Introduction

Arid and semiarid zones, generally defined as regions where the rate of evaporation is significantly greater than that of precipitation, occupy more than 40% of Earth’s terrestrial area. These areas are home to roughly more than a third of the world’s population (Cooley et al. 2021; Gaur and Squires 2018). Insufficient water supply is a predominant issue in these regions because water resources are rare but critical for social development and ecosystem security. Therefore, understanding hydrological processes is crucial for the efficient management and sustainable development of scarce water resources in arid regions. Precipitation, surface water, and groundwater are critical elements of the hydrological cycle. Owing to the high uncertainty and unevenness of precipitation and surface-water occurrence, groundwater typically plays an important and reliable role in arid zones (Ashraf et al. 2021). However, identifying recharge mechanisms and elucidating the evolution of groundwater in arid zones is challenging because of the unique climatic conditions in arid regions (Moeck et al. 2020; Xu and Beekman 2019).

Several approaches can be used to elucidate the water cycle processes, particularly to estimate surface runoff and groundwater recharge (Barua et al. 2021; Li et al. 2015). These approaches include, e.g., the water balance approach (Weatherl et al. 2021), multiple tracers method (Gui et al. 2022; Shi et al. 2021a, b), monitoring water-table fluctuation, and using numerical models (Wang et al. 2021a, b). Because each technique has different uncertainties and limitations, one or more methods should be selected to analyze the hydrological process based on the actual situation or unique purpose. Because of drought, using water balance methods to estimated groundwater recharge is not ideal due to the unreliable amount of water present. Moreover, because the measurement error of hydrological processes may be larger than the groundwater recharge amount (de Vries and Simmers 2002), some direct measurement methods, such as the water-level fluctuation method, are not suitable for arid regions because the associated error tends to be significant. Nevertheless, as the natural tracers, hydrochemical features, and isotopic compositions of water are the result of environmental effects (e.g., climate) or the interaction between the water and medium, they are often used to elucidate the hydrological evolution process and cycle history (Huang et al. 2021; Wang et al. 2015, 2016, 2017a, b, c, 2019). In particular, using isotope techniques has unique advantages in arid regions, such as tracing the sources of water and calculating their contributions (Li et al. 2016; Wang et al. 2022a, b). Recently, Gamboa et al. (2019) investigated the major ion and the multi-isotopic composition—87Sr/86Sr, δ11B, δ34S(SO4), and δ18O(SO4)—of groundwater from the Central Depression in northern Chile to identify the origin and evolution of groundwater solutes. Those researchers also developed a conceptual model to understand the hydrology cycle. Similarly, Alvarez et al. (2022) identified the transformation process between rainwater, lake water, and groundwater using major ions, stable isotopes, and the different surface-water/groundwater configurations in this area. Krishan et al. (2020, 2021) illustrated the spatial heterogeneity of groundwater recharge sources, and the connectivity between different deep aquifers based on isotopic variations in India. Furthermore, using hydrogeochemical data and groundwater age features, Cartwright et al. (2019, 2020) addressed the groundwater recharge mechanism and flow patterns in Australia, and highlighted the fact that river water may be vital for local ecosystem health. Similar studies have been conducted in Northwest China (Gui et al. 2022; Li et al. 2019a, b) where researchers have elucidated the relationship between glaciers, river water, and groundwater using hydrochemical and isotopic methods.

As an essential arid region, Central Asia (CA) has also been facing a severe water crisis of late (Guo et al. 2018; Jiang and Zhou 2021). CA stretches from the Caspian Sea in the west to western China and Mongolia in the east, and from Afghanistan and Iran in the south to Russia in the north (Karthe et al. 2015; Fig. 1a). Considering the similarity of the hydrological cycles, this report also included China’s Xinjiang region in the study area. Historically, CA was closely tied to the Silk Road trade routes, acting as a crossroads for the movement of people, goods, and ideas between Europe and the Far East. At present, it is still an important part of the trading and agricultural area. Water resources are an essential element for the survival of local people, as well as for maintaining local economic development and ecological security. However, the current research focuses on small locally specific catchments, and the lower resolution is not sufficient to understand the hydrological cycle in CA as a whole. Furthermore, systematic hydrological cycle reviews for this region are very rare, which leads to a lack of clarity on the recharge mechanisms and evolution processes of the river and groundwater. In addition, the climatic conditions in CA can significantly affect water quality (Tweed et al. 2011). Therefore, it is necessary to evaluate the suitability of CA for irrigation and drinking based on current water chemistry data.

Fig. 1
figure 1

a Map showing the geographical location of the study area in the world while highlighting the drylands (Metternicht et al. 2015). b An ground surface elevation map of Central Asia (Shuttle Radar Topography Mission data)

Full size image

Although several studies using hydrochemical and isotopic methods have been conducted on individual catchments distributed in CA (Sun et al. 2016; Wang et al. 2020; Wu et al. 2019a, b, 2020; Xiao et al. 2015, 2012; Zhang et al. 2019), systematic review is lacking, which hinders our understanding of regional water resources. Therefore, 222 sets of chemical data and 390 sets of isotope data were collected to build a baseline database, which includes hydrochemical and isotopic data for precipitation, river water, and groundwater in the study area. These data were then reviewed to provide an overview of the hydrological cycle in CA, and elucidate the origin and natural evolution of surface water and groundwater. Finally, a conceptual model of the hydrological process was constructed based on these data. Baseline databases not only allow for an understanding of the current status of water quality in the study area, but they also provide a comparison for future changes in the evolution of water quality and potential recharge source conversion. Additionally, attempts were made to highlight how to carry out future research efforts to aid water resource management in the arid region.

Study area and data sources

Study area description

Central Asia consists of the five former Soviet Union states Kazakhstan, Turkmenistan, Uzbekistan, Tajikistan, and Kyrgyzstan (Fig. 1a). It stretches from the Caspian Sea in the west to China in the east and from Afghanistan in the south to Russia in the north. Because the melting water from glaciers and springs in the Tianshan Mountains (Mt. Tianshan) and Pamir Mountains feed most of the rivers in the study area (Chen et al. 2018), some major rivers in the Xinjiang Uygur Autonomous Region of China also originate from these mountains. Consequently, Xinjiang is also included in the study area, which together constitutes the CA referred to in this study (Fig. 1a). Therefore, the study area covers approximately 7 million km2, with Mt. Tianshan and the Pamir Mountains at the center, and covers five plains and basins. As the areas with the highest elevation in the study area, the elevation distribution of the Pamir Mountains and Mt. Tianshan is approximately 5,500–7,000 meters above sea level (m asl) (Fig. 1b), and a series of rivers, plains, basins, and low-elevation mountains are distributed around them. The lowest points in CA are in the coastal region of the Caspian Sea, the western part of Kazakhstan and Turkmenistan, at –151 m asl. There are four major rivers (defined as catchment areas larger than 140,000 km2) in the study area: the Amu Darya River, Syr Darya River, Ili River, and Tarim River (Fig. 1b).

Central Asia is situated in the center of the Eurasian continent (Fig. 1a), far away from the ocean, with a prevailing continental climate, therefore, it is characterized by low precipitation and high evaporation, particularly in the plains and basins (Jiang et al. 2020; Peng et al. 2018). Consequently, this region can be characterized as arid or semiarid, and the climate is mostly controlled by westerly air circulation currents (Chen et al. 2018; Liu et al. 2019). However, extremely low precipitation often occurs in deserts, whereas high frequency and large amounts of precipitation occur mainly in the high-elevation mountains—for example, annual precipitation below 100 mm was measured in Aral-Caspian desert regions and the Tarim Basin, whereas precipitation of up to 1,000 mm/year frequently occurs on Mt. Tianshan and in the Pamir Mountains (Fig. 2). Combined with glacial meltwater, precipitation feeds the four major rivers and many oasis plains (Barandun et al. 2021; Chen et al. 2019). Notably, the annual precipitation in habitable plains and oases is approximately 200–300 mm/year (Yang et al. 2020; Fig. 2). The water crisis has become increasingly severe owing to global climate change and human activities in recent years, which have led to increased impacts on agricultural production, animal husbandry, water resources, economies, and human health (Mirlas et al. 2015, 2021, 2022). The average air temperature peaks can reach 30 °C in July and fall below –2 °C in January.

Fig. 2
figure 2

Distribution of the average annual precipitation in the study area over 1970–2000 (data source: (Fick and Hijmans 2017)) and the locations of sampling sites

Full size image

Geology and hydrogeology

The study area is located to the west of the Central Asian Orogenic Belt (Wang et al. 2022a, b). The Pamir Mountains and the Mt. Tianshan area are exposed to a large number of pre-Quaternary strata consisting of crystalline and sedimentary rocks. Quaternary deposits are widely developed in the front of the mountains (Jiang et al. 2015; Tan et al. 2019; Wang et al. 2018). The high-elevation mountain zone mainly consists of the Precambrian basement represented by high-grade Archean and Proterozoic gneisses, schists, carbonates, and clastic rocks, as well as overlying Cambrian–Silurian sedimentary rocks, carboniferous basaltic to intermediate volcanic rocks, and Permian volcanic and sedimentary rocks. Furthermore, numerous Ordovician–Permian granitoids have intruded the Precambrian basement and Paleozoic strata.

The water-bearing strata consist primarily of unconsolidated Quaternary deposits in the study area, and they are mainly distributed in the Quaternary alluvial-pluvial sediments in the piedmont plains of the four major rivers. Because unconsolidated sediments are permeable, they allow a certain amount of surface runoff to infiltrate and recharge the aquifer, particularly in the piedmont fans. Moreover, groundwater, flowing from high to low altitudes and discharged by springs, wells, and via evaporation (Ostrovsky 2007), is naturally recharged by glacial meltwater and precipitation in high-elevation mountains. Therefore, mountainous areas are usually recharge regions, and both the Aral and the Taklimakan deserts are the groundwater discharge region (Fig. 3).

Fig. 3
figure 3

Hydrogeological map of the study area. (Data source: Institute of Hydrogeology and Environmental Geology, Chinese Academy of Geological Sciences, 2019.07)

Full size image

Data and analysis

Data sources

Stable isotopic data (hydrogen and oxygen isotopes) and hydrogeochemical data for precipitation, river water, and groundwater were collected for analysis. The analytical accuracy of the isotopes varied owing to different data sources, and is listed in Table S2 in the electronic supplementary material (ESM). The isotope data were expressed as delta values, defined as the per mil deviation from the Vienna Standard Mean Ocean Water (VSMOW) using the following formula,

$$\delta =\frac{R_{\mathrm{sample}}-{R}_{\mathrm{standard}}}{R_{\mathrm{standard}}}\times 100$$
(1)

where R represents the isotope ratios (2H/1H or 18O/16O) of the sample and standard.

To understand the isotopic characteristics of precipitation in the study area, data were collected from three long-term monitoring sites set up by the International Atomic Energy Agency (IAEA) in the study area, including 185 sets of δ18O and δD values. These three sites are in Tashkent in Uzbekistan, and Urumqi and Hetian in Xinjiang region, China. The sampling periods for the Tashkent, Urumqi, and Hetian sites were from 1969–1983, 1986–2003, and 1988–1992, respectively. Local meteoric water lines (LMWL) were established using linear regression based on these data.

The hydrochemical and stable isotope data of river water and groundwater were collected from published literature and are listed in Tables S1 and S2 in the ESM. These data covered the four major rivers in the study area (Wu et al. 2019a, b, 2020). The stable isotopic compositions of river water and groundwater were expressed as standard δ notation in units per mil relative to VSMOW too. The measurement accuracy of δ18O and δD has been included in Table S2 in the ESM. Due to equilibrium Rayleigh condensation processes, deuterium excess (d-excess) was defined as offsets in δD from the global meteoric water line (GMWL); thus, it can be used to identify the influence of evaporation on the isotope compositions in various types of water using the following equation (Dansgaard 1964):

$$\text{d-excess}=\delta \mathrm{D}-8\times {\delta}^{18}\mathrm{O}$$
(2)

Data collected from literature focus on the four river basins and do not cover the whole of CA. Therefore, hydrochemical and isotopic results of different water bodies focus on the central-eastern part of CA. In addition, aquifers in the plains tend to be multilayered as deposition occurs in multiple stages during geological history. Owing to insufficient groundwater data from the different aquifers, all groundwater is discussed as a whole.

Statistical analysis method and software

Linear regression in Excel was used to assess the relationships between δ18O and δD values in water and to plot X and Y graphs. Arc Map v.10.3 was used to prepare a map of the study area and the digital elevation model (DEM) was acquired from the Shuttle Radar Topography Mission (STRM) to derive Fig. 1. Saturation index (SI) values were either collected from the literature or calculated using PHREEQC software (Parkhurst and Appelo 1999).

A Piper tributary diagram was designed to reveal the dominant cations and anions, and differences in the chemistry of the waters. It was also used to help understand the sources of dissolved ions in the waters (Karmegam et al. 2011; Teng et al. 2016).

The suitability indices for drinking and irrigation can be expressed as excessive sodium concentrations in the water and sodium adsorption, which were calculated as follows (Shaki and Adeloye 2006)

$$\mathrm{Na}\%=\mathrm{Na}^+/\left(\mathrm{Na}^++\mathrm{K}^++\mathrm{Ca}^{2+}+\mathrm{Mg}^{2+}\right)$$
(3)
$$\mathrm{SAR}={\mathrm{Na}}^{+}/\sqrt{\left({\mathrm{Ca}}^{2+}+{\mathrm{Mg}}^{2+}\right)/2}$$
(4)

where all the ionic concentrations are expressed in μeq/L of the corresponding ions.

In addition, the total hardness (TH) of water was measured by adding the concentrations of calcium and magnesium, and converting this value to an equivalent concentration of calcium carbonate (CaCO3) in milligrams per liter (mg/L) of water, which can be calculated as follows (Eyankware et al. 2018; Sawyer and McCarty 1967):

$$\mathrm{TH}=0.1\times \left({\mathrm{Ca}}^{2+}+{\mathrm{Mg}}^{2+}\right)$$
(5)

where the Ca2+ and Mg2+ concentrations are expressed in μmol/L and the unit of TH is mg/L. According to World Health Organization guidelines (WHO 2010), water TH at concentrations below 60 mg/L is generally considered soft, 60–120 mg/L is moderately hard, 120–180 mg/L is hard, and >180 mg/L is very hard.

Results

To facilitate the data presentation and discuss hydrochemistry characteristics, the results are presented based on the four major catchments; the Amu Darya River Basin, Syr Darya River Basin, Ili River Basin, and Tarim River Basin. All data are displayed in Table S1 in the ESM, while the summary statistics are presented in Table 1.

Table 1 Summary statistics for the chemical and stable isotope composition of river water and groundwater in Central Asia
Full size table

Characteristics of δ 18O, δD, and d-excess in precipitation

The δ18O and δD values of precipitation in Toshkent, Urumqi, and Hetian sites set by IAEA (Fig. 2) were collected from IAEA and WMO (2021). The δ18O values ranged from −29.81 to 3.78‰, with an average of −11.39‰. The minimum δD value was −236‰, and the maximum was 45.5‰, with an average of −79.34‰. As for the monthly means, the isotopic values were notably depleted from December to February of the following year, with a minimum average of −20.58‰ and −143.4‰ for δ18O and δD, respectively. Furthermore, more positive values occurred from June to August, (–5.4‰ and –37.8‰ for δ18O and δD, respectively). Based on these 185 datasets, the LMWL for CA was derived as δD = 7.52 δ18O + 6.29 using linear regression (R2 = 0.9515, n = 185; Fig. 4). The d-excess values of precipitation ranged from –44.5 to –54.84‰, with an average of 11.75‰.

Fig. 4
figure 4

Distribution of stable isotopic values compared with the local meteoric water line (LMWL) of Central Asia. The gray dashed line is the LMWL of δD = 7.52 δ18O + 6.29

Full size image

Hydrochemical and stable isotopic characteristics of river water

Amu Darya River

The Amu Darya River originates in the Pamirs Mountains and flows east to west. River water sample data collected from the literature were mainly from the upper reaches of the river near the Pamirs. The δ18O values of these samples varied from –17.8 to –9.3‰, with a mean value of –14.02‰, whereas the δD values varied from –132.7 to –66.1‰, with a mean value of –99.44‰ (Fig. 4; Liu et al. 2015; Wu et al. 2020). River water δ18O and δD values showed a significantly negative trend and were distributed in the lower left corner of the graph. In addition, the river water d-excess values had a wide range from 2.7 to 21.9‰, with an average of 12.67‰.

The pH of the Amu Darya River water was found to be alkaline, ranging from 7.4 to 9.0, while the TDS of the river water varied from 59.6 to 1,400 mg/L, with an average of 332.4 mg/L (Wu et al. 2020). The major cations and anions (meq %) were plotted on a Piper diagram and further projected onto the central diamond field to evaluate the hydrochemical facies and different types of water (Fig. 5). In the cation plot, most of the Amu River samples fell in the left corner, where the Ca2+ values are higher (>50%), demonstrating the dominance of Ca2+ in the river water. In contrast, the anion diagram showed that most of the water samples were distributed in the lower left corner, indicating that it is characterized by the dominance of HCO3− over Cl− and SO42−. The results of the central diamond field in the Piper diagram show the overall chemical characteristics of the river water samples. Alkaline earth elements (Ca2+ and Mg2+) were dominant over alkaline elements (Na+ and K+) and weak acids (HCO3−) over strong acids (SO42− + Cl−) in these water samples, respectively. Therefore, the river water samples from the upper Amu Darya were distributed in the zone for Ca–HCO3 types.

Fig. 5
figure 5

Piper ternary diagram of the major ions for water samples from the Amu Darya River, Syr Darya River, Ili River, and Tarim River. The mean values of major ions were used for the data from the Syr Darya River, and raw data were used for the others

Full size image

Syr Darya River

Similar to Amu Darya, Syr Darya also originates in the Pamirs Mountains, flows from east to west, and finally empties into the Aral Sea. The Syr Darya water sample data were collected mainly from the upper and middle reaches (Wu et al. 2019a). The δ18O values of the water samples from Syr Darya River ranged from −14.7 to –6.2‰, with a mean value of –11.1‰, and the δD values ranged from −104.8 and −45.8‰, with a mean value of –74.2‰. Moreover, the d-excess value in Syr Darya showed a wide range, from 2.5 to 32.7‰, with an average of 13.9‰.

The river water of Syr Darya is weakly alkaline, with a pH value ranging between 7.95 and 9.31. Moreover, the TDS values ranged from 342 to 4,010 mg/L, with a mean value of 1,140 mg/L. More than 95% of the river water samples had a TDS value greater than 500 mg/L, suggesting that the river water is slightly saline with higher hardness (Zhang et al. 2019). According to the Piper diagram, the main anions were SO42−, followed by HCO3− and Cl− , in order of concentration, whereas the main cations were dominated by Ca2+ and Na+. Spatially, a high TDS commonly occurs in the region of the Aral Sea, which is the lower reaches, whereas a low TDS is typically found in the upper reaches. As a result, most of the Syr Darya river water was characterized by significantly more alkaline earth elements (Ca2++Mg2+) than alkali elements (Na++K+), while strong acids (SO42− + Cl−) exceeded weak acids (HCO3−) in abundance.

Ili River

The Ili River originates from Mt. Tianshan, flows from China to Kazakhstan, and ends at Balkhash Lake. The isotopic distribution range of the river water is relatively narrow, in which the δ18O and δD values vary from −89.4 to –71.4‰ with a mean value of – 80.61‰ and from −13.1 to – 0.3‰ with a mean value of −11.7‰, respectively. The d-excess was generally larger than 10‰, with an average of 12.97‰, and ranged from 9.7 to 17.1‰ in the Ili River Basin (Ma et al. 2016).

The pH of the Ili River water varied from 7.43 to 8.26 (Ma et al. 2016). Meanwhile, the TDS values of the Ili River water were relatively low, with a minimum value of 91.3 mg/L, a maximum value of 608 mg/L, and a mean value of 288.8 mg/L (Shen et al. 2021). The order of cation concentration from high to low in the Ili River water was Ca2+, Na++, Mg2+, and K+, whereas that for the anions were HCO3−, SO42−, and Cl−. The Piper diagram showed that most of the river water samples fell in the HCO3−–Ca2+ zone (Fig. 5).

Tarim River

The Tarim River originates from the Kunlun Mountains in the eastern part of the Pamirs and flows through the vast Taklimakan Desert. Therefore, the spatial variabilities of the hydrogeochemical and isotopic values are very significant. Stable isotope data for water from the Tarim River showed that δD values varied from −96.1 to – 22.2‰ with a mean value of –62.57‰, and δ18O values ranged from −14.1 to –5.0‰ with a mean value of −9.40‰. Furthermore, the d-excess had a wide range, from –2.5 to 39.7‰, and the mean value was 12.65‰ (Sun et al. 2016).

The pH value of the Tarim River water varied from 8.3 to 9.7 (Xiao et al. 2012). Meanwhile, the TDS varied from 158 to 11,190 mg/L with a mean value of 1,000 mg/L, which tended to be lower in the upstream areas and higher in the downstream desert areas. Because the dominant cations were Ca2+ + Mg2+, and the major anion was HCO3−, the water belongs to the Ca–Mg–HCO3 type.

Hydrochemical and stable isotopic characteristics of groundwater

Previous studies on groundwater hydrochemistry and isotopes in the Amu Darya-Syr Darya catchments were mainly focused on the Aral Sea, which is the end point of the two rivers in the lowest part of the basin (Schettler et al. 2013a, b; Vetshteyn et al. 1983). Therefore, the groundwater samples were collected in this artesian basin, mainly from artesian springs and wells. The δD and δ18O values of these groundwater samples from the Aral Sea drainage basin varied from −93 to +31‰, with a mean value of –50.01‰, and δ18O values ranged from −14.5‰ to +6.2‰ with a mean value of −5.6‰ (Fig. 6). The d-excess values ranged from –79.6 to 42.8‰, with a mean value of –4.8‰. Lastly, the pH values were close to 7 and the TDS ranged from 1,400 to 53,300 mg/L (Table S3 in the ESM).

Fig. 6
figure 6

Plot of the δD and δ18O in groundwater across Central Asia. The gray dashed line is the local meteoric water line of δD = 7.52 δ18O + 6.29

Full size image

The groundwater samples from the Ili River Basin were collected from unconsolidated deposits (Ye 2015). The results showed that the δD and δ18O values of groundwater varied from −109.09 to –72.75‰ with a mean value of –81.25‰, and δ18O values ranged from −14.08‰ to –10.57‰ with a mean value of −11.92‰ (Fig. 6). Furthermore, the d-excess values varied from 3.55 to 24.62‰, with a mean value of 14.0‰, and more than 90% of the samples had d-excess values greater than 10‰. The pH values ranged from 6.33 to 8.48, and the TDS ranged from 197.8 to 3,300 mg/L (Table S3 in the ESM).

Wang et al. (2020) showed that the δD and δ18O values of groundwater in the Tarim Basin varied from −85.8 to –17.2‰ with a mean value of –62.22‰, and δ18O values ranged from −11.48 to –0.9‰ with a mean value of −8.88‰ (Fig. 6). The d-excess values varied from –13.48 to 18.75‰, with a mean value of 9.56‰. The hydrogeochemical results of the groundwater samples collected from the Tarim Basin were neutral to weakly alkaline, with pH values that ranged from 7.2 to 9.8. The TDS varied from 14.5 to 27,500 mg/L, with a mean value of 2,180 mg/L (Table S3 in the ESM). Groundwater with low TDS is mostly of type Ca–Mg–HCO3 , whereas high-salinity groundwater is of the Na–Cl type (Xiao et al. 2015).

Discussion

Water recharge sources and hydrological processes

Isotopic characteristics of precipitation

Precipitation plays an important role in recharging surface water and groundwater (Wang et al. 2019). Stable hydrogen and oxygen isotopes are often good tracers for identifying the recharge sources of river water and groundwater, as a result, they have been used in many studies (Geris et al. 2022; Liu et al. 2022; Riedel and Weber 2020). Based on the monitoring sites established by the IAEA in the study area, the LMWL in CA was established using linear regression to analyze the relationship between precipitation, river water, and groundwater.

Both the slope and intercept of the LMWL in the study area were lower than those of the global meteoric water line (GMWL: δD = 8 δ18O + 10, Craig 1961), and the mean d-excess value of these 185 precipitation data was 11.75‰. Although water vapor is dominated by the westerlies in these areas, this also indicates that evaporation may play an important role during precipitation events. Usually, the secondary evaporation of falling raindrops is likely to occur in arid regions due to a dry climate, and this will result in isotopic non-equilibrium fractionation occurring within the moisture (Wang et al. 2017a, b, c, 2019).

Isotopic characteristics of river water

The δD and δ18O distribution of the river water fell along the LMWL and very close to it in the plot of δD versus δ18O (Fig. 4), indicating that the river water mainly originates from local precipitation. In addition, some of the water vapor that forms precipitation in CA originates from the Mediterranean region, which has a significantly high d-excess value (Bershaw 2018; Shi et al. 2021a, b). Figure 7 shows that most of the river water also has a d-excess value that is close to or greater than 10‰, indicating a close relationship between river water and rainwater, because the d-excess value of global precipitation is 10‰. However, the amount of rainwater is very low in most parts of the study area, and precipitation rarely occurs, particularly in the desert regions, which has resulted in a severe shortage of local water resources (Chen et al. 2021; Li et al. 2020). Therefore, it is difficult to maintain these four major rivers in the study area using precipitation alone, suggesting that there were other sources of water feeding the rivers.

Fig. 7
figure 7

Distribution of d-excess values for river water in the study area. The blue solid line marked 10‰ is the average value of global precipitation

Full size image

According to the altitude effect of stable isotopes in precipitation, the δ18O value will decrease by 0.17–0.44‰ when the altitude increases by 100 m (Bershaw and Lechler 2019; Juhlke et al. 2019), which causes the precipitation in high-altitude mountainous areas to have significantly negative characteristics. In addition, the isotopes of glacial meltwater also have negative characteristics (Boral and Sen 2020). As a result, the δD and δ18O values of river water samples obtained from the Amu Darya were clearly negative, because most of them were taken from the high-altitude upper reaches, the Pamir Mountains. On one hand, there is a significant positive correlation between rainwater amount and altitude in most arid areas (Wang et al. 2017a, b, c, 2019), resulting in abundant rainfall in high-altitude regions, which can provide water sources for rivers. On the other hand, glacial meltwater in high-altitude areas is also another important source of recharge for river water, which has been proven in these areas (Chen et al. 2019; Guo et al. 2019). This also explains the negative isotopic values in water samples from the Ili River and Syr Darya River.

The isotopic characteristics of river water are unavoidably affected by evaporation because it is very strong in the study area. According to the Rayleigh Fractionation theory (Craig et al. 1956; Yang et al. 2019a, b), the δD and δ18O values in water will be more positive if it undergoes intensive evaporation. Because some samples of water from the Tarim River and the Syr Darya River were collected from the middle and lower reaches, they were significantly affected by evaporation, resulting in more positive δD and δ18O values (Fig. 4).

Isotopic characteristics of groundwater

Figure 6 shows that the groundwater in the study area also originates from precipitation because almost all the isotopic scatter points of groundwater are distributed along the LMWL. Compared with river water, the range of stable groundwater isotopes is narrower and more positive. Evaporation may occur in the process of recharging groundwater from rivers in arid regions, which can lead to the creation of positive isotopes (Hao et al. 2019; Wang et al. 2016). Therefore, it can be inferred that the groundwater may inherit the isotopic characteristics of river water and undergo evaporation in the study area. Specifically, in plains and basins, river water infiltration is an important source of groundwater recharge, owing to limited precipitation.

Whether it is river water or groundwater, they all show high d-excess values in the Ili River Basin, and these values are larger than 10‰, with some of them being close to 20‰ (Fig. 7). This indicates that, in addition to precipitation, local water vapor recirculation also plays an important role in the isotopic signatures of river water and groundwater in the Ili River Basin (Wang et al. 2019). The geomorphological and topographic features of this area also provide the necessary conditions for water recirculation (Cao et al. 2022). Similarly, the mean d-excess values of groundwater in the Tarim River Basin are very close to 10‰, indicating a close relationship between groundwater and precipitation. However, there is a wide range of the d-excess values for groundwater in the artesian basin of the lower Amu Darya-Syr Darya. The lower d-excess values may have been caused by recharge occurring during different geological historical periods.

Factors controlling the water chemistry

River water

Gibbs plots can be used to identify the influences of rock weathering, evaporation, and precipitation on water chemistry. Figure 8 shows that most of the Tarim River samples fall into the evaporation end-member, indicating that evaporation is a key factor in the chemical formation of river water. The Syr Darya River had similar characteristics (Fig. 8). In contrast, the Amu Darya River water is mainly concentrated in the rock-weathering end-member. The Amu Darya River water samples were mainly collected from the upper reaches of mountainous areas, therefore, weathering influenced the chemical features of the river water. The chemical features of the Ili River water lie between rock weathering and evaporation, but most fall into the rock-weathering end-member, indicating that rock weathering still plays a more dominant role than the other processes.

Fig. 8
figure 8

Gibbs diagrams of the water from the Amu Darya River, Syr Darya River, Ili River, and Tarim River

Full size image

Considering the characteristics of the lithology in the study area (Li et al. 2019a, b; Niu et al. 2021; Yang et al. 2019a, b), the chemical type river water controlled by rock-weathering end-members is usually the Ca–HCO3 type, but it belongs to the Na–Cl type if the river water is more dominated by evaporation. Therefore, the chemical type of river water, including the Amu Darya River, Ili River, and the upper reaches of the Tarim River, are type Ca–HCO3, have a low TDS value, and are controlled by rock weathering. The chemical type of river water in the lower reaches of the Tarim River is type Na–Cl, has a high TDS value, and is controlled by evaporation.

Groundwater

The solutes in groundwater usually come from the interaction between water and rock; therefore, dissolution and precipitation control the chemical features of groundwater. The saturation index (SI) of minerals in groundwater can be used to understand the under-saturation, saturation, and oversaturation states of the mineral phases.

Minerals such as dolomite, aragonite, and calcite were in an oversaturated state in most of the groundwater (more than 93%) in the Ili River Basin, indicating that these three minerals are difficult to dissolve in groundwater to change their chemical composition. The SI values of other minerals such as magnesite, gypsum, and halite, were less than 0, suggesting that these minerals can still dissolve into groundwater.

The calculated SI of the groundwater in the Tarim Basin was ranked in the following order: dolomite > calcite > aragonite > magnesite > gypsum > halite (Xiao et al. 2015; Table S2 in the ESM). The SI of the first three minerals were greater than 0, suggesting that they were all in an oversaturated state. This is likely due to the extremely arid climate of the Tarim Basin. Evaporation will continue to concentrate the ions in groundwater, driving minerals such as dolomite, calcite, and aragonite, toward saturation and over-saturation. These three minerals determine the concentrations of Ca2+, Mg2+, and HCO3– in groundwater. However, the SI of gypsum and halite were less than 0, suggesting that they were under-saturated, and can still dissolve into groundwater and increase the concentrations of Na+, Cl–, Ca2+, and SO42–. The increase in Ca2+ also promotes the oversaturation of dolomite, calcite, and aragonite. In addition, there was a clear negative correlation between (Ca2+ + Mg2+ – HCO3– – SO42–) and (Na+ – Cl–) for the groundwater in the Tarim Basin, indicating that the cation exchange reaction is an important controlling factor of the ionic composition for groundwater.

Water quality for drinking and irrigation

Rock weathering and evaporation control the natural ionic compositions of rivers in the study area; therefore, the unique lithology and climate are the dominant factors in determining the hydrochemistry of river water. Although human activities have also affected water quality (Liu et al. 2021), here the focus was on the role of natural processes in determining the chemical composition of water. The ionic concentrations in most of the river water samples in the study area were remarkably higher than those for the global averages (Meybeck 2003); nevertheless, some were still suitable for drinking and irrigation based on the indicator mentioned in the preceding.

As most of the river water samples in the Amu Darya were in the headwater region, both the Na+% and SAR of the water samples fell between good and excellent, indicating that all water samples were suitable for irrigation purposes. Over 65% of the river water samples from the Amu Darya belonged to the moderately hard category. Overall, river water from Amu Darya Rivers falls in the safe category for drinking and irrigation purposes, with the exception of a few samples (Wu et al. 2020).

Based on the surface-water samples collected from the Syr Darya, approximately 90.48% (38 samples) of the water samples were suitable for irrigation, and only 2.38% (one sample) were unsuitable for irrigation based on the Na+%. Nevertheless, all water samples were excellent according to the SAR. The minimum value of TH was 169.9 mg/L, which is very close to the limit of the very hard category.

All river water and groundwater samples collected from the Ili River were deemed excellent for irrigation, based on Na+%, but only 62.5% of river water and 40% of groundwater were suitable for irrigation based on the SAR. Less than 30% of the river water samples were very hard but more than 80% of the groundwater samples were very hard, indicating that river water quality was significantly better than that of groundwater in the samples collected from the literature.

The high salinity content in the water is one of the main characteristics of the river water in the Tarim Basin. It has caused many adverse effects not only on production and the life of the residents but also on the ecological environment along the river. According to the Na+%, more than 90% of the river water samples were suitable for irrigation, but only 32% were suitable for irrigation according to the SAR. Owing to its high salinity, river water and groundwater quality in the Tarim Basin is not very suitable for drinking. Over 40% of the river water samples belonged to the very hard category and less than 10% of the groundwater samples were of good quality.

Factors that affect water’s suitability for drinking and irrigation include not only Na+%, SAR, and TH values, but also the presence of substances such as heavy metals, organic pollutants, antibiotics, radionuclides, etc. (Rajasekar et al. 2022; Zhou et al. 2022). In particular, some anthropogenic activities, such as irrigation and mining, may have caused water pollution in local areas (Liu et al. 2021), resulting in some water samples not meeting drinking and irrigation standards. Owing to a lack of data, this study focused only on evaluating natural factors. The deterioration of water quality caused by anthropogenic factors is an issue that requires urgent attention in the future.

Conceptual model

According to the discharge amount of the four rivers in the study area, the rivers are characterized by strong seasonality, with peak flows occurring in summer, reaching their maximum in summer when glaciers melt and rainfall amounts are at their maximum (Hao et al. 2015; Isupova 2019; Lobanova and Didovets 2019). It is difficult to distinguish the contribution of glacial meltwater from that of rainfall to river flow. However, the composition of stable isotopes in water can also be used to track hydrological processes. As part of this study, there was an attempt to use isotope-based end-member mixing analysis (EMMA; Hooper et al. 1990) to perform hydrograph separation in the study area. The estimations showed that the contribution of the Pamir Mountains’ glacier meltwater to the runoff of the Amu Darya River was ~50%, whereas the contribution of the Mt. Tianshan glacier meltwater to the runoff of the Ili River was ~40% (Table S3 in the ESM). This is consistent with the results of Chen et al. (2019) who calculated the contribution of glacial meltwater in the different catchments over Mt. Tianshan. The EMMA was not applied to the Syr Darya and Tarim River catchments because of the lack of reliable endmembers in this study.

Based on the aforementioned analysis of the geological characteristics, hydrogeological features, and climatic conditions of CA, as well as the hydrochemical and isotopic data of precipitation, river water, and groundwater collected from the published literature, a general conceptual model of the hydrological process in the study area was established (Fig. 9).

Fig. 9
figure 9

A general conceptual model of the hydrological processes in the study area

Full size image

In the headwater region of these four major rivers, precipitation and glacier meltwater are the main recharge sources according to the isotope data. Rock weathering is the dominant factor in river water chemistry in this region, and low-salinity river water is found here. Driven by dry and extremely dry climatic features, strong evaporation in CA controls the chemical composition of river water, resulting in the formation of high salinity river water. Typically, the TDS values of river water increase gradually along the river. Although some river water is not suitable for irrigation due to high Na+% or high SAR indices, nor is it very safe for drinking according to the standards set by the WHO, the river water is still the most important source for irrigation and drinking in CA, considering the scarcity of water resources in the study area.

The infiltration of precipitation and river water is the main source of groundwater recharge in the study area. The interaction between groundwater and rocks in the aquifer influences the groundwater’s chemistry. Dolomite, aragonite, and calcite are oversaturated in most groundwater samples, and more than half of the groundwater samples are unsuitable for irrigation and drinking, especially in the Tarim Basin. This finding indicates that groundwater resources are not able to alleviate the local water crisis, although groundwater is generally regarded as a stable and high-quality water resource.

Summary and conclusion

Central Asia, the farthest place from the ocean, has long suffered from water resource shortages owing to its arid climate. Although several studies focusing on individual catchments in CA have been conducted using hydrochemical and isotopic data, knowledge of hydrological processes is still lacking. A comprehensive understanding of this region’s hydrological processes and water quality is of great significance for the rational development and utilization of water resources. Hydrochemical and isotope data on precipitation, river water, and groundwater were collected in the Amu Darya, Syr Daya, Ili, and Tarim rivers, which are the four major rivers in CA. Based on these data, which were published in the literature, the relationships between precipitation, river water, and groundwater were analyzed and clarified. In addition, the factors controlling the hydrochemical features and solute sources of the water were revealed.

According to the δD and δ18O values of precipitation provided by IAEA, the LMWL was established as follows: δD = 7.52 δ18O + 6.29 (R2 = 0.9515; n = 185), where the slope and intercept are lower than those of the GMWL. The isotopic characteristics of precipitation are controlled by the water vapor source and secondary sub-cloud evaporation. The abundant precipitation in mountains, such as the Pamir Mountains and Mt. Tianshan, is the main source of water for the river because the isotopic scatter values are closely distributed along the LMWL. Rock weathering is a dominant factor in river water chemistry, but evaporation gradually becomes the controlling factor in the middle and lower reaches of the river because of the high evaporation in the study area. The isotopic values of the river water were more positive, indicating that the river water had undergone intensive evaporation. In addition, the isotopic characteristics of that groundwater were very similar to those of river water, but the scatter values were more concentrated. This indicates that the relationship between groundwater and river water is very close, and that river water might be the only recharge source for groundwater. In the majority of the groundwater samples collected, the SI values of dolomite, calcite, and aragonite were greater than 0, indicating that these minerals in the groundwater have become oversaturated and have begun to precipitate because of strong evaporation and long-term interaction between the groundwater and aquifer. Based on information derived from hydrogeochemical and isotopic data, a conceptual hydrological model of the study area was established to clarify the water cycle process. Finally, the suitability of river water and groundwater for irrigation and drinking was evaluated using the Na+%, SAR, and TH. The results show that most of the river water and groundwater samples are suitable for irrigation, but not for drinking, because of their high TH values.

Although a large amount of data was collected in this study to illustrate the hydrological processes in CA, long-term field observations of isotopic and ionic compositions from river water, groundwater, and glacier meltwater should be intensified in the future to understand the potential evolution of hydrological processes driven by climate change. First, the contribution of glacial meltwater to river water must be accurately quantified. In headwater regions, such as Mt. Tianshan and the Pamir Mountains, glacial meltwater is the most important source of river water in addition to precipitation, but the current level of understanding of glacial meltwater is still poor. Second, the long-term series monitoring of river water is not enough to clearly understand the changing processes of water quality over time. This will help to expand knowledge of water quality throughout the year. Third, groundwater quality at different depths in the study area should be clarified by more extensive monitoring in the future to protect human health, because poor-quality groundwater is often found in arid regions, and most of it is not suitable for irrigation or drinking. In addition, because some anthropogenic activities, such as irrigation and mining, may have caused surface water and groundwater pollution in CA, these types of activities need to be monitored more closely in the future.