Hydrochemical characteristics and exchange dynamics between surface water and groundwater in an arid river basin

The Weigan River Basin, located in the arid region of northwest China, faces severe water scarcity. The complex interactions between surface water and groundwater pose a critical challenge for accurately assessing total water resources. To elucidate the exchange mechanisms and fluxes, this study employs a comprehensive analytical approach integrating hydrochemistry, stable isotopes (δ¹⁸O, δD), and Bayesian mixing models (MixSIAR). Hydrochemical analysis reveals a patterned spatial evolution of water chemistry characteristics within the basin. Mountainous surface waters predominantly exhibit a HCO₃·SO₄-Ca·Mg type, controlled by the weathering of carbonate and silicate rocks. Groundwater chemistry evolves along the flow path from an HCO₃·Cl-Na·Ca type to an HCO₃·SO₄-Na·Ca type, revealing groundwater recharge from surface water rich in SO₄^2-. In the plains, groundwater undergoes further evaporation and concentration, cation exchange adsorption, and human activities, eventually discharging into surface water and causing elevated Na⁺ levels in rivers. Based on these insights, MixSIAR model quantification reveals a clear and statistically significant spatiotemporal transformation pattern. In mountainous sections (Heizi River, Karasu River, Tairweichuk River, upper reaches of both the Muzhati River and Weigan River), surface water serves as the primary groundwater recharge source (dry period contribution: 59%–70%; wet period contribution: 54%–59%). Conversely, in the plain areas of the lower reaches of both the Muzhati and Weigan Rivers, groundwater replenishes surface water (dry period contribution: 53%–55%; wet period contribution: 56%–63%). Seasonally, surface water contribution during the dry period is on average 7.6% higher than during the wet period. In contrast, groundwater contribution in the plain region is on average 5.5% higher during the wet period than during the dry period. Through a research approach combining geochemical tracing and quantitative modeling, this study not only reveals the water cycle patterns in the Weigan River basin but also provides quantifiable scientific basis for precise simulation and management of water resources in arid inland river basins.

1 Introduction

Water is a fundamental resource underpinning both human survival and ecosystem stability. However, rapid economic growth, urbanization, and shifting consumption patterns have intensified global water demand, pushing many regions toward a critical water crisis (Liu et al., 2022). In arid inland basins, where renewable water resources are extremely limited, water scarcity has emerged as a key constraint on socio-economic development and ecological security (Azadi et al., 2025). With surface water resources being scarce and periodic unstable, groundwater has become an indispensable component in sustaining agricultural production and human livelihoods (Wu et al., 2025). However, this natural regime has been profoundly altered by intensive groundwater abstraction in the Weigan River Basin (Ma et al., 2024). The specific mechanisms and fluxes of surface water and groundwater interactions under this new anthropogenic dominance remain poorly quantified. The Weigan River basin is a data-deficient region, making the selection of appropriate methods for quantifying surface water and groundwater conversion critically important.

Numerous methods have been developed to investigate surface-groundwater interactions, including hydrological surveys, water balance analysis, numerical simulations, and environmental tracer techniques (Mnati et al., 2023; Zhao et al., 2024). Among these, environmental tracer methods have gained prominence due to their capacity to integrate multi-source data and provide insights into water movement processes (Dinka and Olumana, 2017; Xiao et al., 2023; Yang et al., 2024). In particular, environmental isotopes and hydrochemical indicators have been widely and effectively applied to identify and quantify surface water and groundwater exchanges (Li et al., 2017; Mahlangu et al., 2020). Substantial research demonstrates the applicability of these methods across varied hydrological settings. For example, hydrochemical statistics have been successfully employed to reveal river and groundwater connectivity (Kachadourianmarras et al., 2020; Mladenov et al., 2022; Gao et al., 2025), while stable isotopes of hydrogen and oxygen (²H, ¹⁸O) have consistently proven effective in tracing water cycling processes across diverse basins (Koh et al., 2015; Liao et al., 2020; Tang and Han, 2021). Hydrochemical analyses are highly effective because water-rock interactions during subsurface flow impart distinct compositional signatures to groundwater, differentiating it from surface water (Bajracharya et al., 2020). Statistical analysis of these facies can thus delineate mixing zones and flow paths, even with limited spatial and temporal data (Xu et al., 2017; Bayou et al., 2024). Concurrently, stable isotopes of hydrogen and oxygen serve as conservative tracers of the water molecule itself. In arid regions, the characteristic evaporative enrichment of ¹⁸O and ²H in surface water creates a strong isotopic contrast with often less-evaporated groundwater, making them ideal for identifying (Kim et al., 2023; Tantama et al., 2023; Xiangxiang et al., 2025).

Environmental tracers undergo complex material and energy exchanges during transformation in aquatic systems, with their components susceptible to non-mixing processes such as evaporation and water-rock interactions (Haibo et al., 2016). However, traditional tracer-based quantitative analysis methods (e.g., terminal mixing models) heavily rely on the prior identification of hydrological endpoints within the study area. In arid basin regions characterized by complex hydrogeological conditions, uncertainties in terminal selection can lead to biased estimates of transformation fluxes (Hare et al., 2021). To advance beyond qualitative analysis and achieve robust quantification of MixSIAR model offer significant advantages. MixSIAR model incorporates uncertainty in end-member composition and propagates analytical errors through a probabilistic framework, generating reliable confidence intervals for source contributions (James et al., 2023; Wang et al., 2024). The quantitative results from the MixSIAR model have deepened our understanding of the interaction between surface water and groundwater in complex river basins.

As a major tributary of the Tarim River, the Weigan River plays a pivotal role in maintaining ecological integrity and socio-economic vitality across the oases of southern Xinjiang. However, recurrent downstream water shortages and ecosystem degradation highlight the urgent need for better management of the region’s water resources. This task requires a comprehensive and quantitative understanding of the hydrological connectivity between surface and groundwater systems, which remains poorly quantified. Previous research has largely focused either on surface hydrology or on regional groundwater flow, with limited basin-scale investigations into coupled transformation mechanisms. To address this gap, this study integrates hydrogeochemical analyses and MixSIAR model to elucidate the hydrochemical evolution and quantify surface water and groundwater exchange fluxes within the Weigan River Basin. The findings aim to advance mechanistic understanding of water cycle processes in arid basins and provide a scientific foundation for evidence-based water allocation planning and sustainable oasis management in northwestern China.

2 Materials and methods

2.1 Study area

The Weigan River Basin is located at geographical coordinates 80°13′–84°05′ east longitude and 39°29′–43°05′ north latitude, situated at the northern edge of the Tarim Basin (Figure 1). It extends northward from the Tianshan Mountains, southward to the north bank of the Tarim River, eastward adjacent to Luntai County, and westward bordering Aksu City and Alar City. The terrain primarily consists of mountains and plains (Ren et al., 2024). This watershed exhibits a typical temperate continental arid climate. Based on meteorological observations from 1961 to 2010, the contrast between the wet period (May–September) and the dry period (October–April) is pronounced (Su et al., 2024). During the wet period, the average temperature in the watershed exceeds 15 °C. Abundant heat drives the melting of glaciers and snowpack in the Tianshan Mountains, making this the primary recharge period for rivers. Precipitation during this time accounts for approximately 60%–75% of the annual total, with a monthly average precipitation of about 25 mm. During the dry period, the average temperature in the basin can drop below 0 °C, with scarce precipitation accounting for less than 30% of the annual total. Overall, the basin’s long-term average precipitation is only 70.9 mm/a, while the long-term average potential evaporation reaches 2000–2,500 mm.

Figure 1

Figure 1. Overview of the study area and distribution of sampling points.

The Weigan River originates from the southern slopes of the Tianshan Mountains, formed by the convergence of five major tributaries (Muzhati River, Kamuslang River, Tairweichuk River, Karasu River, and Heizi River) (Su et al., 2023). The multi-year average runoff volumes for each tributary are 14.51 × 10⁸ m³/a, 6.55 × 10⁸ m³/a, 1.09 × 10⁸ m³/a, 2.16 × 10⁸ m³/a, and 4.01 × 10⁸ m³/a respectively (Su et al., 2023). Spatially, each river exhibits a decreasing trend from upstream to downstream due to factors such as evaporation, seepage, and artificial diversion. In terms of timing, runoff from May to September accounts for 70%–80% of the annual total, primarily due to the combined contribution of meltwater from high-altitude glaciers and snowpack, along with summer precipitation. From October to April, runoff constitutes only 20%–30% of the annual volume. Since the Muzhati River and Weigan River traverse both mountainous and plain terrain, this study delineates their upper and lower reaches. The boundary for the Muzhati River is set at Wenbashi Township, while the boundary for the Weigan River is defined as 10 km downstream from the mountain pass exit.

The Baicheng Basin is a large Cenozoic synclinal depression oriented roughly east to west. Within the basin, 200–500 m of loose Quaternary sediments have accumulated, providing ample space for groundwater storage and migration (Figure 2). The basin harbors abundant unconfined groundwater, with multiple-layered confined aquifers found only within the urban area of Baicheng County. The aquifer consists primarily of sand, pebbles, and gravel layers, exhibiting relatively good permeability. The Quaternary deposits in the alluvial plain region reach a thickness of 400–1200 m, increasing gradually from north to south. The lithology of the plain area transitions from gravel layers in the north to sandy soil and clay in the south, with permeability gradually decreasing. The main sources of groundwater recharge include river infiltration, canal infiltration, field infiltration, and underground runoff from mountainous areas to the foothill plains. Groundwater is consumed through evaporation in shallow aquifers and underground runoff toward the Tarim River in the south (Qiao et al., 2013).

Figure 2

Figure 2. Hydrogeological profile of the Weigan River Basin.

2.2 Sample collection and testing

Field sampling of surface water and shallow groundwater in the Weigan River Basin was conducted during March 2024 (dry period) and August 2024 (wet period). This schedule was designed to capture the characteristic hydrological and hydrochemical conditions of each season during periods of relative stability. We avoided the deep winter (e.g., January) due to inaccessible frozen conditions. And the peak melt month (e.g., July) to minimize the influence of transient, sediment-laden flood pulses, thereby ensuring that our samples represent the more stable and representative regimes of the basin. In arid inland basins, the zone of active hydraulic exchange between surface water and groundwater is often conceptually generalized to extend 0–3 km from the river channel (Wang et al., 2025). For the Weigan River Basin, this generic range is strongly supported by local hydrogeological evidence. Data from pumping tests and a network of monitoring wells within the alluvial-pluvial plain reveal a consistently steep hydraulic gradient of 3‰–5‰ within the 0–2 km corridor adjacent to the main river, which is the primary driver of vigorous exchange. Beyond approximately 2.5 km, the gradient flattens markedly to less than 1‰, and aquifer lithology transitions to finer-grained sediments (Hejie, 2014). This significant reduction in driving force confines the dominant, large-volume exchange to within the 0–3 km range, thereby justifying its application as our sampling boundary. Accordingly, sampling sites were strategically distributed along the river corridor, encompassing the major hydrogeological units from the mountainous recharge zones to the downstream plain discharge areas.

During the dry period, a total of 43 water samples were collected, including 18 surface water samples (WG-1 to WG-6 and BC-1 to BC-12), and 25 groundwater samples (WGD-1 to WGD-8 and BCD-1 to BCD-17). During the wet period, 35 water samples were collected: 16 surface water samples (WGX-1 to WGX-4 and BCX-1 to BCX-12), and 19 groundwater samples (WGXD-1 to WGXD-5 and BCXD-1 toBCXD-14) (Figure 1). Two parallel samples were collected for both surface water and groundwater at each sampling period. Surface water samples were obtained directly from natural river reaches, while groundwater samples were taken from local domestic and agricultural irrigation wells. The sampling depth of groundwater (8–35 m) corresponded to the shallow unconfined aquifer most responsive to river–aquifer interactions. To minimize vertical stratification effects, groundwater was extracted after approximately 10 min of pumping, when temperature and electrical conductivity had stabilized. All samples were collected following standardized field protocols to ensure comparability. Prior to sampling, polyethylene bottles were rinsed 4–5 times with the respective water sample. After collection, bottles were immediately sealed with Parafilm, stored in ice-filled insulated containers at 4 °C, and transported in darkness to the laboratory for analysis within 7 days. While groundwater extracted from irrigation wells may reflect anthropogenic recharge associated with agricultural practices, this feature provides valuable insight into human-induced alterations of the natural water cycle. Hence, the hydrochemical signatures obtained represent not only the intrinsic groundwater characteristics but also the cumulative impacts of agricultural and domestic water use under current environmental conditions (Li et al., 2019). This sampling design ensures both scientific rigor and policy relevance, enabling a robust evaluation of the coupled natural and human processes governing water quality and availability in arid basins.

On-site measurements of pH, TDS, and temperature were conducted using a portable DGX-GM862 probe (Germany). Samples for major ions and δD/δ¹⁸O were analyzed at the Public Technology Center of the Northwest Institute of Eco-Environment and Resources, Chinese Academy of Sciences.

The conventional ion chromatograph model used is the Dionex ICS-6000, paired with the Milli-Q IQ7000 ultra-pure water system. Deionized water with a resistivity of 18.2 MΩ/cm is employed. All tubing and connectors in the ion chromatograph system are made of acid- and alkali-resistant PEEK material. Sample pretreatment involves filtering the water sample through a 0.22 μm microporous water system filter membrane before testing. In the field of hydrogeochemical research, it is widely accepted that data quality is excellent when the absolute value of the charge balance error (CBE) falls within ±5%. Data with a CBE between ±5% and ±10% are generally considered reliable and acceptable (Zhang et al., 2024). Among all samples collected in this study, CBE values fell within the acceptable range of ±10%, with 85% of samples exhibiting CBE values below ±5%. This indicates that the hydrochemical data from this study are of good quality and suitable for subsequent statistical analysis, genesis interpretation, and hydrogeochemical modeling.

δD and δ¹⁸O were analyzed using the Isotope Ratio Infrared Spectroscopy (IRIS) system and the Liquid Water Isotope Analyzer (LWIA, 912-0008-1001, LGR Inc, USA) for testing and analysis. The repeatability/precision of δ¹⁸O is 0.1‰, and that of δD is 0.3‰. The hydrogen and oxygen stable isotope content is expressed as the thousandth deviation from V-SMOW (Chen et al., 2024), denoted as Equation 1:

$\delta =\frac{R_{\text{sample}}-R_{\text{standard}}}{R_{\text{standard}}}\times 1000‰(1)$

In the equation, δ represents the degree of change in the composition of δD and δ¹⁸O in the water sample relative to the standard sample, expressed in ‰; R_sample represents the ratio of ²H/¹H or ¹⁸O/¹⁶O in the water; R_standard represents the ratio of ²H/¹H or ¹⁸O/¹⁶O in the standard sample.

2.3 Model principles and software

Statistical analysis of the main water chemical parameters and stable hydrogen and oxygen isotopes of surface water and groundwater in the Weigan River basin was conducted using Microsoft Excel 2013. Origin 2018 was used to plot Piper trilinear diagram, Gibbs diagrams, and diagrams of the proportional relationships between major ions to analyze water chemical types and their causes. This paper employs RStudio to develop a Bayesian mixture model (MixSIAR) and incorporates a Gaussian error estimation formula within the model to estimate the confidence interval for the mixture ratio. This model is based on the Dirichlet distribution and constructs a logical prior distribution within a Bayesian framework. The MixSIAR model incorporates uncertainties related to various isotopic compositions, multiple sources, and discrimination factors (Neil et al., 2025; Boumaiza et al., 2025). The formula is Equations 2–5:

$X_{i,j}={\displaystyle \sum _{k=1}^K}{P}_k\left(S_{i,j}+C_{j,k}\right)+{\epsilon }_{i,j}(2)$

$S_{i,j}\sim N\left({\mu }_{j,k},{\omega }_{j,k}^2\right)(3)$

$C_{j,k}\sim N\left({\sigma }_{j,k},{\tau }_{j,k}^2\right)(4)$

${\epsilon }_{i,j}\sim N\left(0,{\phi }_j^2\right)(5)$

In the equation, X_i,j is the j isotope value of mixture i (i = 1, 2, 3, …, N; j = 1, 2, 3, …, J); P_k is the contribution rate of water source k; S_j,k is the j isotope value of water source k (k = 1,2,3 … ,k), which is normally distributed with mean ${\mathrm{\mu }}_{\mathrm{j},\mathrm{k}}$ and variance ${\omega }_{j,k}^2$. C_jk is the fractionation factor of water source k at isotope j, expressed as the mean ${\mathrm{\sigma }}_{\mathrm{j},\mathrm{k}}$ and standard deviation ${\mathrm{\tau }}_{\mathrm{j},\mathrm{k}}^2$. ${\epsilon }_{i,j}$ is the residual error for additional non-quantified differences for individual components, represented by the mean value 0 and standard deviation ${\mathrm{\phi }}_{\mathrm{j}}^2$.

Under the effects of evaporative fractionation, δ¹⁸O/δD cannot serve as a conservative tracer. This paper introduces a tritium surplus correction method to calibrate measured δ¹⁸O values (Bandara et al., 2024), formulated as Equation 6:

${{\mathrm{\delta }}^{18}\mathrm{O}}_{\text{corrected}}={{\mathrm{\delta }}^{18}\mathrm{O}}_{\text{measured}}+\left[\left({\mathrm{d}}_0-{\mathrm{d}}_{\text{measured}}\right)/\left(8-\mathrm{S}\right)\right](6)$

In the equation, ${\mathrm{\delta }18\mathrm{O}}_{\text{corrected}}$ is the corrected ${\mathrm{\delta }}^{18}\mathrm{O}$ value; ${\mathrm{\delta }18\mathrm{O}}_{\text{measured}}$ is the measured ${\mathrm{\delta }}^{18}\mathrm{O}$ value; ${\mathrm{d}}_0$ is the initial tritium surplus value; ${\mathrm{d}}_{\text{measured}}$ is the measured tritium surplus; S is the slope of the water body evaporation line.

3 Results

3.1 Hydrochemical characteristics and water types

Statistical analysis of concentrations for key hydrochemical parameters in surface water and groundwater within the Weigan River basin (Table 1). During the dry period and wet period, the pH values of surface water ranged from 7.77–8.11 and 7.41–8.00, respectively, with average values of 7.99 and 7.75. The TDS values of surface water ranged from 260.69–3,631.90 mg/L and 137.90–860.34 mg/L, respectively, with average values of 677.85 mg/L and 319.32 mg/L. During dry period and wet period, the concentration order of major anions and cations in surface water remains consistent. The concentration order for major anions is uniformly: HCO₃⁻ > SO₄^2- > Cl⁻. The concentration order for major cations is: Ca²⁺ > Na⁺ > Mg²⁺ > K⁺. The pH values of the groundwater ranged from 7.17–8.87 and from 7.45–8.17, respectively, with average values of 7.84 and 7.79. The TDS values of groundwater ranged from 189.59–3,477.87 mg/L and 200.46–5,742.02 mg/L, respectively, with average values of 992.86 mg/L and 799.69 mg/L. The concentration order of major anions in groundwater is: HCO₃⁻ > Cl⁻ > SO₄^2-. The concentration order of major cations is: Na⁺ > Ca²⁺ > Mg²⁺ > K⁺.

Table 1

Table 1. Characteristics of surface water and groundwater chemical parameters in different periods in the Weigan River Basin.

Based on the concentrations of water chemistry indicators in both surface water and groundwater, it can be determined that both types of water exhibit weakly alkaline properties. Within the Baicheng Basin, surface water TDS decreased during the wet period compared to the dry period. Conversely, surface water in the plain areas shows elevated TDS values during the wet period. Groundwater TDS within the Baicheng Basin follows the same variation pattern as surface water, with locally elevated TDS values observed in the plain areas. The chemical components in surface water are relatively stable, with peak concentrations of Mg²⁺, Na⁺, Cl⁻, and SO₄^2- observed at sampling point WG-5 during the dry period and at sampling point BCX-1 during the wet period. The chemical composition of groundwater exhibits distinct spatio-temporal variations. Spatially, the concentrations of major ions in groundwater across the plain region are significantly higher than those in the Baicheng Basin. Temporally, ion concentrations during the dry period are markedly higher than during the wet period. Groundwater shows relatively high coefficients of variation for Mg²⁺, Na⁺, Cl⁻, and SO₄^2-.

The Piper trilinear diagram (Figure 3) reveals that groundwater exhibits a more complex and diverse hydrochemical type compared to surface water. During the dry period, surface water chemistry types are dominated by HCO₃·Cl-Na·Ca and HCO₃·SO₄-Ca·Mg, accounting for 75% and 79% of the total, respectively. During the wet period, surface water chemistry types are dominated by HCO₃·Cl·SO₄-Na·Ca and HCO₃·SO₄-Ca·Mg, accounting for 75% and 79% of the total, respectively. During both dry and wet periods, groundwater chemistry types were dominated by HCO₃·Cl-Na·Ca and HCO₃·SO₄-Na·Ca, respectively accounting for 73% and 68% of the total.

Figure 3

Figure 3. The Piper trilinear diagram of surface water and groundwater in the Weigan River basin ((a) dry period, (b) wet period).

3.2 Stable hydrogen and oxygen isotope composition

Due to the lack of long-term atmospheric precipitation isotope observation data within the Weigan River basin, this study utilized statistical data from GNIP stations established through collaboration between the International Atomic Energy Agency (IAEA) and the World Meteorological Organization (WMO). Stations proximate to the study area include Hotan, Urumqi, and Alaer. During the dry period (October–April), δD ranges at the three sites were −236.00‰– 39.7‰, −205.00‰– 6.00‰, and −223.70‰– 80.80‰, respectively; δ¹⁸O ranges were −29.81‰– 6.85‰, −27.97‰– 7.63‰, and −26.7‰ to −7.04‰, respectively. During the wet period (May–September), δD ranges were −62.40‰– 45.50‰, −121.50‰ to −8.9‰, and −42.7‰ to −13.5‰; δ¹⁸O ranges were −8.00‰–3.78‰, −14.70‰–1.8‰, and −6.64‰–3.28‰. Comparing data across stations reveals that δD and δ¹⁸O values during the wet period are significantly higher than those during the dry period.

Based on monthly atmospheric precipitation isotope data from three stations, the linear model for atmospheric precipitation (LMWL) equation for the Weigan River basin was obtained via least squares fitting: δD = 7.47δ¹⁸O+ 4.05 (R² = 0.95). The slope of the local precipitation line equation is smaller than that of the global precipitation line (GMWL) (δD = 8δ¹⁸O+ 10). The results indicate that the study area is distant from oceans, characterized by a relatively arid climate, and water vapor is influenced by evaporation processes, which is consistent with actual conditions (Figure 4).

Figure 4

Figure 4. Relationship between δD and δ¹⁸O of monthly atmospheric precipitation in the Weigan River basin.

As shown in Table 1, the δD values of surface water during the dry period and wet period ranged from −77.92‰ to −56.71‰ and −81.31‰ to −59.68‰, respectively, with average values of −68.86‰ and −69.71‰. The δ¹⁸O values ranged from −12.20‰ to −9.33‰ and −12.77‰ to −9.86‰, with average values of −10.82‰ and −11.19‰, respectively. The δD values for groundwater in different periods ranged from −80.93‰ to −56.83‰ and −81.71‰ to −55.43‰, with average values of −69.41‰ and −72.48‰, respectively. The δ¹⁸O values ranged from −12.14‰ to −8.87‰ and −12.70‰ to −9.53‰, with average values of −10.76‰ and −10.68‰, respectively.

The linear relationships between δD and δ¹⁸O for surface water and groundwater in the Weigan River basin are δD = 6.76 δ¹⁸O+ 5.01 (n = 30, R² = 0.90) and δD = 7.34 δ¹⁸O+ 10.32 (n = 40, R² = 0.89), respectively. Both slopes are smaller than that of the LMWL. Most surface water and groundwater data points cluster in the upper-left quadrant of the regional precipitation line (Figure 5). Surface water and groundwater data points overlap significantly.

Figure 5

Figure 5. Relationship between δ¹⁸O and δD of surface water and groundwater in the Weigan River basin.

3.3 Surface water and groundwater conversion ratio based on MixSIAR

This study employed the Bayesian Stable 1sotope Assignment Reaction (MixSIAR) model to quantitatively calculate the transformation ratio between surface water and groundwater. The selection of end members for the MixSIAR model must satisfy core criteria including conservativeness, independence, and sufficient spatio-temporal representativeness. Although δ¹⁸O is a commonly used tracer for studying water mixing processes, its conservativeness is compromised under conditions of intense evaporation. Therefore, this study first applied the d-excess method to correct all δ¹⁸O test data from water samples for evaporation effects, thereby restoring their conservativeness and ensuring compliance with the fundamental assumptions of the mixing model.

Precipitation (P), upstream surface water (SW_up), and upstream groundwater (GW_up) are designated as potential recharge sources. Downstream surface water (SW_down) or downstream groundwater (GW_down) is defined as the mixture requiring analysis. For the downstream sections of the Muzhati River and Weigan River, the water body at the outlet of the respective upstream sections is selected as the source. To enhance the spatiotemporal representativeness of the endpoints, this study utilized long-term data from three adjacent GNIP stations (Hotan, Urumqi, and Alaer). After monthly precipitation-weighted calculations, δ¹⁸O values were −14.89‰ during the dry period and −6.37‰ during the wet period. Surface water and groundwater endpoints represent the average corrected δ¹⁸O values from the upper reaches of each river (Table 2). As shown in the results of Section 3.2, significant spatial variations exist between the corrected precipitation and the upstream surface water and groundwater endpoints. However, we candidly acknowledge that characterizing surface water and groundwater endpoints based on sampling during both dry and wet periods inherently limits temporal representativeness. Furthermore, the number of water samples and the iteration step size also represent potential sources of model uncertainty.

Table 2

Table 2. End-members and mixture data input for the MixSIAR model.

Based on the optimized endpoint data, the MixSIAR model calculated the mean surface water to groundwater conversion ratios and their 95% confidence intervals for each river in the study area (Table 3; Figure 6). During both dry and wet periods, the Heizi River, Karasu River, and Tairweichuk River exhibit surface water recharge to groundwater. During dry period, surface water contributes 60%, 59%, and 65% to groundwater recharge for these rivers, respectively. During wet period, the contribution ratios are 56%, 54%, and 59%, respectively. There is no significant conversion relationship between surface water and groundwater in the Kamuslang River. The conversion relationship between the Muzhati River and Weigan River exhibits distinct spatial differentiation. In the upstream section, surface water replenishes groundwater, while in the downstream section, groundwater replenishes surface water. Surface water recharge contributions in the upper reaches of the Muzhati River were 55% (dry period) and 65% (wet period), while groundwater recharge contributions in the lower reaches were 63% (dry period) and 55% (wet period). For the Weigan River, surface water recharge contributes 70% (dry period) and 57% (wet period) in the upper reaches, while groundwater recharge contributes 53% (dry period) and 56% (wet period) in the lower reaches.

Table 3

Table 3. Contribution proportion of different end-members and 95% confidence interval.

Figure 6

Figure 6. End-members contribution proportion and 95% Confidence interval plot ((a) is dry period, (b) is wet period).

4 Discussion

4.1 Key factors in water chemistry and ion sources

Gibbs diagrams can be used to analyze the distribution characteristics of surface water and groundwater with respect to three controlling factors: rock weathering, precipitation, and evaporative concentration (Hegeu et al., 2023). According to the Gibbs diagram (Figure 7), most surface water samples during both dry and wet periods were influenced by water-rock interactions, with only a few dry period samples from the plain region falling within the evaporation zone. Groundwater throughout the Baicheng Basin was influenced by water-rock interaction during both periods, while groundwater in the plain area was affected by both water-rock interaction and evaporation. Similar ion sources in surface water and groundwater suggest a potential close transformation relationship between them.

Figure 7

Figure 7. Gibbs diagram of surface water and groundwater ((a) and (b) are dry period, (c) and (d) are wet period).

The hydrochemical characteristics of the Weigan River basin are primarily controlled by water-rock interactions. The ratios of γCa²⁺/γNa⁺ and γHCO₃^-/γNa⁺ can be used to infer which rock minerals dissolved to form the current hydrochemical profile (Sun et al., 2021). During dry periods, surface water ions primarily originate from silicate rock dissolution, whereas during wet periods, surface water is influenced by both silicate and carbonate rock dissolution (Figure 8). Groundwater in the Baicheng Basin derives mainly from silicate rock dissolution in both dry and wet periods, while groundwater in the plain area results from the combined action of silicate and evaporite rocks. This phenomenon aligns with the results of Gibbs diagram analysis.

Figure 8

Figure 8. Relationship between Ca²⁺/Na⁺ and HCO₃⁻/Na⁺ in surface water and groundwater.

Further analysis of ion ratio coefficients was conducted on surface water and groundwater within the study area. Surface water sampling points generally cluster near the 1:1 line of γNa⁺/γCl⁻ (Figure 9a). Groundwater samples above the 1:1 line indicate that Na⁺ in the water originates not only from rock salt dissolution but also from the dissolution of silicates such as sodium feldspar, or may be influenced by cation exchange adsorption. Samples below the 1:1 line are primarily affected by human agricultural activities (Selvakumar et al., 2022). Surface water and groundwater samples during the wet period generally cluster near the 1:1 line, indicating that Ca²⁺, Mg²⁺, and HCO₃⁻ in the water primarily originate from the dissolution of carbonate rocks (Figure 9b). SO₄^2- mainly stems from the dissolution of evaporite rocks (gypsum). In contrast, samples from the dry period deviate significantly from the 1:1 line. Surface water and groundwater exhibit identical water-rock interaction characteristics across different periods. Surface water samples are predominantly located above the 1:1 line, indicating that carbonate dissolut0ion is primarily dominated by calcite (Figure 9c). Most groundwater samples are situated above both the 1:1 and 1:2 lines, suggesting that groundwater dissolution is primarily driven by the combined dissolution of calcite and dolomite. Based on the above analysis, groundwater within the region may be influenced by cation exchange adsorption. By examining the relationship between γ(Na⁺+K⁺-Cl^-) and γ((Ca²⁺+Mg²⁺) -(HCO₃⁻ + SO₄^2-)), it is possible to determine whether cation exchange adsorption has occurred in the groundwater body (Zhan et al., 2015). Most groundwater samples from the Baicheng Basin cluster near the y = -x line, indicating cation exchange adsorption occurs within the basin (Figure 9d). However, some samples from the plain area deviate from this line, suggesting they are less affected by cation exchange adsorption (Liu et al., 2020).

Figure 9

Figure 9. Proportional relationship between major ions in surface water and groundwater.

4.2 Quantitative transformation patterns and controlling conditions for surface water and groundwater

Based on the calculation results from the MixSIAR model (Table 3), the conversion relationship between surface water and groundwater can be preliminarily determined. The Heizi River, Karasu River, and Tairweichuk River primarily exhibit surface water replenishing groundwater. The surface water chemistry of these three rivers is predominantly characterized by the HCO₃·SO₄-Ca·Mg type. Groundwater chemistry evolves along the flow path from an HCO₃·Cl-Na·Ca type to an HCO₃·SO₄-Na·Ca type, revealing groundwater recharge from surface water rich in SO₄^2-. Simultaneously, this systemic evolution from Ca²⁺-enriched surface water to Na⁺-enriched groundwater reflects significant cation exchange occurring within the aquifer after surface water infiltration. From a hydrogeological perspective, these rivers flow through deeply incised valleys with riverbed sediments dominated by highly permeable sand, gravel, and cobble. Furthermore, the groundwater table consistently lies below the river water level, establishing a stable hydraulic gradient that drives continuous surface water infiltration. This stable groundwater recharge pattern for mountainous rivers is prevalent in arid inland river basins. For example, in the Qilian Mountains of the upper Heihe River basin, studies similarly indicate that surface runoff is the primary recharge source for shallow groundwater (Li et al., 2018).

There is no significant transformation relationship between surface water and groundwater in the Kamuslang River. The δ¹⁸O values and major ion concentrations of both surface water and groundwater in this river remain relatively stable in both time and space, with highly overlapping ranges. The hydrochemical types of surface water and groundwater show distinct differences. This phenomenon is likely attributed to the presence of relatively impermeable clay layers beneath the riverbed, which impede vertical exchange. Similar decoupling between surface water and groundwater systems has been observed in certain sections of the Hexi Corridor due to fault-induced water barriers or lithological changes (Feng, 2010).

The upper reaches of the Muzhati River and Weigan River exhibit surface water recharge to groundwater, while the lower reaches show groundwater discharge to surface water. The controlling factor in this transformation process is the change in hydrogeological structure along the flow direction. In the upper reaches, the rivers lie atop alluvial fans with steep topography and high permeability, where groundwater lies at significant depths, facilitating recharge through surface water seepage. As the rivers enter the downstream alluvial plains, the terrain flattens, groundwater levels rapidly shallow, and the hydraulic gradient reverses, causing groundwater to discharge into the rivers. This transition pattern is a typical hydrological process in arid inland regions. Similar transition patterns have been observed in the Aksu River basin at the northern margin of the Tarim Basin and the Manas River basin in the Junggar Basin (Yang et al., 2017; Feng et al., 2024).

4.3 Uncertainty analysis of mixing ratios and statistical significance assessment

To quantify the reliability of model results, this study assessed the statistical significance of differences across rivers and periods based on the mean contribution ratios and their 95% confidence intervals (CI) calculated using MixSIAR (Table 3). The core of analyzing the significance of computational results lies in comparing the overlap of their contribution ratio confidence intervals. If the 95% confidence intervals of two sets of estimated values do not overlap, it indicates that their difference is statistically significant (p < 0.05) (Cumming and Finch, 2005).

Statistical results indicate that the spatial patterns of surface water-groundwater conversion relationships within the study area are statistically significant. “Mountainous” rivers (Heizi River, Karasu River, and Tairweichuk River) exhibit surface water replenishing groundwater. Taking the dry period as an example, the mean surface water contribution rates and their 95% confidence intervals for these three rivers were 60% (CI: 48%–72%), 59% (CI: 48%–70%), and 65% (CI: 55%–75%), respectively. The confidence intervals for surface water contribution and groundwater contribution do not overlap, statistically confirming the dominant role of surface water. For “mountain-plain” rivers (Muzhati River and Weigan River), surface water replenished groundwater in mountainous areas, while groundwater replenished surface water in plains. Taking the Muzhati River during the dry period as an example, surface water in the upstream section contributes 65% (CI: 55%–75%) to groundwater, while surface water in the downstream section contributes 40% (CI: 35%–45%) to its own volume. The non-overlapping confidence intervals statistically confirm the reversal of the conversion relationship between surface water and groundwater. The Weigan River exhibits a similar conversion pattern.

In terms of seasonal dynamics, the conversion ratios among rivers in the Wei River basin show minimal variation. The 95% confidence intervals for conversion ratios during dry and wet periods exhibit high overlap across different rivers. This indicates that groundwater and surface water exchange processes within the basin remain relatively stable on a seasonal scale. However, we observe that in mountainous areas, the proportion of surface water replenishment during the dry period is slightly higher than during the wet period. This may be attributed to the rapid rise in groundwater levels during the wet season, which reduces the hydraulic gradient between river water and groundwater, thereby decreasing the recharge rate (Boano et al., 2014). Conversely, in the plains area, the proportion of groundwater contributing to surface water during the wet period increases compared to the dry period. This is likely due to the significant infiltration of agricultural irrigation water during the wet season, leading to an increase in the volume of groundwater converted into surface water (Cao et al., 2013).

5 Conclusion

This study systematically elucidates the interaction mechanisms between surface water and groundwater in the Weigan River basin by integrating hydrochemistry, stable hydrogen and oxygen isotopes, and the MixSIAR model. It provides quantitative analysis and uncertainty assessment of their exchange fluxes. Key findings are as follows.

1. Surface water in the Weigan River basin is primarily influenced by water-rock interactions. In mountainous areas, intense surface water recharge (rich in Ca²⁺ and SO₄^2-) is the dominant factor controlling the chemical composition of shallow groundwater. This promotes cation exchange adsorption within aquifers, leading to groundwater evolution toward the HCO₃·SO₄-Na·Ca type. In the plains, groundwater chemistry is influenced by evaporation-concentration processes and human activities. Its discharge into surface waters also contributes to elevated Na⁺ concentrations in river channels. The spatial evolution of the basin’s hydrochemical characteristics provides reliable hydrochemical evidence for identifying water cycle processes across different geomorphological units.

2. The MixSIAR model results indicate that the direction of water cycle transformation is strictly controlled by topography and hydrogeological structure, forming a spatial transformation pattern of “surface water infiltration in mountainous areas and groundwater discharge in plains.” In mountainous river channels, the average surface water contribution rates during the dry and wet seasons are 63.8% and 56.2%, respectively. In contrast, for the lower reaches of the Muzhati and Weigan Rivers flowing through the foothill plains, groundwater contributes 54.0% and 59.5% on average during the dry and wet seasons, respectively. Seasonally, surface water replenishment in mountainous areas averages 7.6% higher during the dry season than the wet season, while groundwater replenishment in plains areas averages 5.5% higher during the wet season than the dry season.

3. This study restored the conservativeness of δ¹⁸O under intense evaporation conditions through deuterium surplus correction, ensuring the prerequisite for applying the mixing model. Subsequently, the MixSIAR model was employed to quantify conversion ratios and provide uncertainty ranges (95% CI). Confidence interval analysis confirmed statistically significant spatial variations in transformation relationships across different regions. Seasonal variations, however, were insignificant in most areas, indicating that the fundamental patterns of surface-groundwater exchange within the basin remain relatively stable at seasonal scales.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Author contributions

YX: Data curation, Investigation, Writing – original draft, Software, Visualization. CR: Methodology, Conceptualization, Writing – original draft. JZ: Writing – original draft, Visualization, Conceptualization. JH: Formal Analysis, Writing – original draft, Methodology. YZ: Validation, Writing – original draft. WW: Investigation, Writing – original draft. HC: Project administration, Writing – original draft. JY: Conceptualization, Writing – review and editing, Supervision. AL: Resources, Project administration, Writing – review and editing, Funding acquisition.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This work was financially supported by the National Natural Science Foundation of China (Grant No. U2443207; Grant No. 72304245).

Conflict of interest

Author JH was employed by Yunnan Geological Engineering Survey Design and Research Institute Co., Ltd.

The remaining author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

Azadi, Y., Yaghoubi, J., Gholizadeh, H., Gholamrezai, S., and Rahimi-Feyzabad, F. J. (2025). Social barriers to water scarcity adaptation: a grounded theory exploration in arid and semi-arid regions. Agricultural Water Manage 309, 109338.

CrossRef Full Text | Google Scholar

Bajracharya, R., Nakamura, T., Ghimire, S., Shakya, B. M., and Tamrakar, N. K. J. W. (2020). Identifying groundwater and River water interconnections using hydrochemistry, stable isotopes, and statistical methods in Hanumante River, Kathmandu Valley, central Nepal. Water (Basel). 12 (6), 1524. doi:10.3390/w12061524

CrossRef Full Text | Google Scholar

Bandara, U. G. C., Diyabalanage, S., Barth, J. A. C., and Chandrajith, R. (2024). Geochemical and isotope characterization of groundwater and assessment of surface water mixing in the coastal karst aquifer basin in northwestern Sri Lanka. Appl. Geochem. 170, 106098. doi:10.1016/j.apgeochem.2024.106098

CrossRef Full Text | Google Scholar

Bayou, W. T., Mohammed, M., Ayenew, T., Mebrahtu, T. K., and Wohnlich, S. J. (2024). Hydrochemistry and isotope hydrology of groundwater and surface water in the Sor and Gebba watershed, southwestern Ethiopia. Environ. Earth Sci. 83 (10), 316. doi:10.1007/s12665-024-11620-3

CrossRef Full Text | Google Scholar

Boano, F., Harvey, J. W., Marion, A., Packman, A. I., Revelli, R., Ridolfi, L., et al. (2014). Hyporheic flow and transport processes: mechanisms, models, and biogeochemical implications. Rev. Geophys. 52 (4), 603–679. doi:10.1002/2012rg000417

CrossRef Full Text | Google Scholar

Boumaiza, L., Chesnaux, R., Stotler, R. L., Zahi, F., Mayer, B., Leybourne, M. I., et al. (2025). Multiple environmental tracers combined with a constrained Bayesian isotope mixing model to elucidate nitrate and sulfate contamination in a coastal groundwater system. Sci. Total Environ. 959, 178265. doi:10.1016/j.scitotenv.2024.178265

PubMed Abstract | CrossRef Full Text | Google Scholar

Cao, G., Zheng, C., Scanlon, B. R., Liu, J., and Li, W. J. W. R. R. (2013). Use of flow modeling to assess sustainability of groundwater resources in the North China Plain. Water Resour. Res. 49 (1), 159–175. doi:10.1029/2012wr011899

CrossRef Full Text | Google Scholar

Chen, H., Li, J., Zhao, C., Wang, C., Hao, W., Meng, L., et al. (2024). Seasonal and spatial variations of water recharging in the Yangtze River using hydrogen and oxygen stable isotopes. J. Hydrol. X. 639 (000), 10. doi:10.1016/j.jhydrol.2024.131563

CrossRef Full Text | Google Scholar

Cumming, G., and Finch, S. J. A. P. (2005). Inference by eye: confidence intervals and how to read pictures of data. Am. Psychol. 60 (2), 170–180. doi:10.1037/0003-066X.60.2.170

PubMed Abstract | CrossRef Full Text | Google Scholar

Dinka, M. O. (2017). Hydrochemical composition and origin of surface water and groundwater in the Matahara area, Ethiopia. Inland Waters 7(3): 297–304.

CrossRef Full Text | Google Scholar

Feng, Q. J. (2010). “Water resources in the Hexi corridor and its cycle,” in Sustainability in Food and Water. Alliance for Global Sustainability Bookseries. Editors A. Sumi, K. Fukushi, R. Honda, and K. Hassan (Dordrecht: Springer). 18, 299–310. doi:10.1007/978-90-481-9914-3_30

CrossRef Full Text | Google Scholar

Feng, F., Wu, S., Feng, Q., Jin, S., Gou, S., and Huang, Q. J. (2024). Hydrological processes in the Aksu River source region during the intense ablation period based on water chemistry and isotopes. J. Geogr. Sci. 34 (11), 2111–2127. doi:10.1007/s11442-024-2285-0

CrossRef Full Text | Google Scholar

Gao, Z., Huang, T., Chen, J., Tian, H., Tan, M., Niu, Y., et al. (2025). Comprehensive hydrochemical analysis, controlling mechanisms, and water quality assessment of surface and groundwater in a typical intensive agricultural area, Northern China. Water 17 (2), 276. doi:10.3390/w17020276

CrossRef Full Text | Google Scholar

Haibo, C., Jiahua, W., Rong, W., and Baodong, X. J. (2016). Characterizing the interaction of groundwater and surface water in the karst aquifer of Fangshan, Beijing (China). Hydrogeol. J. 25(2): 1–14.

Google Scholar

Hare, D. K., Helton, A. M., Johnson, Z. C., Lane, J. W., and Briggs, M. A. J. N. C. (2021). Continental-scale analysis of shallow and deep groundwater contributions to streams. Nat. Commun. 12 (1), 1450. doi:10.1038/s41467-021-21651-0

PubMed Abstract | CrossRef Full Text | Google Scholar

Hegeu, H., Kshetrimayum, K. S., and Gaichunglu, G. (2023). Identification of surface water-groundwater interaction using hydrochemistry and multivariate statistical technique in the alluvial plain of Nagaland, India. Arab. J. Geosci. 16, 290. doi:10.1007/s12517-023-11358-5

CrossRef Full Text | Google Scholar

Hejie, G. J. (2014). Saline-Alkali soil amelioration and groundwater utilization using vertical drainage method in Weigan River irrigation district of China. Water Resour. Res 3, 160–165.

Google Scholar

James, R., Amasi, A. I., Wynants, M., Nobert, J., Mtei, K. M., Njau, K. J. J. o. S., et al. (2023). Tracing the dominant sources of sediment flowing towards Lake Victoria using geochemical tracers and a Bayesian mixing model. J. Soils Sediments 23 (3), 1568–1580. doi:10.1007/s11368-023-03440-y

CrossRef Full Text | Google Scholar

Kachadourianmarras, A., Alconadamagliano, M. M., Mendoza, E., Silva, R. J. W., and Herrerías-Azcue, F. (2020). Characterization of surface evidence of groundwater flow systems in Continental Mexico. Water (Basel). 12 (2459), 2459. doi:10.3390/w12092459

CrossRef Full Text | Google Scholar

Kim, D. H., Lee, J. Y., Cho, S. Y., Moon, H. S., Jung, Y. Y., Park, Y., et al. (2023). Determining groundwater-surface water interaction at coastal Lagoons using hydrogeochemical tracers. J. Soil Groundwater Environ 28(2): 1–11.

Google Scholar

Koh, E. H., Kaown, D., Yoon, Y. Y., and Lee, K. K. (2015). 3H-3He groundwater ages of the layered aquifer system in the agricultural fields of Jeju volcanic island, South Korea. Egu General Assem. Conf.

Google Scholar

Li, J., Li, F., and Liu, Q. J. C. (2017). PAHs behavior in surface water and groundwater of the Yellow River estuary: evidence from isotopes and hydrochemistry. Chemosphere 178 (JUL), 143–153. doi:10.1016/j.chemosphere.2017.03.052

PubMed Abstract | CrossRef Full Text | Google Scholar

Li, X., Cheng, G., Ge, Y., Li, H., Han, F., Hu, X., et al. (2018). Hydrological cycle in the heihe River Basin and its implication for water resource management in endorheic basins. JGR. Atmos. 123 (17), 890–914. doi:10.1002/2017jd027889

CrossRef Full Text | Google Scholar

Li, Z., QingchunYang, Y. M., HonhyunWang, H. L., and JiannanBian, J. D. M. (2019). Isotopic and geochemical interpretation of groundwater under the influences of anthropogenic activities. J. Hydro 576.

CrossRef Full Text | Google Scholar

Liao, D., Pang, Z., Xiao, W., Hao, Y., Sun, G. J. W., Yang, X., et al. (2020). Constraining the water cycle model of an important karstic catchment in Southeast Tibetan Plateau using isotopic tracers (2H, 18O, 3H, 222Rn). Water (Basel). 12 (12), 3306. doi:10.3390/w12123306

CrossRef Full Text | Google Scholar

Liu, J., Wang, M., Gao, Z., Chen, Q., Li, F. J. A. G., and Li, F. (2020). Hydrochemical characteristics and water quality assessment of groundwater in the Yishu River basin. Acta Geophys. 68 (8), 877–889. doi:10.1007/s11600-020-00440-1

CrossRef Full Text | Google Scholar

Liu, W., Liu, X., Yang, H., Ciais, P., and Wada, Y. J. (2022). Global water scarcity assessment incorporating green water in crop production, Water Resour. Res 58.

CrossRef Full Text | Google Scholar

Ma, Z., Wang, W., Zhang, Z., Ji, D., Wang, J., Zhao, M., et al. (2024). River–groundwater interactions in the arid and semiarid areas of northwestern China. Hydrogeol. J. 32 (1), 37–57. doi:10.1007/s10040-023-02691-w

CrossRef Full Text | Google Scholar

Mahlangu, S., Lorentz, S., Diamond, R., and Dippenaar, M. J. (2020). Surface water-groundwater interaction using tritium and stable water isotopes: a case study of Middelburg. Afr. Earth Sci 171. 103886

CrossRef Full Text | Google Scholar

Mladenov, N., Parsons, D., Kinoshita, A. M., Pinongcos, F., Mueller, M., Garcia, D., et al. (2022). Groundwater-surface water interactions and flux of organic matter and nutrients in an urban, Mediterranean stream. Sci. Total Environ. 811, 152379.

CrossRef Full Text | Google Scholar

Mnati, M., Al-Aboodi, A., and Hassan, A. A. J. (2023). Simulation of interaction between groundwater and surface water in safwan-zubair area, south of Iraq. BJES 23, 50–55. doi:10.33971/bjes.23.1.7

CrossRef Full Text | Google Scholar

Neil, E., Fu, H., and Si, B. (2025). A continuous semi-nonparametric isotope-based mixing model for multimodal water uptake patterns. Ecohydrology 18, e70003. doi:10.1002/eco.70003

CrossRef Full Text | Google Scholar

Qiao, X., Yang, Y., and Yang, D. J. C. (2013). Willingness to pay for ecosystem services and influencing factors in Weigan River Basin: willingness to pay for ecosystem services and influencing factors in Weigan River Basin. Chin. J. Eco-Agric 20: 1254–1261. doi:10.3724/sp.j.1011.2012.01254

CrossRef Full Text | Google Scholar

Ren, C., Zhang, P., Deng, X. Y., Zhang, J., Wang, Y. Y., Wang, S. H., et al. (2024). Unveiling the dynamics and influence of water footprints in arid areas: a case study of Xinjiang, China. Water 16 (8), 1164. doi:10.3390/w16081164

CrossRef Full Text | Google Scholar

Selvakumar, S., Chandrasekar, N., Srinivas, Y., Selvam, S., Kaliraj, S., Magesh, N. S., et al. (2022). Hydrogeochemical processes controlling the groundwater salinity in the coastal aquifers of Southern Tamil Nadu, India. Mar. Pollut. Bull. 174, 113264. doi:10.1016/j.marpolbul.2021.113264

CrossRef Full Text | Google Scholar

Su, J. W., Long, A. H., Chen, F. L., Ren, C., Zhang, P., Zhang, J., et al. (2023). Impact of the construction of water conservation projects on runoff from the Weigan River. Water 15 (13), 2431. doi:10.3390/w15132431

CrossRef Full Text | Google Scholar

Su, J. W., Zhang, P., Deng, X. Y., Ren, C., Zhang, J., Chen, F. L., et al. (2024). Predicting runoff from the Weigan River under climate change. Appl. Sciences-Basel 14 (2), 541. doi:10.3390/app14020541

CrossRef Full Text | Google Scholar

Sun, J., Yan, B., Li, Y., Sun, H., Wang, Y., and Chen, J. (2021). Characterization and cause analysis of shallow groundwater hydrochemistry in the plains of Henan Province, China. Sustainability 13 (22), 12586. doi:10.3390/su132212586

CrossRef Full Text | Google Scholar

Tang, Y., and Han, R. J. W. (2021). Stable carbon and sulfur isotope characteristics of stream water in a typical karst small catchment, Southwest China. Water (Basel). 13 (4), 523. doi:10.3390/w13040523

CrossRef Full Text | Google Scholar

Tantama, E. E., Putra, D., and Marliyani, G. J. (2023). “The origin of shallow groundwater in Randublatung basin and its surroundings based on stable isotopes (18O and 2H),“ in 4th International Conference on Earth Science, Mineral and Energy.

Google Scholar

Wang, F., Liu, L., Xu, W., Li, Y., Ruan, Q., Cao, W. J. E., et al. (2024). Multiple stable isotopic approaches for tracing nitrate contamination sources: implications for nitrogen management in complex watersheds. Ecotoxicol. Environ. Saf. 269 (000), 9. doi:10.1016/j.ecoenv.2023.115822

PubMed Abstract | CrossRef Full Text | Google Scholar

Wang, L., Sun, Y., Yang, C., and Dong, Y. (2025). Spatial variations in groundwater-surface water interaction patterns in the shule River Basin, Northwest China: insights from stable isotopes and hydrochemistry.

Google Scholar

Wu, X., Feng, X., and Fu, B. J. G. (2025). Climate aridification and intensified human interference undermined water storage in the Lower Yellow River region. Geogr Sustain 6, 100303, doi:10.1016/j.geosus.2025.100303

CrossRef Full Text | Google Scholar

Xiangxiang, C., Lu, L., Yutong, L., Xueqing, Z., Suhua, M., and Shan, L. J. (2025). Chemical characteristics and controlling factors of groundwater in an intensive surface water-groundwater conversion zone: a case study of the Zhengzhou section of the lower Yellow River. Front. Environ. Sci. 13 (13), 1580232. doi:10.3389/fenvs.2025.1580232

CrossRef Full Text | Google Scholar

Xiao, Y., Zhang, J., Long, A., Xu, S., Guo, T., Gu, X., et al. (2023). Hydrochemical characteristics and Formation mechanism of Quaternary groundwater in Baoshan Basin, Western Yunnan, China. Water (Basel). 15 (15), 2736. doi:10.3390/w15152736

CrossRef Full Text | Google Scholar

Xu, W., Su, X., Dai, Z., Yang, F., Zhu, P., and Huang, Y. J. H. J. (2017). Multi-tracer investigation of river and groundwater interactions: a case study in Nalenggele River basin, northwest China. Hydrogeol. J. 25, 2015–2029. doi:10.1007/s10040-017-1606-0

CrossRef Full Text | Google Scholar

Yang, G., He, X. L., Li, X. L., Long, A. H., Xue, L. Q., and Engineering, B. (2017). Transformation of surface water and groundwater and water balance in the agricultural irrigation area of the Manas River Basin. China 10 (4), 107–118.

Google Scholar

Yang, H., Wei, J., and Shi, K. J. W. (2024). Hydrochemical and isotopic characteristics and the spatiotemporal differences of surface water and groundwater in the Qaidam Basin. China 16 (1), 26.

Google Scholar

Zhan, Y., Teng, Y., and Zuo, R. (2015). Chemical characteristics and geochemical evolution of groundwater in the North China Plain. Egu General Assem. Conf.

Google Scholar

Zhang, Z., Liu, F., Chen, S., Jiang, W., Zhang, H., Gao, Z., et al. (2024). Hydrogeochemical and isotopic insights into groundwater evolution in the agricultural area of the Luanhe Plain. Environ. Earth Sci. 83 (20), 1–13. doi:10.1007/s12665-024-11857-y

CrossRef Full Text | Google Scholar

Zhao, L., Su, C., Liu, W., and Xie, X. J. (2024). Understanding surface water-groundwater conversion relationship and associated Hydrogeochemistry evolution in the upper reaches of Luan River Basin, north China. J. Earth Sci. 35 (3), 1010–1023. doi:10.1007/s12583-022-1629-x

CrossRef Full Text | Google Scholar