weigan River and the Kuqa River, and the annual streamflow is 26.6 × 108 m3 and 4.6 × 108 m3 (Wang et al. 2021), respectively, with the flood season from June to September (Figure 1). The oasis is located in the southern Tianshan Mountains and the northern Tarim Basin, which is sensitive to ecological environment changes. It was a typical and complete piedmont alluvial fan plain, whose elevation ranges from 945 to 1,147 m, and declines from north to south within the oasis. The climate of our study region is a typical temperate arid continental climate (Wang et al. 2021). The annual average air temperature is 11.4 °C, and the annual frost-free period is 209–226 days. The annual average precipitation is 74.6 mm, with about 76% of the total precipitation mainly occurring between May and September (Figure 1) (obtained from the Chinese National Meteorological Information Centre). The annual runoff into the Wei-Ku Oasis is about 31.2 × 108 m3, and more than 90% of the total runoff is directly diverted into cultivated land for agricultural irrigation (Wang et al. 2021). This caused the main stream to dry up within the oasis in recent decades, and only a small amount of flood was discharged in flood season (Chen et al. 2019). Depth to phreatic water level ranged from 2 to 6 m across the whole oasis during 2000–2018, and phreatic water level showed a decreasing trend from northwestern (higher altitude) to southeastern (lower altitude), implying that shallow groundwater in the oasis flows from north to south, and finally into the mainstream of Tarim River (Wang et al. 2021).
Locations of the Tarim Basin (a), Weigan-Kuqa River Basin (b), and water sampling sites (c), and the intra-annual variations of monthly average air temperature, precipitation (d), and streamflow (e) in the Wei-Ku Oasis. The map of the Tarim Basin was modified from Wang et al. (2021).
Locations of the Tarim Basin (a), Weigan-Kuqa River Basin (b), and water sampling sites (c), and the intra-annual variations of monthly average air temperature, precipitation (d), and streamflow (e) in the Wei-Ku Oasis. The map of the Tarim Basin was modified from Wang et al. (2021).
The dominated natural plants in the Wei-Ku Oasis are Sophora alopecuroides, Tamarix spp., Populus euphratica, Karelinia caspica, Phragmites communis, Halostachys caspica, Glycyrrhiza, and Alhagi pseudalhagi (Zhao 2006). The main types of land use include cropland, grassland, and bare areas. This oasis is covered by the Quaternary sediments of different hydrogeologic units, including pebbly sandstone, fine siltstone, fine sandstone, silty fine sandstone, sandy gravel, and sandstone. The northern margin and northeastern part of the oasis are proluvial sediments, covered by pebbly sandstone and sandstone (Wang et al. 2021). The southeastern part of the oasis is alluvial sediment, covered by fine sandstone (40–60 m of thickness). The central part is alluvial–proluvial sediment, the sediments are from pebbly sandstone to fine siltstone. Furthermore, the main soil types include oasis soil, inland saline soil, brown desert soil, meadow soil, and aeolian sandy soil.
MATERIALS AND METHODS
Data collection
During the period from August 2018 to August 2019, water samples were obtained along the Weigan River and the Kuqa River, as well as within the Wei-Ku Oasis in the flood season (August) and non-flood season (April and November), including surface water samples and groundwater samples (Figure 1(c)). Surface water samples in upstream (SWU) were collected from the tributaries of Weigan River and Kuqa River (river water). Surface water samples in oasis (SWO) were obtained from the Wei-Ku Oasis, and included river water, reservoir water, and channel water, where the channel water for agricultural irrigation would be from stream water and groundwater pumping from confined aquifer. Groundwater samples were taken from regularly used boreholes (long-term observation wells) and pumped wells (industrial wells, irrigation wells, and domestic wells) within the oasis. Considering the hydrogeological characteristics in the Wei-Ku Oasis, groundwater samples were grouped into three categories: shallow groundwater (well depth < 20 m), middle groundwater (20 m < well depth < 100 m), and deep groundwater (well depth > 100 m) (Wang et al. 2013). Moreover, 15 shallow groundwater samples were collected from the upstream to treat as lateral groundwater flow samples for the Wei-Ku Oasis. A total of 186 water samples were obtained, including 97 surface water samples (26 in upstream and 71 in oasis) and 89 groundwater samples (67 shallow groundwater samples, 21 middle groundwater samples, and 1 deep groundwater samples). All water samples were filtered through a 0.22 μm filter, and immediately sealed in 100-mL high-density polyethylene bottles, and then stored in coolers until analysis.
Stable water isotopes (δ2H and δ18O) and chemical measurements of the water samples were analyzed at the State Key Laboratory of Desert and Oasis Ecology, Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences. δ2H and δ18O of water were determined by a liquid water isotope analyzer (LGR DLT-100), and expressed relative to the Vienna Standard Mean Ocean Water (VSMOW) in per million (δ, ‰). The analytical precision for δ2H and δ18O were ±0.8‰ and ±0.1‰, respectively. The total dissolved solids (TDS) of water samples were measured in the fields using a multi-parameter meter (YSI ProPlus), with a TDS measurement precision of 0.01 mg/L.
In this study, streamflow data at the outlets of the mountainous tributaries (Weigan River and Kuqa River) in the Weigan-Kuqa River basin from 1960 to 2018 were collected from the Weigan River Basin Authority (Figure 1(b)), which was observed daily by the hydrological observational stations (Wang et al. 2021). Groundwater level data of phreatic aquifer in the oasis were obtained from the Water-salt Monitoring Station of the Weigan River Basin Authority, China. Meteorological data (air temperature and precipitation) were obtained from the Chinese National Meteorological Information Centre (http://cdc.cma.gov.cn). Land-use and land-cover (LULC) data was derived from the C3S Global Land Cover products, with a 300 m resolution from 2016 to 2018 (http://cds.climate.copernicus.eu/). Moreover, data statistical analyses were conducted using the Origin (version 2021), ArcGIS (version 10.6), and MATLAB (version R2018a). The IDW (inverse distance weighting) method was applied to interpolate the data of stable isotope, TDS, and LULC into a 250 m grid data.
End-member mixing analysis method










RESULTS
Stable isotopic composition
The stable isotopic compositions of surface water and groundwater are summarized in Table 1 and Figure 2. The isotopic values of various water components exhibited a large range. The stable isotopes of surface water in upstream ranged from −81.2‰ to −45.5‰ for δ2H (mean δ2H = −66.5‰) and from −12.0‰ to −7.6‰ for δ18O (mean δ18O = −10.1‰) were lighter than precipitation isotopes in the Hetian station (δ18O = −5.5‰) but heavier than precipitation isotopes in the Wulumuqi station (δ18O = −10.7‰) in Xinjiang (data from the Global Network of Isotopes in Precipitation). The δ18O and δ2H of surface water in the oasis (from −11.4‰ to −7.0‰ and from −75.8‰ to −42.3‰) were close to the isotopes of surface water in upstream, indicating that surface water in the oasis was mainly from river water. Additionally, the shallow groundwater data (mean δ2H = −69.5‰, mean δ18O = −10.0‰) was similar with the values of surface water in the oasis, suggesting that shallow groundwater was mainly recharged from local surface water (Wang et al. 2020). Furthermore, in the Wei-Ku Oasis, the isotopes of middle water (δ18O = −10.2‰) were heavier than those of deep groundwater (δ18O = −10.8‰), while lighter than those of shallow groundwater, but the difference of isotopes among various aquifers was small. This may indicate that the interactions between different aquifers were frequent (Wang et al. 2013; Zeng et al. 2020).
Stable isotopes and salinity of water samples obtained in the Weigan-Kuqa River Basin
. | Water components . | |||||
---|---|---|---|---|---|---|
SWU . | SWO . | SG . | MG . | DG . | ||
δ2H (‰) | Max | −45.5 | −42.3 | −52.0 | −64.1 | / |
Min | −81.2 | −75.8 | −83.9 | −79.9 | / | |
Mean | −66.5 | −64.8 | −69.5 | −70.5 | −74.4 | |
δ18O (‰) | Max | −7.6 | −7.0 | −7.5 | −9.1 | / |
Min | −12.0 | −11.4 | −12.2 | −11.8 | / | |
Mean | −10.1 | −9.7 | −10.0 | −10.2 | −10.8 | |
TDS (mg/L) | Max | 803.6 | 1,753.1 | 28,783.0 | 3,419.4 | / |
Min | 192.4 | 300.4 | 807.4 | 713.3 | / | |
Mean | 427.7 | 660.5 | 9,357.1 | 1,519.8 | 373.9 |
. | Water components . | |||||
---|---|---|---|---|---|---|
SWU . | SWO . | SG . | MG . | DG . | ||
δ2H (‰) | Max | −45.5 | −42.3 | −52.0 | −64.1 | / |
Min | −81.2 | −75.8 | −83.9 | −79.9 | / | |
Mean | −66.5 | −64.8 | −69.5 | −70.5 | −74.4 | |
δ18O (‰) | Max | −7.6 | −7.0 | −7.5 | −9.1 | / |
Min | −12.0 | −11.4 | −12.2 | −11.8 | / | |
Mean | −10.1 | −9.7 | −10.0 | −10.2 | −10.8 | |
TDS (mg/L) | Max | 803.6 | 1,753.1 | 28,783.0 | 3,419.4 | / |
Min | 192.4 | 300.4 | 807.4 | 713.3 | / | |
Mean | 427.7 | 660.5 | 9,357.1 | 1,519.8 | 373.9 |
Note: SWU, surface water in upstream; SWO, surface water in oasis; SG, shallow groundwater; MG, middle groundwater; DG, deep groundwater; TDS, total dissolved solids; Max, maximum value; Min, minimum value; Mean, mean value.
Relationship between δ18O and δ2H for surface water and groundwater in the Wei-Ku Oasis. GMWL, global meteoric water line; LMWL in NW China, local meteoric water line in northwest China (Guo et al. 2019a); SWL, surface water line in the oasis; GWL, groundwater line in the oasis. Precipitation isotopic data were from the GNIP (Global Network of Isotopes in Precipitation) for the Hetian and Wulumuqi stations.
Relationship between δ18O and δ2H for surface water and groundwater in the Wei-Ku Oasis. GMWL, global meteoric water line; LMWL in NW China, local meteoric water line in northwest China (Guo et al. 2019a); SWL, surface water line in the oasis; GWL, groundwater line in the oasis. Precipitation isotopic data were from the GNIP (Global Network of Isotopes in Precipitation) for the Hetian and Wulumuqi stations.
The diagrams of δ18O vs. δ2H for surface water and groundwater samples in the Wei-Ku Oasis illustrated that the data of surface water and groundwater were both located near the GMWL (global meteoric water line) and LMWL in northwest China (local meteoric water line; Guo et al. 2019a), and distributed in a relatively scattered way (Figure 2). The groundwater line (GWL) based on these groundwater data was , and its slope and intercept were both smaller than those of GMWL (
) (Craig 1961) and LMWL in northwest China (
) (Guo et al. 2019a), possibly due to evaporative enrichment in the arid inland oasis (Wang et al. 2013). Additionally, the surface water line (SWL) based on the surface water samples in the oasis was
, and the slope and intercept were smaller than those of GMWL, but greater than those of the LMWL in northwest China and the Tarim Basin (
) (Sun 2015). This indicated that surface water in the oasis did not experience strong evaporation, maybe due to the rapid transport of river water to irrigated fields through irrigation canals (Li et al. 2019b).
The spatial distribution of δ18O for surface water and shallow groundwater in the Wei-Ku Oasis were prepared (Figure 3). As shown in Figure 3, stable isotopic values of surface water and shallow groundwater varied among sampling sites, and showed obviously spatial dynamic. This may be attributed to the different effects of air temperature, evaporation, and agricultural activities (well-developed irrigation canal system and agricultural drainage ditch network) by redistributing water (Wang et al. 2020, 2022). In general, except for the northeast corner, the δ18O value of surface water in the oasis increased from northwest to southeast, while the spatial variation trend of shallow groundwater δ18O in the oasis was not obvious (Figure 3). The δ18O of surface water in the Wei-Ku Oasis ranged from −11.4‰ to −7.0‰, and the low value was mainly distributed in the northwestern part, while the high value was mainly distributed in the northeast and southern edge of the oasis. Additionally, shallow groundwater δ18O in the Wei-Ku Oasis ranged from −12.2‰ to −7.5‰, and showed a relatively large spatial heterogeneity. The high value of shallow groundwater δ18O was mainly distributed in the northern edge and southern edge of the oasis, possibly due to different sources of shallow groundwater among sampling sites. The low value of shallow groundwater δ18O was distributed in several sporadic areas on the central oasis, possibly due to recharge with river water from higher elevation (Guo et al. 2019a).
Spatial distribution of δ18O for surface water and shallow groundwater in the Wei-Ku Oasis: (a) surface water and (b) shallow groundwater.
Spatial distribution of δ18O for surface water and shallow groundwater in the Wei-Ku Oasis: (a) surface water and (b) shallow groundwater.
Figure 4 shows the seasonal variations of stable isotopes for surface water and groundwater across the Wei-Ku Oasis. In the oasis, the δ18O and δ2H of surface water and groundwater showed obviously temporal dynamic, and varied among different aquifers, mainly due to the seasonal variations of water source, air temperature, and precipitation in the study region. Generally, the mean values of δ18O and δ2H in surface water and groundwater for non-flood season were both heavier than those for flood season in the oasis (Figure 4), and were contrary to the seasonal variations of rainfall and runoff (Wang et al. 2021), possibly due to the massive recharge from river water with lighter isotopes in flood season (Guo et al. 2019a). The seasonal variation of mean δ18O value was the largest in surface water, followed by shallow groundwater, while the smallest in middle groundwater mainly due to the seasonal differences of recharge sources and processes among different water components (Wang et al. 2013). However, the difference of δ18O seasonal variation was not obvious among various water components (Figure 4), possibly indicating the close interaction between different aquifers (Zeng et al. 2020). Furthermore, stable isotopes of shallow groundwater were lighter than those of surface water and heavier than those of middle groundwater both in flood season and non-flood season.
Box and whisker plots showing the variations of stable isotopes and TDS for surface water and groundwater in flood season and non-flood season across the Wei-Ku Oasis.
Box and whisker plots showing the variations of stable isotopes and TDS for surface water and groundwater in flood season and non-flood season across the Wei-Ku Oasis.
Dissolved ions
TDS content varied significantly between different water components (Table 1). In general, the TDS content of groundwater samples in the study area was the largest, followed by surface water in the oasis, while the smallest in river water samples, mainly due to strong evaporation in such arid climate (Liu et al. 2018b). River water in the upstream had low salinity, and its TDS showed a small range (TDS = 428 mg/L) (Table 1). Compared with the river water samples in upstream, surface water in the oasis had slightly higher salinity (TDS = 300–1,753 mg/L), suggesting that surface water in the oasis was subjected to strong evaporation.
Furthermore, the salinity of groundwater exhibited a significant difference between various aquifers, and showed a wide range for shallow and middle groundwater samples (Table 1). Generally, the salinity of deep groundwater in the oasis was the smallest, followed by the middle groundwater, while the largest in shallow groundwater, which may be attributed to the different water sources for each groundwater components (Wang et al. 2013). Shallow groundwater in the oasis had high salinity (TDS = 9,357 mg/L), which was mainly because surface water provided a significant salinity contribution to groundwater in the study region (Wang et al. 2021). The salinity of middle groundwater samples (TDS = 1,520 mg/L) was higher than the value of deep groundwater (TDS = 373.9 mg/L), but lower than the value of shallow groundwater, which implied that phreatic water could leak downward to confined water, and in turn lead to the interaction between salt and fresh water (Zeng et al. 2020).
TDS concentrations of surface water and shallow groundwater in the Wei-Ku Oasis showed spatial differences among the sampling sites (Figure 5). Generally, TDS values of shallow groundwater in the oasis showed an increasing trend from northwest to southeast. TDS concentration of surface water ranged from 300 to 1,753 mg/L in the oasis. The low value of surface water TDS was mainly distributed in the west and several sporadic areas on the central and north part, while the high value was mainly distributed near the main stream of Tarim River in the south, which may be attributed to evaporative concentration effect of surface water ions along the flow path (Liu et al. 2018b). In addition, the TDS value of shallow groundwater in the oasis ranged from 807 to 28,783 mg/L was low in the northwest part and high in the southeast part, which may be attributed to the little recharge of groundwater from river water in the southeast desert area. Furthermore, the spatial variation trend of middle groundwater TDS was not obvious in the Wei-Ku Oasis, while the shallow groundwater TDS value showed significant horizontal zoning, which may be due to the obvious differences of water sources, hydrochemical formation mechanism, and human interference intensity between various aquifers (Wang et al. 2013).
Spatial distribution of TDS for surface water and shallow groundwater in the Wei-Ku Oasis: (a) surface water and (b) shallow groundwater.
Spatial distribution of TDS for surface water and shallow groundwater in the Wei-Ku Oasis: (a) surface water and (b) shallow groundwater.
Figure 4 exhibits the seasonal variations of TDS in surface water and groundwater across the Wei-Ku Oasis. The TDS concentration of surface water and groundwater showed different fluctuation in flood season and non-flood season, and seasonal variation of TDS in the oasis was high during the sampling period. Generally, the mean TDS value of surface water and groundwater in non-flood season was higher than that in flood season in the oasis (Figure 4), possibly due to the massive recharge from river water with lower salinity in flood season (Wang et al. 2021). The difference of TDS value among seasons was the largest in shallow groundwater, followed by surface water, while the smallest in middle groundwater, suggesting that middle groundwater was relatively steady. The average TDS concentration of surface water in the oasis was 604 mg/L in flood season and 1,207 mg/L in non-flood season, which was contrary to the seasonal variations of rainfall and runoff. The shallow groundwater TDS value was 8,167 mg/L in flood season and 9,539 mg/L in non-flood season, while the middle groundwater TDS value was 1,467 mg/L and 1,553 mg/L in flood season and non-flood season, respectively.
The sources of groundwater
The end-members analysis method was applied to identify water sources and to quantify the contribution proportions of each end member to shallow groundwater in the Wei-Ku Oasis. The potential water sources of shallow groundwater include surface water (leakage or infiltration from canal, riverbank, reservoir, and irrigation), lateral groundwater flow from the upstream, and confined water (leakage recharge) (Table 1 and Figure 2), in which river water and groundwater pumping from confined aquifer are used for irrigation in this area (Chen et al. 2019). The concentrations of δ2H and δ18O for surface water, lateral groundwater flow (δ2H = −78.2‰, δ18O = −11.5‰), confined water and shallow groundwater revealed considerable differences in the oasis (Table 1). Therefore, both δ2H and δ18O were selected as tracers to confirm if the three end-members exist. Shallow groundwater in the oasis was located outside the triangle spanning the three end-members (Figure 6), indicating that the three members did not exist. Confined water would recharge shallow groundwater directly or indirectly. The directly recharge of confined water to phreatic water was little and could be neglected, because groundwater level dropped continuously and caused phreatic water leakage downward to recharge confined water due to groundwater overexploitation in the Wei-Ku Oasis (Wang et al. 2021). Additionally, surface water samples in the oasis contain the pumping confined water used for agricultural irrigation, so the indirect recharge of confined water to shallow groundwater has been reflected in the end-member of surface water.
End-member mixing diagram using the mean δ2H and δ18O values for the shallow groundwater in the Wei-Ku Oasis.
End-member mixing diagram using the mean δ2H and δ18O values for the shallow groundwater in the Wei-Ku Oasis.
Hence, the two end-members analysis method (Equations (1) and (2)) was used to estimate the contribution of surface water and lateral groundwater flow to shallow groundwater in this study. The mean value of δ18O in lateral groundwater flow from the upstream (−11.5‰), surface water (−9.7‰) and shallow groundwater (−10.0‰) was used for this calculation (Table 1). In general, surface water and lateral groundwater flow accounted for 88 and 12% of shallow groundwater in the Wei-Ku Oasis, respectively. Obviously, shallow groundwater in the oasis mainly originated from local surface water infiltration, and lateral groundwater flow from the upstream also played an important role in shallow groundwater in the arid inland oasis.
DISCUSSION
Groundwater recharge
The importance of precipitation as one water source of shallow groundwater was relatively little, and was neglected during estimating the sources. Snowmelt played an important role in the runoff of mountainous areas in spring (Li et al. 2019b). However, the contribution of snowmelt to shallow groundwater in the oasis region was weak, because snow cover was little in the oasis and snowmelt rapidly evaporated or sublimated into the atmosphere during the melt period (Wang et al. 2021). In the Wei-Ku Oasis, annual average precipitation was about 74.6 mm (Figure 1), while annual potential evaporation was about 2,401 mm (Chen et al. 2019), resulting in relatively dry soil with large soil water holding capacity. In addition, depth to phreatic water level in the oasis was generally above 5 m, which may lead to very weak recharge from rainfall to groundwater in the arid inland oasis (Wang et al. 2021). Thus, precipitation can be neglected during estimating groundwater sources.
As shown before, surface water in the Wei-Ku Oasis accounted for 88% of the sources of shallow groundwater, while lateral groundwater flow from the upstream accounted for 12%, suggesting that surface water was the dominant source of shallow groundwater. The stable isotopic values of shallow groundwater were close to those of surface water rather than those of lateral groundwater flow (Figures 2 and 6), maybe due to the different water migration mechanisms between the two end-members (Wang et al. 2013). Furthermore, runoff into the Wei-Ku Oasis is basically consumed in this area, most of which is used for agricultural irrigation, and then migrated downward to supply groundwater (Zhang et al. 2014; Wang et al. 2021). Wang et al. (2021) pointed out that the intra-annual distribution of runoff exhibited three peaks due to reservoir regulation (Figure 1), corresponded to crop irrigation time (from March to April, May to August, and October to November) in the region, indicating that the recharge of groundwater from river water mainly occurred during the irrigation season. In addition, a certain degree of increase of phreatic water level from January to April confirmed the recharge of phreatic water from lateral groundwater flow (Figure 7; Wang et al. 2013), because there is no agricultural irrigation and almost all of river water is stored in mountain reservoir during this period (Wang et al. 2021). Guo et al. (2019a) also found that lateral groundwater flow and irrigated return water were the main sources of shallow groundwater using stable isotope data in arid oasis. Furthermore, we examined the shallow groundwater sources and its dominant contributors using one-year data in this study. More detailed long-term and continuous investigation are required to explore the groundwater source dynamics (different locations and seasons) in the arid oasis in the context of climate change and intense anthropogenic interference.
Intra-annual variation of monthly average groundwater level depth in phreatic aquifer for G1 and G2 groundwater observation wells in the Wei-Ku Oasis in 2019.
Intra-annual variation of monthly average groundwater level depth in phreatic aquifer for G1 and G2 groundwater observation wells in the Wei-Ku Oasis in 2019.
According to the above results, we analyzed groundwater recharge process in the Wei-Ku Oasis. Runoff formed in mountainous region is the dominant source of surface water and groundwater in the oasis, which is dried-up in the oasis and could not reach the Tarim River (Chen et al. 2019). In the flood season, rainfall infiltrates into the dry soil layer, some is consumed by soil moisture evaporation, and the other part transfers downward to supply phreatic water by gravity and capillary force. However, the recharge of groundwater from rainfall is very little due to scarce precipitation and strong evaporation in the arid oasis (Ma et al. 2013). At the same time, the leakage water from natural streambed, canals, reservoirs, and agricultural irrigation could infiltrate directly or indirectly to recharge shallow groundwater by gravity (accounting for 88%), as the predominant source in the oasis (Wang et al. 2021). In addition, lateral ground from the upstream flows into the Wei-Ku Oasis to recharge shallow groundwater by the hydraulic gradient within aquifer system (accounting for 12%), which process is continuous and slow, and affected by the magnitude and direction of hydraulic gradient (Ma et al. 2013; Wang et al. 2013). Furthermore, deep confined water also is an important source of shallow groundwater within the arid oasis (Guo et al. 2019b). The irrigation water from pumping deep groundwater could infiltrate into phreatic aquifer (Chen et al. 2019). Confined water could leak upward to recharge shallow groundwater, but it is limited due to groundwater level decline caused by groundwater overexploitation (Zeng et al. 2020). This indicates that human activities play an important role on the groundwater recharge in the arid inland oasis (Castellano et al. 2019; Li et al. 2019a). Zeng et al. (2020) noted little difference of stable isotopes among phreatic water, shallow confined water, and deep confined water, and pointed out that groundwater mixing occurred among different aquifers in arid inland basin.
Groundwater discharge
According to the spatiotemporal characteristics of groundwater TDS and isotopes, the front range of the Wei-Ku Oasis is recharge area of groundwater, and the Tarim River is discharge area. Accordingly, four main routes of the oasis groundwater discharge are possible: phreatic water evapotranspiration, groundwater extraction, agricultural drainage, and lateral flow into the Tarim River by hydraulic gradient (Wang et al. 2021). Under natural conditions, evapotranspiration in the shallow phreatic aquifer is an important route of groundwater discharge, and mainly occurs in the southwest part of the oasis, because phreatic water evaporation is closely related to groundwater level depth, the lithology of the unsaturated zone and vegetation coverage in the region (Wang et al. 2013; Guo et al. 2019b). As the groundwater level decreases, the area of effective phreatic evaporation decreases in the arid oasis, thus leads to the decreasing of phreatic evaporation amount (Eamus et al. 2015). In the past 20 years, groundwater evapotranspiration showed a decreasing trend in the Wei-Ku Oasis (Table 2), which was mainly attributed to phreatic decline during 2000–2018 in the context of land-cover change and human activities (Chen et al. 2019).
The information related to climate condition, human activities, and groundwater within the Wei-Ku Oasis in 2000, 2010, and 2018
Parameters . | 2000 . | 2010 . | 2018 . |
---|---|---|---|
Air temperature (°C) | 11.5 | 11.4 | 10.9 |
Precipitation (mm) | 50.5 | 97.9 | 69.7 |
Cropland area (km2) | 3,969.3 | 4,302.8 | 4,442.0 |
Phreatic water evapotranspiration (108 m3) | 13.1 | 3.7 | 1.6 |
Depth to phreatic water level (m) | 2.7 | 4.0 | 5.7 |
Total groundwater extraction (108 m3) | 0.5 | 2.8 | 3.2 |
Agricultural water discharge (108 m3) | 3.1 | 2.3 | / |
Agricultural salt discharge (104 tons) | 170.6 | 150.9 | / |
Parameters . | 2000 . | 2010 . | 2018 . |
---|---|---|---|
Air temperature (°C) | 11.5 | 11.4 | 10.9 |
Precipitation (mm) | 50.5 | 97.9 | 69.7 |
Cropland area (km2) | 3,969.3 | 4,302.8 | 4,442.0 |
Phreatic water evapotranspiration (108 m3) | 13.1 | 3.7 | 1.6 |
Depth to phreatic water level (m) | 2.7 | 4.0 | 5.7 |
Total groundwater extraction (108 m3) | 0.5 | 2.8 | 3.2 |
Agricultural water discharge (108 m3) | 3.1 | 2.3 | / |
Agricultural salt discharge (104 tons) | 170.6 | 150.9 | / |
Note: Agricultural water discharge, agricultural water discharge by drainage canals; Agricultural salt discharge, agricultural salt discharge by drainage canals. The data of groundwater and human activities were from Wang et al. (2021).
In addition, groundwater exploitation for industry, domestic, and agricultural use is an important route of groundwater discharge in the arid inland oasis (Asoka et al. 2017). In recent decades, the water demand in the oasis has continued to increase due to the oasis expansion and rapid population growth, and surface water resources could not meet the water demand, thus leads to groundwater overexploitation for irrigation (Shukla et al. 2018; de Graaf et al. 2019). During 2000–2014, groundwater extraction exhibited an increasing trend due to the rapid farmland expansion, with an increasing rate of 2,809 × 104 m3/a (Table 2). However, groundwater extraction suddenly decreased and gradually stabilized after 2015 in the Wei-Ku Oasis, mainly because the ‘strictest water resource management system’ was implemented to effectively curb the disorderly exploitation of groundwater (Chen et al. 2019). Moreover, agricultural water discharge by drainage canals is one of the important ways for groundwater discharge in the Wei-Ku Oasis, which could lower the groundwater level and salinity to effectively prevent soil salinization (Castellano et al. 2019; Wang et al. 2021). Agricultural discharge showed a decreasing trend in our study area during 2000–2018, with a decreasing rate of 939 × 104 m3/a (Table 2), mainly due to the decrease of groundwater level and the widespread promotion of efficient water saving irrigation methods (Liu et al. 2018a; Porhemmat et al. 2018). Groundwater flows from the recharge area toward the discharge area driven by hydraulic gradient, and ultimately reaches the main stream of the Tarim River (Wang et al. 2020).
Influence of anthropogenic activities on groundwater dynamic
In recent decades, with the socio-economic development, anthropogenic activities have imposed intensive impacts on groundwater dynamics in the Wei-Ku Oasis (Wang et al. 2020). From 1990 onwards, with the decrease of river water as a result of increased agricultural irrigation in the middle and down reaches, the river channel in the oasis became dry (Chen et al. 2019). Thus, groundwater level in the Wei-Ku Oasis markedly dropped, due to surface hydrological process changes and groundwater extraction (Wang et al. 2021). Land-cover change and groundwater overexploitation were the dominant driving factors of groundwater dynamic changes in the Wei-Ku Oasis in the past decades (Chen et al. 2019). In the oasis, cropland was the dominant land-cover type, and its area gradually increased during 2000–2018 (55.9% in 2000, 62.5% in 2018) (Figure 8), which caused the increase of agricultural irrigation water demand. This, in turn, leads to the decrease of groundwater level and storage, mainly due to groundwater overexploitation (Berihun et al. 2019; Riley et al. 2019). In addition, groundwater overexploitation for agricultural irrigation in the Wei-Ku Oasis accelerated the decline of groundwater level and storage in the past decades (Table 2). This, in turn, affected the recharge and discharge of aquifer system, decreasing the amount of rainfall infiltration, canal leakage, irrigation water infiltration, phreatic evapotranspiration, and agricultural drainage (Wang et al. 2021).
Distribution of annual averaged LULC across the Wei-Ku Oasis for 2000 and 2018. LULC, land use and land cover.
Distribution of annual averaged LULC across the Wei-Ku Oasis for 2000 and 2018. LULC, land use and land cover.
Agricultural irrigation and drainage activities imposed intensive impacts on groundwater dynamic, including water quantity and quality (Porhemmat et al. 2018; Wang et al. 2021). In the past decades, large numbers of agricultural irrigation canals were constructed in the oasis due to the widespread promotion of water saving irrigation methods, which could increase canal water utilization coefficient and then affect groundwater distribution (Liu et al. 2018b). Thus, the leakage water from canals gradually reduced in the arid oasis, and could cause phreatic decline due to decreasing groundwater recharge from channel water (Zhang et al. 2014). In the Wei-Ku Oasis, the water saving irrigation area accounted for more than 50% of the total cultivated area in 2018, while the recharge of shallow aquifer from irrigation water infiltration gradually reduced from 2000 to 2018. That was because water saving irrigation methods could improve agricultural water use efficiency and save water resources, also could decrease deep percolation, and thus, reduce groundwater recharge from irrigation water infiltration (Liu et al. 2018a; Porhemmat et al. 2018). In addition, agricultural drainage activities noticeably influenced groundwater level and salinity (Table 2). Castellano et al. (2019) reported that the improved agricultural drainage could discharge some water and salt from irrigated area, thus leading to the decrease of groundwater level and salinity in shallow aquifer. Porhemmat et al. (2018) found that irrigated agriculture has a significant impact on groundwater chemistry in arid and semi-arid areas. However, we did not analyze the spatiotemporal variation of groundwater chemistry and its response to agricultural activities, and future studies on these aspects are needed. According to the above analysis and supported by the previous research, a conceptual diagram of groundwater recharge and discharge was proposed in the dried-up river oasis of arid inland basin (Figure 9).
Conceptual diagram of the groundwater recharge and discharge in the dried-up river oasis of arid inland basin.
Conceptual diagram of the groundwater recharge and discharge in the dried-up river oasis of arid inland basin.
CONCLUSION
A combined application of stable isotopic indicators and EMMA were performed to quantify the sources of shallow groundwater and interpret the recharge and discharge of groundwater under the influence of human activities in the Wei-Ku Oasis. The stable isotopes and salinity of surface water and groundwater showed obviously spatiotemporal dynamics and seasonal differentiation. Except for the northeast corner, the δ18O value of surface water in the oasis was heavier from northwest to southeast, while the spatial variation trend of shallow groundwater δ18O was not obvious. The TDS value of shallow groundwater increased from northwest to southeast, while the spatial variation trend of middle groundwater TDS was not obvious, mainly due to evaporation, mineral leaching, and mixing. Surface water and groundwater in non-flood season had higher values of stable isotopes and TDS than those in flood season, possibly due to the dominant recharge from river water in flood season. The difference of δ18O seasonal variation was not obvious among various water components, indicating the close interaction between different aquifers. The seasonal variation of TDS value was the largest in shallow groundwater, followed by surface water, while the smallest was in middle groundwater.
Anthropogenic activities are dramatically affecting groundwater dynamics in the Wei-Ku Oasis. Shallow groundwater in the oasis was mainly supplied by surface water and lateral groundwater flow from upstream, accounting for 88 and 12%, respectively, implying that surface water was the dominant source. The main routes of groundwater discharge include phreatic water evapotranspiration, groundwater extraction, agricultural drainage, and lateral flow into the Tarim River. Land-cover change and groundwater overexploitation are the main factors responsible for groundwater dynamic changes. Our findings could be useful for further understanding groundwater sources and cycling, and highlight the importance of anthropogenic activities on groundwater dynamic in arid inland oasis regions.
FUNDING
This research was jointly supported by the International Cooperation Program of Chinese Academy of Sciences (131965KYSB20210045), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDA20100303), and the Chinese Academy of Sciences ‘Light of West China’ Program (2018-XBQNXZ-B-015).
CONFLICT OF INTEREST
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
DATA AVAILABILITY STATEMENT
Data cannot be made publicly available; readers should contact the corresponding author for details.