Preamble

J Arid Land (2025) 17(11): 1604–1622 Science Press Springer-Verlag Hydrochemical characteristics and transformation relationships between different water bodies in the Qixing Lake region of the Hobq Desert, China XI Cheng , YAN Min , ZUO Hejun , LIU Ruimin 1 College of Desert Control Science and Engineering, Inner Mongolia Agricultural University, Hohhot 010018, China; State Key Laboratory of Water Engineering Ecology and Environment in Arid Area, Inner Mongolia Agricultural University, Hohhot 010018, China

Abstract

Desert lakes are an important link in the water cycle and an important reservoir of water resources in arid and semi-arid areas, playing an important role in maintaining the stability of the regional natural environment. However, studies on the hydrochemical evolution and transformation relationships between desert lake groups and potential water sources are limited. Taking the Qixing Lake, the only lake group within the Hobq Desert in China, as the area of interest, this study collected samples of precipitation water, Yellow River water, lake water, and groundwater at different burial depths in the Qixing Lake region from July 2023 to October 2024. The hydrochemistry of different water bodies was analyzed using a combination of Piper diagrams, Gibbs diagrams, ratio of ions, and MixSIAR mixing models to reveal the transformational relationships of lake water with precipitation, groundwater, and Yellow River water. Results showed that both groundwater and surface water in the study area are weakly-to-strongly alkaline, with HCO as the dominant anion and Na , and K as the main cations.

The hydrochemical type of groundwater and some lakes was dominated by HCO , whereas that of other lakes was dominated by Cl and HCO . The hydrochemistry of groundwater and Yellow River water in the Qixing Lake region was controlled mainly by a combination of evaporite saline and silicate rock mineral dissolution. The local meteoric water line (LMWL) of the study area proved that regional water bodies are strongly affected by evaporative fractionation. The MixSIAR model revealed that shallow groundwater is the main recharge source of the lake group in the Qixing Lake region, accounting for 59.0%–64.2% of the total. The findings can provide references for the identification of water sources in desert lakes and the development and utilization of water resources in desert lake regions.

Keywords

hydrochemical type; cation exchange; stable isotope; MixSIAR model; desert lake sources; Qixing Lake Citation:

XI Cheng, YAN Min, ZUO Hejun, LIU Ruimin. 2025. Hydrochemical characteristics and transformation relationships between different water bodies in the Qixing Lake region of the Hobq Desert, China. Journal of Arid Land, 17(11):

1 Introduction

Water is an important resource for the survival of human beings and all living things (Luo et al., 2020). Water scarcity and the resulting environmental crisis is a major problem globally, especially in arid areas (Zhang et al., 2020). China is one of the countries with the largest extent of desertification in the world, the contradiction between the supply and the demand of water © Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences, Science Press and Springer-Verlag GmbH Germany, part of Springer Nature 2025

resources in China is becoming increasingly obvious, and the problems of water shortages and water pollution are becoming ever more serious and complex (Zhang et al., 2021; Wang et al., 2023b). In arid and semi-arid areas, the transformation between precipitation, surface water, and groundwater is an important process in the hydrological cycle, prerequisite for watershed-scale management and protection of water resources (Qin et al., 2021). As an important link in the water cycle and an important reservoir of water resources, lakes have a high degree of ecological vulnerability and high ecological value, and the sustainable development and utilization of lake water resources can be of great importance to regional socioeconomic development (Duan et al., 2021).

China has a large number, wide distribution, and full range of lake types. The desert lake represents a unique lake type that is important for maintaining and enhancing the natural environment in desert areas (Chen et al., 2004; Zhang et al., 2019). Arid areas with dry climate, high evapotranspiration, and scarce effective precipitation recharge can form lakes with neither surface nor subsurface runoff connected to the ocean, and no direct hydraulic connection to other catchment areas. Thus, an independent water circulation system is fundamental to the long-term existence of lakes in arid areas. Many studies have investigated the formation and evolution of desert lakes, reporting rich results regarding the dynamic evolution of the water level and area of such lakes (Zhou et al., 2019), the salinization processes (Tweed et al., 2011), the impact of climate (Woolway et al., 2020), and the underlying mechanism of the water cycle (Chen et al., 2020), thereby providing a scientific basis for the protection and management of lakes in arid areas. However, owing to lack of recent monitoring data, research on the water environment and evolution of lakes in arid areas has centered mainly on single large- and medium-sized lakes.

Most lakes in arid areas are closed lakes, representing the lowest drainage datum in the basin plain. Therefore, such lakes are recharged by precipitation, surface runoff, and subsurface runoff, and if there is no runoff out of the lakes, evaporation is the only mode of lake discharge (Fu et al., 2021). The Badain Jaran Desert in China has a large number of lakes and is an area of active desert lake research. Under the extremely arid environment, strong evaporation has considerable influence on the water level and the hydrochemistry characteristics of the lake group, and the contribution of groundwater recharge to the lakes is an important reason for the existence of the characteristic sand dune/lake landscape (Chen et al., 2012). Moreover, the lake group also affects the regional groundwater circulation pattern and the groundwater recharge intensity (Wu et al., 2014). Differences in the recharge process among the lakes of the Badain Jaran Desert lake group lead to marked variation in the hydrochemistry characteristics, owing to variation in the hydrodynamic conditions, aquifer characteristics, recharge and drainage paths, and other factors (Luo et al., 2017; Cao et al., 2022). Consequently, consensus is lacking regarding qualitative explanation of the sources of lake water.

Lakes interact with groundwater and other potential recharge sources to varying degrees (Urrutia et al., 2019). Therefore, the hydrochemical components of natural water bodies are the products of long-term interactions between water and its surroundings, and they also provide a record of the formation, transport, and transformation pathways of water bodies to a certain extent. Stable isotopes of hydrogen (δD) and oxygen (δ O) are affected by evaporative fractionation, condensation, and mixing, and water sources of different origins are identifiable by their characteristic isotopes and relatively stable contents. Therefore, the joint application of hydrochemistry and stable isotope tracer techniques is an effective approach to improving understanding of the complex hydrological processes of desert lakes (Zhang et al., 2019).

However, there are fewer studies on the hydrochemical evolution and transformation relationships of desert lake groups based on combined hydrochemical and isotopic approaches, and the understanding of the differences between different desert lakes remains unclear.

The Hobq Desert of China is a desert subject to the combined action of the monsoon and the Yellow River, and the Qixing Lake is the only lake group distributed within the Hobq Desert (Xi et al., 2024). As a sensitive unit of regional environmental change and interaction between human and natural environments, the Qixing Lake is located in an arid area with little precipitation,

strong evaporation, and no natural surface runoff recharge. Owing to insufficient water environment monitoring capacity and inadequate systematic and comprehensive information, the current status of the water environment and the water cycle characteristics of the Qixing Lake remain unclear. Additionally, the Yellow River, as the northern boundary of the Hobq Desert, is only approximately 10 km from the Qixing Lake. The elevation of the Qixing Lake is higher than that of the Yellow River. In this context, does the Yellow River actually have lateral recharge to the Qixing Lake? Is there a hydraulic connection between them? Under the influence of an intensely evaporative wind and sandy environment, what are the water sources that can be maintained for a long period of time in the Qixing Lake? What is the proportion? Currently, there is a lack of studies on hydrochemical types and transformation relationships in different water bodies of the Qixing Lake region. Therefore, in this context, it is important to consider whether the Yellow River actually recharges the Qixing Lake and what the rate of contribution is, and the relationships between the Qixing Lake and the Yellow River, as well as between the Qixing Lake and other water bodies.

Therefore, through hydrochemistry and stable isotope analysis methods, this study explored the hydrochemical characteristics of the Qixing Lake, and elucidated the complex transformational relationships between different water bodies. The findings can help reveal the hydrological, hydrophysical, and hydrochemical importance of desert lakes represented by the Qixing Lake, and provide a resource to support both the sustainable development and utilization and the ecological protection and restoration of desert lakes and wetlands. 2 Materials and methods

2.1 Study area

The Hobq Desert (Fig. 1a [FIGURE:1]) is located in Ordos City, Inner Mongolia Autonomous Region, China, and is an active sand sea in the semi-arid areas of northern China. It extends for approximately 370 km from west to east and approximately 15–50 km from north to south, presenting as a long narrow band with an area of 17,300 km . The sand dunes of the Hobq Desert almost entirely cover the Yellow River terrace on the northern edge of the Ordos Plateau, and the terrain rises in a step-like fashion from north to south. The Inner Mongolia section of the Yellow River Basin is an important sand source area for regional sandstorms and dust storms, and it is the main source of sediment in the middle and lower reaches of the Yellow River. The Yellow River is also the northern boundary of the Hobq Desert and represents an important source of water supporting the existence of oases at the edge of the desert and providing for ecological construction.

The Hobq Desert Qixing Lake Ecotourism Zone, also known as Hobq National Desert Park, is a national 4A-level desert ecotourism resort located in Duguitala Town, Hanggin Banner, Ordos City. There are seven lakes in the park: Dadaotu Lake (DDT), Zhangjinao Lake (ZJN), Zhahandaotu Lake (ZHDT), Wuritu Lake (WRT), Yueliangshen Lake (YLS), Naren Lake (NR), and Yikeershen Lake (YKES). These lakes together form the Qixing Lake, which is the only lake group in the Hobq Desert. This study took the Qixing Lake region (40°37′40″–40°48′00″N, 108°15′00″–108°36′00″E) as the study area, which is in the transition zone between arid and semi-arid areas. It has a typical temperate continental monsoon climate with annual average precipitation of 250 mm and annual average evaporation of 2100–2955 mm.

According to the lithological characteristics of the aquifer and the conditions of groundwater storage and hydrodynamic characteristics, the main sources of groundwater recharge in the study area are infiltration recharge from atmospheric precipitation, back-seepage recharge from agricultural irrigation water (Yellow River), and lateral runoff recharge from the desert dive in the core area of the southern part of the Hobq Desert; and the main modes of discharge are evaporation from the groundwater and discharge from the underground runoff. Hanggin Banner belongs to the Ordos Plateau, and the rock formation is nearly horizontal and slightly north-dipping, with a regional dip angle of about 1° (Fig. 1b). Due to the limitation of the

Overview of the topographic and geomorphologic map of the Hobq Desert (a) and hydrogeologic zoning and groundwater discharge direction in Hanggin Banner (b), as well as the information of three boreholes near the Qixing Lake and the hydrogeologic situation (c). HK 06, HK 07, and ZK 256 are borehole designation; Q2 and Q4 are Quaternary deposits of the Pleistocene (Q2) and Holocene (Q4). Note that the boundary of Hanggin Banner is based on the standard map ( bsfw/bzdt/), and the boundary of the standard map has not been modified. monoclinic water storage structure in the study area, the groundwater basically flows from south to north, and is finally discharged to the Yellow River. On the way of runoff, groundwater is discharged to the surface and gullies through submerged aquifers, or it is discharged through strong evaporation by pooling into lakes in low-lying places, which is the main discharge pathway of groundwater in this region. Among them, the Yellow River water and groundwater have a close mutual recharge and discharge relationship. The Qixing Lake is located in the alluvial plain on the south bank of the Yellow River, and the groundwater types are loose rock pore diving and pressurized water (Ge et al., 2016). The Yellow River alluvial plain is gray-yellow fine sand and silt of the Quaternary Holocene, and the water separator is located at 50–80 m (Fig. 1c), which is sandy clay. The mineral composition of the soil in the Qixing Lake region is mainly quartz, followed by feldspar, calcite, muscovite and chlorite, with low content of dolomite and tremolite (Meng, 2023).

2.2.1 Sampling methods

Categorical sampling was conducted from July 2023 to October 2024 in major water bodies in the study area. A total of 36 groups of water samples were collected (Fig. 2 [FIGURE:2]): 3 groups of Yellow River samples (one group for every 15 km upstream and sample numbers of YR1–3), 7 groups of precipitation samples (sample numbers of PP1–7 and sampling dates of 11 July 2023, 3 August 2023, 25 August 2023, 2 October 2023, 24 April 2024, 4 August 2024, and 17 October 2024), 5 groups of groundwater samples (3 groups in Daotu village and sample numbers of GW-DT1–3 for groundwater depths of 17, 80, and 300 m, respectively; 2 groups in Dalatu village and sample numbers of GW-DLT1–2 for groundwater depths of 16 and 120 m, respectively), and 21 groups of Qixing Lake samples (3 groups taken from each of the 7 lakes with sample numbers of DDT1–3, WRT1–3, ZJN1–3, ZHDT1–3, YLS1–3, NR1–3, and YKES1–3). Combined with the

hydrogeological situation, this study regarded the groundwater located above the water barrier at depths of 16 and 17 m as shallow groundwater, and the groundwater below the water barrier at depths of 80, 120, and 300 m as deep groundwater. All samples were collected by rinsing polyethylene sampling bottles three times with raw water, allowing them to be filled with water samples. The mouths of the filled bottles were wrapped with Parafilm M sealing film, and these bottles were sealed, preserved, and returned to the laboratory for testing.

Overview of the Qixing Lake region and distribution of the sampling points for different water bodies.

DDT, Dadaotu Lake; WRT, Wuritu Lake; ZJN, Zhangjinao Lake; ZHDT, Zhahandaotu Lake; YLS, Yueliangshen Lake; NR, Naren Lake; YKES, Yikeershen Lake. YR1–3, 3 water sample groups from Yellow River (YR); GW-DT1–3, 3 groups of groundwater samples in Daotu village; GW-DLT1–2, 2 groups of groundwater samples in Dalatu village; DDT1–3, 3 groups of lake samples from DDT; WRT1–3, 3 groups of lake samples from WRT; ZJN1–3, 3 groups of lake samples from ZJN; ZHDT1–3, 3 groups of lake samples from ZHDT; YLS1–3, 3 groups of lake samples from YLS; NR1–3, 3 groups of lake samples from NR; YKES1–3, 3 groups of lake samples from YKES; PP1–7, 7 water sample groups of precipitation (PP).

Measurement methods A portable conductivity water parameter meter (EASY Probe30; Water Land, Beijing, China) was used to determine in-situ water pH and total dissolved solids (TDS). Concentrations of Cl , and Mg were determined in the laboratory using an ion chromatograph (ICS-600; Thermo Fisher, Waltham, USA), and the absolute values of the relative errors of the total anions and cations were less than 5.0%. The concentration of HCO was determined by hydrochloric acid titration, and three replicate tests were performed for each sample (average error<5.0%), following which the average value was taken. The δD and δ O were determined using a high-precision liquid water isotope analyzer (L2130-i/L2140-i; Picarro, Santa Clara, USA), with accuracies of ±0.40‰ and ±0.10‰, respectively. The δD and δ O of water samples were expressed conventionally as δ sample (‰), which is the thousandth deviation of V-SMOW, and can be expressed as follows: sample sample standard standard ×1000‰, where sample is the ratio of H (or O) in the water samples; and standard is the ratio of H (or O) in the standard samples.

Analysis methods Gibbs (1970) analyzed rainwater, river water, lake water, and seawater samples from all over the

world by plotting scatter plots between TDS and Na ) as well as between TDS and ), and classified the hydrochemical fractions of natural water bodies into three controlling endmembers: a rock weathering main control endmember, an atmospheric precipitation main control endmember, and an evaporation concentration main control endmember. Gibbs plots help qualitatively determine the influencing factors of hydrochemical composition, and they have been used widely in the study of water body formation mechanisms.

Here, Gibbs diagrams were used to determine the main controlling factors of hydrochemistry in the studied lakes and other water bodies. The Piper diagrams support analysis and comparison of the derived water quality results for improved understanding of the hydrogeochemical processes (Piper, 1944).

The chlor-alkali indices (CAI-I and CAI-II) are often used to characterize the direction and strength of alternate cation adsorption (Li et al., 2013; Kumar et al., 2018). When the values of both indices are positive, Na and K in groundwater exchange with Ca and Mg in the soil, and reverse cation exchange adsorption occurs (Na +CaX→Ca X, where X is the anion).

When the values of both indices are negative, Ca and Mg in groundwater exchange with Ca and Mg in the soil, and cation exchange adsorption occurs (Na +CaX). The magnitude of the absolute value of each index represents the strength of the alternation. The chlor-alkali indices were calculated using the formulas as follows: )]/Cl )]/(SO The [(Na ] versus [(Ca )–(HCO )] relationship is often used to indicate the occurrence of cation exchange adsorption. When the two are linear and the slope is close to –1, cation exchange adsorption is likely to occur. The [(Na ] denotes the amount of increase or decrease in Na brought about by dissolution of substances other than rock salt, while the [(Ca )–(HCO )] denotes the amount of increase or decrease in Ca brought about by dissolution of substances other than gypsum, calcite, and dolomite (Farid et al., 2013).

To analyze the sources of the major ions in the surface water and groundwater and to clarify the process of hydrochemical evolution, this study utilized the relationships among Ca , and to determine the source of each ion during the water cycle, adopted the relationship between Na and Cl to reveal the sources of rock salt dissolution in the groundwater system, employed the relationship between Ca and HCO to determine the sources of Ca , and used the relationship between Ca and HCO to analyze the participation of precipitated carbonic and sulfuric acids in the dissolution of carbonate rocks in the surface water and groundwater (Han et al., 2009; Xing et al., 2013).

The deuterium excess ( -excess) parameter is also known as the deuterium surplus. The size of -excess is defined as the intercept when the slope of the local meteoric water line (LMWL) is 8.00, characterizing the degree of disequilibrium in the evaporation process and reflecting the degree to which it differs from the global isotopic fractionation of atmospheric precipitation (Valdivielso et al., 2020). In this study, Bayesian mixture model identification was used to calculate the contribution of multiple sources to the mixture. The MixSIAR Bayesian mixture model can quantitatively determine the contribution of potential water sources. This model combines the latest research results of the Bayesian mixture model MixSIR and SIAR, and is used mainly to provide more accurate estimates of the rates of source contributions by choosing the type of source data, fixed/random effects, error terms, and prior distributions. The MixSIAR model incorporates uncertainties associated with a variety of isotopic compositions, multiple sources, and discrimination factors. Before accepting the final output of the model, it is necessary to determine whether the model had converged. In this study, the Gelman–Rubin and Geweke diagnostic tests were used to determine whether the model is close to convergence (Stock et al., 2018; Wang et al., 2019; Egbi et al., 2020), which can be expressed as follows:

where is the -isotope value of mixture =1, 2, …, =1, 2, 3, …, is the contribution of source =1, 2, …, ), crudely calculated by the model; is the -isotope value of source normally distributed with mean value ( ) and variance ( is the fractionation factor of source in the -isotope, expressed as mean value ( ) and standard deviation ( ); and is the residual error of the additional non-quantified differences in individual components, expressed as mean value (0) and standard deviation As the lakes in the Qixing Lake region are all closed lakes without surface runoff recharge, recharge is mainly based on precipitation and groundwater, and discharge is based on evaporation, according to the principle of water balance, the water balance estimation of the lake group can be expressed as: d , where is the volume of lake water (m is the time; d is the change in lake water volume with time change d is the precipitation recharge (m is the groundwater recharge (m ); and is the evaporation (m ). This study used the method of Yan et al. (2024) to extract the lake area and other parameters of lakes, and calculated the amount of lake changes during the pre-sampling period (2018–2020). The regional average precipitation and evapotranspiration data were obtained from the China Meteorological Data Center was 312 and 1232 mm, respectively. 3 Results and discussion 3.1 Hydrochemistry characterization and ion source analysis of different water bodies

3.1.1 Characterization of hydrochemistry

Details of the hydrochemical and hydrogen–oxygen isotope compositions of the surface water and groundwater in the Qixing Lake region are listed in Table 1 [TABLE:1]. The pH and TDS contents of different water bodies were highly variable, and their ionic compositions were characterized by marked differences. The dominant anion in the Yellow River water, Qixing Lake water, and regional groundwater was HCO . The dominant cation of the Yellow River water, regional groundwater, and the lake water of DDT, WRT, ZJN, ZHDT, and YKES was Na ; the dominant cation of YLS water was Ca ; and the dominant cation of NR water was K . The average pH value of the Yellow River water samples was 7.57, which is in the neutral to weakly alkaline range. Among the lake group, ZJN was the most alkaline with a mean pH of 10.63; YKES had the lowest pH of 8.65; and pH of other lakes varied from 8.89 to 10.40 on average. ZHDT had the highest TDS with a mean value of 8550.0 mg/L, followed by WRT, DDT, and ZJN, with a variation range of 1640.9–3081.8 mg/L. These lakes were more salty, probably because of the larger surface of the lakes, which are strongly evaporated but less recharged. The TDS was lower in the YLS, NR, and YKES, with mean value around 400.0 mg/L, which was closer to the range of groundwater, indicating that the groundwater recharge capacity may be higher. The groundwater was low mineralization except for GW-DLT2, and the pH of groundwater did not vary much, ranging from 8.28 to 8.99. The TDS in the lake water of ZHDT, WRT, DDT, and ZJN were substantially higher than that in the groundwater, indicating that the water–rock interaction of the surface water is stronger than that of groundwater, and that the content of easily soluble salts is higher (Wang et al., 2023a).

The Piper diagrams showed that the distribution of water chemical ions in different surface water bodies and the groundwater is more concentrated in the diamond-shaped area (Fig. 3 [FIGURE:3]), indicating that the sources of ions are more consistent and relatively closely linked. The

Characterization of hydrochemical ions in different water bodies YKES1 YKES2 YKES3 Note: TDS, total dissolved solids; YR1–3, 3 water sample groups from Yellow River; GW-DT1–3, 3 groups of groundwater samples in Daotu village; GW-DLT1–2, 2 groups of groundwater samples in Dalatu village; DDT1–3, 3 groups of lake samples from Dadaotu Lake (DDT); WRT1–3, 3 groups of lake samples from Wuritu Lake (WRT); ZJN1–3, 3 groups of lake samples from Zhangjinao Lake (ZJN); ZHDT1–3, 3 groups of lake samples from Zhahandaotu Lake (ZHDT); YLS1–3, 3 groups of lake samples from Yueliangshen Lake (YLS); NR1–3, 3 groups of lake samples from Naren Lake (NR); YKES1–3, 3 groups of lake samples from Yikeershen Lake (YKES). hydrochemical type of water from the Yellow River, GW-DT1, GW-DT3, GW-DLT1–2, DDT, and ZJN was HCO ; the hydrochemical type of water from the GW-DT2 was ; the water of WRT was Cl type; the water of ZHDT was type; and the water of YLS, NR, and YKES was HCO type.

Generally, in arid inland basins, the final destination of groundwater is a salt lake, which represents the terminus of the regional groundwater flow system (Samani et al., 2021). The average TDS values in ZHDT and WRT were as high as 8550.0 and 3081.8 mg/L, respectively, consistent with the modeling of this groundwater flow system. Additionally, the mineralization of lake water was higher than that of river water and groundwater, which might be due to the strong evaporation of lake water that leads to the accumulation of salt in the lake. Through the

Piper diagram for water samples from different water bodies in the study area precipitation of Ca and Mg minerals as well as the cation replacement adsorption, Na became the dominant ion in the lake water. Furthermore, there was a hydraulic connection between the groundwater and the lakes, meaning that the groundwater of HCO type was transformed into the lake water of higher salinity Cl type through evaporation and concentration, i.e., groundwater carried salts into the lake, and the lake water then evaporated, leaving behind an aggregation of salts. This confirmed that groundwater is one of the sources of recharge for the studied lakes (Hao et al., 2019).

Identification of factors controlling hydrochemistry in different water bodies The main controlling factor for the hydrochemistry of Yellow River, most of the groundwater, and YKES, NR, and YLS was rock weathering (Fig. 4 [FIGURE:4]). For ZHDT, the main controlling factor of its hydrochemistry was evaporation. Data points for WRT, DDT, and ZJN were all in the transition zone between the evaporation-concentration endmember and the rock-weathering endmember, suggesting that these lakes are distributed in arid areas with strong evapotranspiration. This is consistent with the characteristics of these lakes on the Inner Mongolian Plateau, which are dominated by groundwater recharge and subject to the arid and semi-arid climate (Ma et al., 2022). Overall, rock weathering dominated in the hydrochemical formation process of all the water bodies in the Qixing Lake region, while the effect of atmospheric precipitation was weak.

The CAI-I and CAI-II values of Yellow River water, lake water, and all groundwater samples were negative (Fig. 5 [FIGURE:5]), characterized by the occurrence of cation exchange adsorption, indicating that positive cation exchange is the main type of cation exchange in the water bodies of the Qixing Lake region. Exchange of Na and K in the rocks with Ca and Mg in the water bodies

Gibbs diagrams showing the hydrochemical formation process of different water bodies in the study area. (a), relationship between Na ) and total dissolved solids (TDS); (b), relationship between Cl ) and TDS.

Distribution of the chlor-alkali indices (CAI-I and CAI-II) (a) and relationship between [(Na and [(Ca )–(HCO )] (b) for different water bodies in the study area occurred, resulting in increases in Na and K and reductions in Ca and Mg in the lake water and groundwater. The [(Na ] versus [(Ca )–(HCO )] relationship of the water in ZHDT deviated markedly from the theoretical value of –1. The relationship between the change in Na due to dissolution of rock salt and changes in Ca and Mg due to dissolution of gypsum, calcite, and dolomite was not obvious. This is mainly due to the fact that the Na and Mg contents of lake water are also affected by dissolution of Na and K feldspar, as well as by the equilibrium effect of carbonate dissolution (Wang et al., 2022).

Ion source analysis of different water bodies Data points for lake water, Yellow River water, and groundwater were closer to the silicate rock control endmember, suggesting that most water samples in the study area are affected by silicate

(Fig. 6a [FIGURE:6]). Data points for ZHDT and WRT were slightly closer to the evaporation control endmember. The Na value of water in most lakes (DDT, ZHDT, WRT, YLS, and YKES) and the Yellow River was close to 1 (Fig. 6b), indicating that these water bodies are affected little by anthropogenic disturbances. Evidently, the data points for most of the groundwater and the water of DDT, WRT, and ZJN were distributed below the (Ca )/HCO =1:1 ratio line (Fig. 6c), indicating that the cation exchange in the water–rock action makes Ca and Mg decrease. The lack of a data point plotted above the (Ca )/(HCO )=1:1 ratio line in Figure 6d indicated that SO in the water samples from the study area may not originate from the dissolution of gypsum in the carbonate rock strata. Data points for most groundwater samples were distributed below the (Ca )/(HCO )=1:1 ratio line, indicating that the concentration value of Ca is less than that of HCO Generally, Na and K originate from atmospheric deposition, dissolution of silicate minerals, and dissolution of evaporate salt minerals, and the dissolution of rock salt releases Na and Cl the water column, which theoretically means that Na and Cl should have the same concentrations (Tay et al., 2014; Li et al., 2022). In the plot of the relationship between Ca and HCO except for the Yellow River water, there were no data points of the water samples distributed above the 1:1 ratio line, suggesting that Ca and Mg do not originate from carbonate dissolution, and that very few other ions, such as carbonic acid, sulfuric acid, and nitric acid, are involved in the dissolution of the carbonate rock (Wang et al., 2021a). Meanwhile, the concentration value of was less than the concentration value of HCO , suggesting that the SO might originate from the dissolution of sulfide minerals in the loess layers around the Yellow River, given the actual geochemical background in the region (Qu et al., 2024).

Relationship between the major ion ratios in the different water bodies in the study area. (a), relationship between Ca and HCO ; (b), relationship between Na and Cl ; (c), relationship between Ca and HCO ; (d) relationship between HCO and Ca

3.2 Hydrogen and oxygen isotope characterization of different water bodies and indicative importance regarding the water cycle 3.2.1 Characteristics and indicative significance of precipitation equivalent curves for different water bodies The LMWL for the study area was determined as δD=6.84δ O+1.38‰ (Fig. 7a [FIGURE:7]), and the global meteoric water line (GMWL) was defined as δD=8.00δ O+10.00‰, as proposed by Craig (1961). It can be seen that the slope of the LMWL in the Qixing Lake region was 6.84, smaller than the slope of the GMWL (8.00). The slope of the LMWL studied by Pei et al. (2023) in the Hetao area of the Yellow River Basin was 5.94; although the Qixing Lake is within their study area, the study sites in the Hetao area contain several mountainous arid areas, leading to smaller slope. The Yinchuan area located in the upper reaches of the Qixing Lake region had slightly better precipitation conditions, with a LMWL slope of 7.21 (Yang et al., 2018). These suggested that atmospheric precipitation in the Qixing Lake region has undergone strong evapotranspiration, which is in line with the local natural environment of drought and low rainfall (Chen et al., 2022).

Comparison between the global meteoric water line (GMWL) and local meteoric water line (LMWL) in the Qixing Lake region and other adjacent areas (a) and relationship between deuterium excess ( -excess) and TDS in lake water (b), as well as the variations in stable isotopes of hydrogen (δD) and oxygen (δ O) among different water bodies (c and d). Bars are standard errors. -excess of the water samples in the Qixing Lake was in the range from –50.18‰ to –14.59‰ (mean of –27.03‰) (Fig. 7b). The -excess of groundwater was in the range of 0.92‰–3.70‰ (mean of 2.01‰), and the -excess of Yellow River water was in the range of 1.94‰–3.32‰ (mean of 2.42‰) (Table 2 [TABLE:2]). Owing to the shallow burial depth of groundwater and the close hydraulic connection with the Yellow River water, during the runoff process, water–rock interaction continuously enriched the δ O values in both the groundwater and Yellow River water, while the δD values remained largely unchanged, resulting in some of the groundwater

Isotope characterization in different water bodies YKES1 YKES2 YKES3 Note: δD and δ O are stable isotopes of hydrogen and oxygen, respectively. -excess, deuterium excess. samples having a excess less than 10.00. Negative excess in the Qixing Lake region revealed the effects of secondary evapotranspiration on isotopic composition (Juan et al., 2020; Xia and Winnick, 2021). The TDS of the lake water in the Qixing Lake region increased with the decrease of excess, showing negative correlation ( =0.24). The smallest excess was found in WRT, and the lake with the largest TDS value was ZHDT, probably because a certain agricultural and livestock operation exists near ZHDT. If the sampling site of ZHDT is removed, the negative correlation between the excess and the TDS of lake water in the study area will be stronger ( =0.92). In the water circle process, hydrogen and oxygen isotopes of different water bodies are mixed and exchanged. When the water body encounters strong evaporation, the kinetic isotope fractionation effect occurs, and the -excess value decreases and even reaches a negative value, while the salts in the water gradually rise (Tantawi et al., 1998; Subyani, 2004). The -excess calculated in this study suggested that strong evapotranspiration in arid areas is an important mechanism leading to salinization of desert lakes.

3.2.2 Water cycle indicated by δD and δ

O in different water bodies The relationship between δ O and δD is a reliable indicator for water body source analysis (Vystavna et al., 2021; Zhu et al., 2021). The hydrogen and oxygen isotope composition of precipitation varies little over a certain range, and the LMWL in the same area can be represented by an equation (Oza et al., 2020; Wolf et al., 2020). The relationship between δD and δ O of the lake water in the Qixing Lake region can be defined as δD=5.18δ O–22.42‰ ( =0.98), and the slopes and the intercepts of δD and δ O fitted line for the surface water were smaller than those of the LMWL, indicating that the surface water is affected by evaporative fractionation after receiving precipitation recharge. The relationship between δD and δ O of the Yellow River water and the groundwater can be defined as δD=7.45δ O–2.95‰ ( =0.92) and δD=7.76δ O–0.22‰ =0.98), respectively, indicating that evaporation of surface water in the Qixing Lake region is the most intense. The smaller slope of the LMWL in the Qixing Lake region coincides with the local water vapor conditions of drought and low rainfall. Owing to the high latitude of the study area, the precipitation process is affected by secondary evaporation, which produces notable fractionation of hydrogen and oxygen isotopes, resulting in more enriched water vapor δ O than δD and producing a low slope of the LMWL. The average range of δD of the lake water in the Qixing Lake region was –37.83‰–12.74‰ and the average range of δ O was –2.90‰–7.47‰ (Fig. 7c and d). The δD and δ O were high in the lake water, with the highest values found in WRT, which deviated from the data of the other water bodies and indicated that WRT has the highest evaporation capacity. In summary, hydrogen and oxygen isotopes were more enriched in lake water, which was related to the higher evaporation intensity of the water due to the open environment of the lakes.

The isotopic composition of shallow groundwater can reflect the overall proportion of convective and advective components of precipitation, and the relationship between groundwater isotopes and precipitation type might also have broader relevance (Chang et al., 2020). Among the groundwater samples, except for GW-DT1, the distribution of data points was relatively concentrated with that of the Yellow River samples (Fig. 7a), indicating that the hydrogen and oxygen isotope composition of these groundwater samples are under the control of the Yellow River, and that there is a strong hydraulic connection between the groundwater system and the Yellow River. 3.3 Transformation relationships among different water bodies and composition of water sources recharging the Qixing Lake In comparison with hydrochemistry methods, isotopic tracers are generally affected less by water evolutionary processes; therefore, isotopic data allow for better discrimination of endmember elements and clearer interpretation of mixing processes (Mahindawansha et al., 2020; Wang et al., 2021b; Wu et al., 2022). Based on the principle of mass conservation of hydrogen and oxygen isotopes and using the Bayesian mixing model, Figure 8 [FIGURE:8] shows the composition of water sources recharging the Qixing Lake. Overall, the lake water recharge was dominated by groundwater, with the total share ranging from 61.0%–69.2% and averaging 66.8%, and ZHDT was the lake that Groundwater recharge to the lake water was dominated by shallow groundwater recharge. As shown in Figure 8b, the proportion of 16–17 m shallow groundwater recharge to the lake water was 38.7%, which accounted for more than half of the proportion of groundwater recharge to the lake water, while the deep groundwater recharge below the regional diaphragm only accounted for 6.3%–9.8%. These results indicated that the main source of lake water is groundwater, and it is mainly shallow groundwater. This is consistent with the findings of Wang et al. (2017), who concluded that surface and soil water in the Ordos region is mainly recharged by groundwater, although the specific proportion of the water source contribution has not been quantified.

The proportion of precipitation recharge to the lake water as a whole ranged from 18.2% to 27.7%, with an average of 21.2%. Among the seven lakes, the recharge of precipitation to the

YLS, YKES, and NR were slightly larger compared to the other four lakes, which may be related to the fact that the areas of the three lakes are smaller, and the effect of precipitation on the water volume of these lakes is more obvious. Studies of Pei et al. (2023) in the Hetao area of the Yellow River Basin showed that atmospheric precipitation in the region is the main source of surface water recharge, up to more than 80.0%, but the Hetao area is a large area, and the surface water composition of the study object is mainly irrigation channels, the Yellow River trunk and tributaries, and seasonal rivers, with fewer lake sampling sites. Therefore, the results of the aforementioned research that precipitation can occupy a certain proportion of the water sources in the Hetao area can be used as a category for reference of the water sources of the lake group of the Qixing Lake. The range of contribution of precipitation recharge to surface water studied by other scholars in this and nearby semi-arid areas can be up to 19.0%–38.0%, which can also justify the Qixing Lake being recharged by precipitation (Han et al., 2021; Zhao et al., 2024).

The Yellow River, as a recharge source, recharged the Qixing Lake with an average percentage of 12.0%. Among the proportions of groundwater recharged by the Yellow River at each depth, the shallow groundwater of 16–17 m was recharged most by the Yellow River (Fig. 8c). We speculated that the process of Yellow River recharge to the Qixing Lake is that the infiltration of Yellow River water first recharges to the regional shallow groundwater, and then the groundwater recharges to the Qixing Lake again. Subsequently, the cycling process of different water bodies in the region, the sources of groundwater at different depths, and the relationships between groundwater at different depths could be the focus of future research.

Composition of water sources recharging the Qixing Lake. (a), recharge proportions of precipitation, Yellow River, and groundwater to each lake of the Qixing Lake group; (b), recharge proportions of precipitation, Yellow River, and groundwater to the Qixing Lake; (c), recharge proportions of Yellow River to groundwater.

Data in Figure 8b are mean±SD. Bars are standard errors.

In addition, based on the principle of water balance, we calculated the amount of precipitation and groundwater recharging the lake. As shown in Table 3 [TABLE:3], the Qixing Lake received an average of 20.6% (±5.7%) recharge from precipitation, which is consistent with the results of precipitation recharge calculated using the Bayesian mixing model. Among the seven lakes, YKES received the least amount of precipitation recharge, which may be related to the deeper water depths of the lake with outcropping springs at the bottom. Groundwater recharge to the Qixing Lake accounted for 75.2%–90.8%, which can be considered reasonable when combined with the previous modeling analysis, after adding the proportion of Yellow River recharge to groundwater followed by the recharge to the Qixing Lake (12.0%) to the proportion of groundwater recharge (66.8%).

Water balance parameters for each lake of the Qixing Lake group Parameter Average area (km 8.67×10 3.83×10 1.88×10 3.42×10 Amount of change in area (m 2.71×10 1.12×10 2.96×10 7.28×10 1.60×10 6.84×10 3.25×10 Average depth of the lake (m) Volume change (m 9.39×10 3.35×10 6.78×10 7.57×10 4.42×10 3.20×10 2.18×10 3.13×10 1.12×10 2.26×10 2.52×10 1.47×10 1.07×10 7.27×10 4.49×10 3.91×10 2.02×10 2.69×10 1.19×10 5.82×10 1.06×10 ) (%) 1.78×10 1.55×10 8.02×10 1.06×10 4.71×10 2.31×10 4.21×10 1.36×10 1.27×10 6.23×10 1.05×10 3.67×10 2.79×10 1.04×10 ) (%) Note: d is the change in lake water volume with time change d is the precipitation recharge; is the groundwater recharge; is the evaporation.

4 Conclusions

Through systematic and complete sample collection of different water bodies in the Qixing Lake region, we comparatively analyzed the hydrochemical evolution characteristics of the Qixing Lake with the method of hydrochemical characterization, and quantified the transformation relationships among different water bodies by MixSIAR hybrid modeling. In the Qixing Lake region, the groundwater and surface water were weakly alkaline to strongly alkaline, the water–rock interaction of the lakes was strong, and the soluble salt content was high. The hydrochemical characteristics of the lake water, groundwater, and Yellow River water were controlled mainly by the joint dissolution of evaporite saline and silicate rock minerals. Sufficient dissolution of mineral components in lake water and groundwater occurred, and shallow groundwater was the main recharge source for the Qixing Lake. In the future, enhanced analysis of the hydrochemical composition of the submerged sediments of the lakes and the sediments of the groundwater layer will help to further understand the specific trajectories and pathways of ion exchange between desert lakes and sediments.

Conflict of interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements This work was supported by the Inner Mongolia Autonomous Region "Unveiling the List of Commanders" Project (2024JBGS0019), the Inner Mongolia Autonomous Region Graduate Student Research Innovation Project (KC2024036B), the Innovative Team on Desertification Control and Sandy Area Resource Conservation and Utilization (BR241301), and the Desert Sand Ecological Protection and Management Technology Innovation Team (NMGIRT2408).

Author contributions Conceptualization: XI Cheng, YAN Min, ZUO Hejun; Data curation: XI Cheng, YAN Min; Methodology: XI Cheng; Formal analysis: YAN Min, LIU Ruimin; Writing - original draft preparation: XI Cheng; Writing - review and editing: YAN Min, ZUO Hejun; Funding acquisition: ZUO Hejun; Investigation: XI Cheng, LIU Ruimin; Resources: YAN Min, ZUO Hejun; Supervision: ZUO Hejun. All authors approved the manuscript.

References

Cao L, Nie Z L, Shen J M, et al. 2022. Stable isotopes reveal the lake shrinkage and groundwater recharge to lakes in the Chang H K, Gonçalves R D, Aggarwal P K, et al. 2020. Groundwater isotope ratios reflect convective and stratiform (paleo) Chen J, Qian H, Gao Y Y, et al. 2020. Insights into hydrological and hydrochemical processes in response to water Chen J S, Li L, Wang J Y, et al. 2004. Groundwater maintains dune landscape. Nature, 432(7016): 459–460.

Chen J S, Sun X X, Gu W Z, et al. 2012. Isotopic and hydrochemical data to restrict the origin of the groundwater in the Badain Jaran Desert, Northern China. Geochemistry International, 50: 455–465.

Chen P, Wang S, Song S, et al. 2022. Ecological restoration intensifies evapotranspiration in the Kubuqi Desert. Ecological Craig H. 1961. Isotopic variations in meteoric waters. Science, 133(3465): 1702–1703.

Duan W L, Zou S, Chen Y N, et al. 2020. Sustainable water management for cross-border resources: The Balkhash Lake Basin Egbi C D, Anornu G K, Ganyaglo S Y, et al. 2020. Nitrate contamination of groundwater in the Lower Volta River Basin of Ghana: sources and related human health risks. Ecotoxicology and Environmental Safety, 191: 110227, doi:

Farid I, Trabelsi R, Zouari K, et al. 2013. Hydrogeochemical processes affecting groundwater in an irrigated land in Central Tunisia. Environmental Earth Sciences, 68: 1215–1231.

Fu C S, Wu H W, Zhu Z C, et al. 2021. Exploring the potential factors on the striking water level variation of the two largest Ge J, Chen J S, Ge L, et al. 2016. Isotopic and hydrochemical evidence of groundwater recharge in the Hopq Desert, NW China.

Journal of Radioanalytical and Nuclear Chemistry, 310: 761–775.

Gibbs R J. 1970. Mechanisms controlling world water chemistry. Science, 170(3962): 1088–1090.

Han D M, Liang X, Jin M G, et al. 2009. Hydrogeochemical indicators of groundwater flow systems in the Yangwu River alluvial fan, Xinzhou Basin, Shanxi, China. Environmental Management, 44: 243–255.

Han Y F, Zhai Y Z, Guo M S, et al. 2021. Hydrochemical and isotopic characterization of the impact of water diversion on water in drainage channels, groundwater, and Lake Ulansuhai in China. Water, 13(21): 3033, doi: 10.3390/w13213033.

Hao S, Li F D, Li Y H, et al. 2019. Stable isotope evidence for identifying the recharge mechanisms of precipitation, surface water, and groundwater in the Ebinur Lake basin. Science of the Total Environment, 657: 1041–1050.

Juan G, Li Z X, Qi F, et al. 2020. Environmental effect and spatiotemporal pattern of stable isotopes in precipitation on the transition zone between the Tibetan Plateau and arid region. Science of the Total Environment, 749: 141559, doi:

Kumar S, Venkatesh A S, Singh R, et al. 2018. Geochemical signatures and isotopic systematics constraining dynamics of fluoride contamination in groundwater across Jamui district, Indo-Gangetic alluvial plains, India. Chemosphere, 205:

Li P Y, Wu J H, Qian H. 2013. Assessment of groundwater quality for irrigation purposes and identification of hydrogeochemical evolution mechanisms in Pengyang County, China. Environmental Earth Sciences, 69: 2211–2225.

Li X Q, Han G L, Liu M, et al. 2022. Potassium and its isotope behaviour during chemical weathering in a tropical catchment affected by evaporite dissolution. Geochimica et Cosmochimica Acta, 316: 105–121.

Luo P P, Sun Y T, Wang S T, et al. 2020. Historical assessment and future sustainability challenges of Egyptian water resources Luo X, Jiao J J, Wang X S, et al. 2017. Groundwater discharge and hydrologic partition of the lakes in desert environment:

Insights from stable H and radium isotopes. Journal of Hydrology, 546: 189–203.

Ma F Y, Chen J Q, Chen J S, et al. 2022. Hydrogeochemical and isotopic evidences of unique groundwater recharge patterns in

the Mongolian Plateau. Hydrological Processes, 36(5): e14554, doi: 10.1002/hyp.14554.

Mahindawansha A, Külls C, Kraft P, et al. 2020. Investigating unproductive water losses from irrigated agricultural crops in the humid tropics through analyses of stable isotopes of water. Hydrology and Earth System Sciences, 24(7): 3627–3642.

Meng Q S. 2023. Spatial distribution of particle size and geochemical characteristies of surface sediments in the hobg desert.

MSc Thesis. Hohhot: Inner Mongolia Agricuiture University. (in Chinese) Oza H, Ganguly A, Padhya V, et al. 2020. Hydrometeorological processes and evaporation from falling rain in Indian sub-continent: insights from stable isotopes and meteorological parameters. Journal of Hydrology, 591: 125601, doi:

Pei S S, Duan L M, Miao P, et al. 2023. Water chemical isotope characteristics and water transformation relationship in Mongolian Section of the Yellow River Basin. Environmental Science, 44(9): 4863–4873. (in Chinese) Piper A M. 1944. A graphic procedure in the geochemical interpretation of water analyses. Eos, Transactions American Geophysical Union, 25(6): 914–928.

Qin W J, Han D M, Song X F, et al. 2021. Environmental isotopes (δ Rn) and hydrochemical evidence for understanding rainfall-surface water-groundwater transformations in a polluted karst area. Journal of Hydrology, 592: 125748, Qu S, Zhao Y Z, Zhang K Y, et al. 2024. Multi-isotopes (δD, δ Owater, Sr, δ S and δ Osulfate) as indicators for groundwater salinization genesis and evolution of a large agricultural drainage lake basin in Inner Mongolia, Northwest Samani S, Moghaddam H K, Zareian M J. 2021. Evaluating time series integrated groundwater sustainability: a case study in Salt Lake catchment, Iran. Environmental Earth Sciences, 80(17): 603, doi: 10.1007/s12665-021-09888-w.

Stock B C, Jackson A L, Ward E J, et al. 2018. Analyzing mixing systems using a new generation of Bayesian tracer mixing models. PeerJ, 6: e5096, doi: 10.7717/peerj.5096.

Subyani A M. 2004. Use of chloride-mass balance and environmental isotopes for evaluation of groundwater recharge in the alluvial aquifer, Wadi Tharad, western Saudi Arabia. Environmental Geology, 46: 741–749.

Tantawi M A, El-Sayed E, Awad M A. 1998. Hydrochemical and stable isotope study of groundwater in the Saint Catherine-Wadi Feiran area, south Sinai, Egypt. Journal of African Earth Sciences, 26(2): 277–284.

Tay C K, Kortatsi B K, Hayford E, et al. 2014. Origin of major dissolved ions in groundwater within the Lower Pra Basin using groundwater geochemistry, source-rock deduction and stable isotopes of H and O. Environmental Earth Sciences, 71:

Tweed S, Leblanc M, Cartwright I, et al. 2011. Arid zone groundwater recharge and salinisation processes; an example from the Lake Eyre Basin, Australia. Journal of Hydrology, 408(3–4): 257–275.

Urrutia J, Herrera C, Custodio E, et al. 2019. Groundwater recharge and hydrodynamics of complex volcanic aquifers with a shallow saline lake: Laguna Tuyajto, Andean Cordillera of northern Chile. Science of the Total Environment, 697: 134116, Valdivielso S, Vázquez-Suñé E, Custodio E. 2020. Origin and variability of oxygen and hydrogen isotopic composition of Vystavna Y, Harjung A, Monteiro L R, et al. 2021. Stable isotopes in global lakes integrate catchment and climatic controls on evaporation. Nature Communications, 12(1): 7224, doi: 10.1038/s41467-021-27569-x.

Wang H, Ni J, Song Q C, et al. 2021a. Analysis of coastal groundwater hydrochemistry evolution based on groundwater flow Wang J, Lu N, Fu B J. 2019. Inter-comparison of stable isotope mixing models for determining plant water source partitioning.

Science of the Total Environment, 666: 685–693. Wang T, Chen J S, Ge J, et al. 2017. Isotopic evidence of allogenic groundwater recharge in the Northern Ordos Basin. Journal of Radioanalytical and Nuclear Chemistry, 314: 1595–1606.

Wang W H, Chen Y N, Wang W R, et al. 2023a. Water quality and interaction between groundwater and surface water impacted Wang Y J, Tang Y, Xu Y, et al. 2021b. Isotopic dynamics of precipitation and its regional and local drivers in a plateau inland Wang Z Y, Su Q, Wang S, et al. 2022. Coupling hydrochemistry and stable isotopes (δ H and δ O) to identify the major factors affecting the hydrochemical evolution of groundwater in the Western Yellow Sea Coast, China. Applied Geochemistry, 138:

Wang Z Y, Xu D Y, Peng D L, et al. 2023b. Future climate change would intensify the water resources supply-demand pressure

Wolf A, Roberts W H, Ersek V, et al. 2020. Rainwater isotopes in central Vietnam controlled by two oceanic moisture sources and rainout effects. Scientific Reports, 10(1): 16482, doi: 10.1038/s41598-020-73508-z.

Woolway R I, Kraemer B M, Lenters J D, et al. 2020. Global lake responses to climate change. Nature Reviews Earth & Environment, 1(8): 388–403.

Wu H W, Song F, Li J, et al. 2022. Surface water isoscapes (δ O and δ H) reveal dual effects of damming and drought on the Wu Y, Wang N A, Zhao L Q, et al. 2014. Hydrochemical characteristics and recharge sources of Lake Nuoertu in the Badain Jaran Desert. Chinese Science Bulletin, 59: 886–895.

Xi C, Zuo H J, Yan M, et al. 2024. Grain size characteristics of different types of surface sediments around Qixing Lake in Kubuqi Desert. Frontiers in Environmental Science, 12: 1409260, doi: 10.3389/fenvs.2024.1409260.

Xia Z Y, Winnick M J. 2021. The competing effects of terrestrial evapotranspiration and raindrop re-evaporation on the deuterium excess continental precipitation.

Earth Planetary Science Letters, 117120, Xing L N, Guo H M, Zhan Y H. 2013. Groundwater hydrochemical characteristics and processes along flow paths in the North China Plain. Journal of Asian Earth Sciences, 70–71: 250–264.

Yan Y, Zuo H J, Yan M, et al. 2024. Lake number and area in Kubuqi Desert from 1986 to 2020. Journal of Desert Research, 44(2): 151–161. (in Chinese) Yang Q C, Mu H K, Wang H, et al. 2018. Quantitative evaluation of groundwater recharge and evaporation intensity with stable oxygen and hydrogen isotopes in a semi-arid region, Northwest China. Hydrological Processes, 32(9): 1130–1136.

Zhang D X, Sial M S, Ahmad N, et al. 2020. Water scarcity and sustainability in an emerging economy: a management perspective for future. Sustainability, 13(1): 144, doi: 10.3390/su13010144.

Zhang M M, Wang S, Gao G Y, et al. 2019. Exploring responses of lake area to river regulation and implications for lake restoration in arid regions. Ecological Engineering, 128: 18–26.

Zhang Y, Liu W X, Cai Y, et al. 2021. Decoupling analysis of water use and economic development in arid region of China–based on quantity and quality of water use. Science of the Total Environment, 761: 143275, doi:

Zhao W J, Qian H, Xu P P, et al. 2024. Tracing groundwater-surface water sources and transformation processes in the Ba River Basin through isotopes water chemistry.

Applied Geochemistry, 106199, Zhou Y, Dong J W, Xiao X M, et al. 2019. Continuous monitoring of lake dynamics on the Mongolian Plateau using all available Landsat imagery and Google Earth Engine. Science of the Total Environment, 689: 366–380.

Zhu G F, Yong L L, Zhang Z X, et al. 2021. Effects of plastic mulch on soil water migration in arid oasis farmland: Evidence of