Optimization of Freshwater–Saline Water Resource Mixing Irrigation Under Multiple Constraints

Abstract

The unique anticline geological structure in the central region of Yingjisha County results in significant spatial variations in groundwater quality. The study shows that the recoverable groundwater reserves account for 13.5% of the natural groundwater supply, and the development potential is considerable. Therefore, this study conducts an in-depth analysis of the spatial distribution characteristics of multiple water sources, integrates agricultural cropping patterns, and delineates irrigation districts accordingly. A water quality-based optimized allocation model for water resources is established. After optimization, the total irrigation water demand is reduced from 3685.8 million m³ to 3229.9 million m³, with total groundwater extraction controlled at 694.0 million m³. The total water shortage rate is 12%, and the decline in groundwater levels has been effectively controlled. Additionally, 116.4 million m³ of saline water is utilized, achieving an 83% utilization rate, which accounts for 16.8% of total groundwater extraction. Consequently, the utilization rate of freshwater decreases from 127% to 64%, while the overall water supply reliability reaches 87.6%. The sequence of water supply and consumption in the model remains consistent with the existing supply structure, demonstrating the rationality of the model parameter settings. This study proposes an optimal freshwater–saline water allocation model, which mixes saline water with reservoir water for dilution and subsequent agricultural irrigation. The approach aims to exploit the potential of saline groundwater and enhance the utilization efficiency of groundwater systems, thereby providing an innovative solution to alleviate water supply-demand conflicts in arid regions.

1. Introduction

Agriculture, as a major water user, accounts for 70% of global water consumption, with irrigation water accounting for over 90% of agricultural water use [1]. China, a populous and agricultural country, has a disconnect in the distribution of water and land resources, with agricultural production highly dependent on irrigation [2,3]. As reported in the 2023 China Water Resources Bulletin, agricultural activities account for approximately 62% of total water withdrawals, with farmland irrigation constituting nearly 61% of agricultural water usage. Due to differences in regional precipitation distribution and long-term overexploitation of groundwater and inappropriate irrigation practices [4], there is a common phenomenon of groundwater levels being either too deep or too shallow in irrigation areas, leading to serious ecological degradation, soil salinization, and other severe ecological problems that significantly hinder the sustainable development of irrigation areas [5,6,7,8]. Yingjisha County suffers from a shortage of freshwater resources, and shallow saline groundwater has significant potential for utilization, which can alleviate the pressure on agricultural water use if used reasonably.

Currently, many studies propose water resources optimization allocation models from the perspective of water distribution or water quality zoning. Wang et al. developed a multi-objective water resources allocation model incorporating both quantity and quality constraints, aiming to achieve a balance between efficient water utilization and quality preservation [9,10]; Zhao et al. proposed the concept of hierarchical water supply and established a rational water allocation model. While existing studies have considered supplying water of different quality levels to users in a prioritized sequence, they have failed to account for the impacts of saline groundwater irrigation on crop growth and secondary soil salinization in agricultural applications. The utilization of high-salinity water resources has emerged as a new research direction in optimal water allocation studies [11,12,13,14]. Therefore, this study incorporates ecological impact factors of saline groundwater by establishing supply priority for groundwater of varying quality, user priority for water allocation, and ecological constraints, to achieve optimized freshwater–saline water allocation.

In recent years, more and more research has been exploring technologies for water quality improvement and comprehensive utilization of water resources in response to the utilization of high salinity water. For example, Zhang Li and others can effectively reduce the salinity in high salinity water through physical and chemical methods, making it suitable for agricultural irrigation [15]. In conclusion, research on the optimization of water resources allocation not only focuses on the rational distribution of water quantity in different regions and qualities but also considers the mutual constraint between water quantity and water quality when facing increasingly serious water shortage and water pollution issues.

The study progress of water resources optimization configuration models has multiple solutions, reflecting the diversity of optimization constraints and objectives. Wu Ze Ning et al. constructed a unified optimization configuration model system framework based on ecological economics for water quality and quantity [16]. Yan Denghua et al. proposed the rational allocation and management of water resources oriented towards ecology based on the analysis of the basin’s “natural-artificial” binary water cycle and the evolution of associated processes and the mechanism of wetland ecological evolution [17]. Yang Gaiqiang et al. established an uncertainty multi-objective optimization model by synthesizing previous research on agricultural water optimization for multiple crops and water sources throughout all growth stages [18]. Xie Xinmin et al. established a “dual-control” water resource allocation model based on the need for ecological civilization construction and the protection of groundwater resources, with the total groundwater and groundwater level as constraints [19]. Fang Guohua et al. established a regional water quantity and quality joint configuration multi-objective model by setting constraints on pollutant loading capacity in water functional areas and water quality requirements for water supply [20]. Su Xiaoling controlled the groundwater levels in canal-well irrigation areas, determined the allocation of surface water and groundwater reasonably, and constructed a model system for temporal and spatial optimization of water resources coupled with numerical simulation of groundwater, with constraints on the water diversion capacity at the canal head, and the monthly allocation results of groundwater and surface water [21]. Li Shuoyang studied the dynamic coupling feedback model of optimized irrigation water allocation and groundwater simulation, exploring the temporal and spatial balance of water resources optimization constrained by total water consumption and reasonable groundwater depth, to achieve water resources management with dual “real” constraints of groundwater level and total water consumption [22]. However, these optimization studies have relatively fewer research on the effects of different irrigation water qualities on soil water salinity and crop growth, neglecting the effective supply of crops through capillary action of shallow groundwater, and the constraints of irrigation conservation measures on changes in water demand [23].

In addition, the water resources management strategies in different regions show significant differences, taking into consideration their geographical characteristics and actual needs [24,25,26,27]. In the optimal allocation of water resources, the amount of groundwater that can be mined more in line with the actual situation is also very important. Nitika, M. et al. [28] estimated the future groundwater level of the basin through the MODFLOW model. Moghaddam, H.K. et al. [29] studied the relationship between exploitable groundwater and groundwater table fluctuation station under the pressure of agricultural groundwater extraction. The study area’s anticline structure has formed a spatial pattern with significant differences in the distribution of groundwater resources between the north and south. Exploring optimized irrigation strategies not only provides a new perspective for the rational allocation of saline groundwater but also effectively enhances the utilization efficiency of water resources in the study area.

Therefore, this article deeply analyzes the distribution characteristics of groundwater space in the background of fold structures, clarifies the relationship between water resources quantity and quality distribution in the hydrogeological background, calculates the exploitable groundwater resources and the available water volume in the entire region. To facilitate water allocation, this study classifies water resources based on total dissolved solids (TDS) concentration using 1 g/L increments. Specifically, (1) groundwater with TDS < 3 g/L is categorized as low-salinity irrigation water (freshwater), including selected groundwater and reservoir water; (2) groundwater exceeding 3 g/L TDS is classified as high-salinity irrigation water (saline water). Based on geological conditions, exploitable water volume, agricultural planting structure, crop salt tolerance, crop growth water level, land salinization, and other ecological factors, irrigation areas are reasonably divided, and a mathematical model with multiple constraints on water quality, water level, and water quantity is established to optimize water resources allocation. The utilization of saline groundwater can enhance the efficiency of groundwater irrigation systems while reducing both groundwater exploitation intensity and total extraction volume. Through optimal allocation of freshwater–saline water resources, the proposed approach employs reservoir water to dilute saline water for irrigation purposes. This strategy not only alleviates agricultural water supply-demand conflicts in Yingjisha County but also provides innovative solutions for regional ecological conservation, demonstrating significant environmental protection benefits.

2. Materials and Methods

2.1. Study Area

The study area is located on the northern margin of the Kunlun Mountains, in the western part of the Tarim Basin. The overall topography slopes from southwest to northeast, characterized by an arid climate with low precipitation and severe water scarcity. From 2018 to 2021, the actual groundwater extraction in the agricultural areas of Yingjisha County exceeded the permissible extraction limit, leading to a continuous decline in groundwater levels. In some areas, groundwater table depth changed by up to 4 m over four years, significantly altering the spatial distribution of groundwater quantity and quality. The location, topography, and land use of the study area are illustrated in Figure 1.

The central region features the Karak Mountain, a low hilly terrain formed by an anticline structure. The mountain extends in a northwest direction (280–290°) with a width of 3–6 km from north to south and a length of 60 km from east to west. Its elevation ranges from 1300 to 1500 m, with a relative height of 50–100 m, classifying it as a mid-mountain landform. The mountain top is characterized by exposed bedrock composed of Tertiary mudstone, sandstone, and conglomerate. The slopes are gentle, rounded, dry, and devoid of vegetation.

The anticline structure restricts groundwater storage, recharge, and discharge, leading to significant variations in water resource distribution. On one side, the rising elevation of the aquifer base causes the aquifer to thin and the water table to become shallower, directing groundwater flow along the uplifted edges. On the opposite side, structural barriers significantly limit recharge, resulting in minimal groundwater replenishment.

Yingjisha County can, thus, be divided into two distinct irrigation zones: the southern and northern regions. The southern region consists of a narrow belt of alluvial and proluvial fan groups originating from the Kunlun Mountains’ foothills, while the northern region features a relatively flat alluvial plain.

2.2. Research Methods

(1) Sample Collection and Testing. In order to determine the spatial distribution characteristics of groundwater in the study area, a total of 6 soil samples and 308 groundwater samples were collected within Yingjisha County. These groundwater samples included 127 phreatic water samples, 173 first confined aquifer samples. Descriptive statistical analysis was used to organize and analyze the measured data from the sampling points. In order to better classify the local silt and sand, this study studied the composition of soil particles in order to find out the proportion of local fine particles below a certain size. According to the specific local irrigation conditions, 6 typical soil samples were collected in Yingjisha County, as shown in

Table 1. Soil sample point.

Number	Crop	Soil Texture	Soil Natural Density/(g·cm−3)	Irrigation Water Quality/(g·L−1)	Irrigation Method
K1	cotton	mealy sand	1.4866	3.8	drop irrigation
K2	cotton	mealy sand	1.4595	4.0	drop irrigation
K7	wheat	sandy, mealy sand	1.3670	2.0	flood irrigation
K9	cotton	sand	1.6600	3.8	drop irrigation
K21	wheat	mealy sand	1.3403	1.0	drop irrigation
K22	wheat	sandy, mealy sand	1.3387	0.8	flood irrigation

below. When collecting soil samples, the depth of 80 cm was dug, and the graded sampling and record were carried out, and the soil moisture TDS measured on site was used for indoor matching.

Weigh 500 g of dry sand and use the screening method to sift the soil sample through a variety of filter sizes; according to the size of the sieve, particles are split by size into several groups. Finally, weigh to determine the proportion of each sample, in order to classify various types of soil.

A Pearson correlation analysis was conducted on soil samples to investigate the relationship between total soluble salts and soil particle size distribution. This analysis provides fundamental data for modeling water–salt transport in the vadose zone, particularly regarding the influence of soil particle size on electrical conductivity in the study area [13]. A correlation equation was established with an R² value of 0.972, indicating a good fit. The correlation analysis equation is as follows:

y = −0.1689x₁ + 1.4845x₂ − 0.545x₃

(1)

(2) Physical Simulation. Irrigation with mildly saline water requires consideration of the crop irrigation water salinity threshold and the soil salinization threshold. Mildly saline water irrigation can be applied under the condition that the soil salinity in the cotton root zone maintains a basic balance of leaching and accumulation, the physico-chemical properties of the tillage layer do not significantly deteriorate, and cotton yield, fiber quality, and water use efficiency do not significantly decrease [30]. This study conducted controlled irrigation experiments based on the climatic characteristics, crop cultivation patterns, and irrigation water sources in the Yingjisha irrigation district. The experiments investigated: (1) salt transport mechanisms under different irrigation conditions with varying lithologies and TDS levels, and (2) the effects of irrigation recharge on soil salinity distribution.

(3) Numerical Simulation. This study established a Hydrus-1D unsaturated zone water–salt transport model based on physically simulated parameters, including soil–water characteristics, irrigation volume, evaporation rate, and initial distributions of both soil moisture and salinity. The model was adjusted based on the comparison between the simulated and observed soil moisture content and salinity at different observation points over time. The final parameters were determined through model validation, and the parameters are shown in

Table 2. Soil moisture characteristics and soil layer salt migration parameters.

ID	θr/ (L3L−3)	θs/ (L3L−3)	a	n	Ks/ (L∙T−1)	P/ (g∙cm−3)	DL/ (cm2∙d−1)	DW/ (cm2∙d−1)
K1	0.037	36	0.0896	1.87	3.04	1.57	3.5	17
K2	0.044	34.2	0.586	1.66	42.5	1.62	2.6	54
K7	0.038	35.34	0.767	1.89	12.2	1.33	3.2	32
K9	0.245	29.8	0.495	2.10	8.40	1.25	12.4	19
K21	0.034	36.2	0.812	1.43	14.3	1.54	6.3	87
K22	0.032	33.7	0.854	1.38	23. 2	1.42	4.8	54

θr (residual water content) represents the irreducible liquid water fraction retained in soil after intense desiccation, which cannot be removed by gravitational drainage or plant root extraction. θs (saturated water content) denotes the maximum volumetric water-holding capacity in completely saturated porous media (unit: L³L⁻³).

.

Using the unsaturated zone one-dimensional water and salt transport model, combined with the non-salinity criteria, soil water and salt predictions for the shallow unsaturated zone under different conditions were made. Based on the non-salinity threshold and the required solute transport model parameters, a correlation between the soil water electrical conductivity X (ms/cm) measured using the EM50 sensor and the total salt content Y (%) from soluble salt tests was established in the laboratory. The correlation yielded an R² value of 0.929, with the fitting equation shown in Equation (2). According to the national standard “Soil Environmental Quality Standards GB15618-2018” [31], and considering both the total salt content in the soil and the impact of salinization on crop growth, a threshold of 0.3% total salt content was set. Using the following relationship, it was calculated that when the soil solution electrical conductivity threshold is below 0.91 ms/cm, it can be considered non-salinized.

y = 0.8266x + 0.0318

(2)

Based on field investigation data from sampling sites, including irrigation water quality and irrigation schedules, this study conducted saline water irrigation simulations to predict soil electrical conductivity (EC) variations at fixed groundwater table depths. The simulation results were used to determine the critical water table depth at which EC values remained below the salinity threshold for non-saline conditions in the unsaturated zone over a 120-day period. The calculated thresholds for irrigation water salinity and corresponding computational results are presented in

Table 3. Optimal water level and water quality for crop irrigation.

ID	Lithology	Current Water Level/(m)	Optimal Water Level Buried/(m)	TDS of Irrigation Water/(g·L−1)
K1	mealy sand	5.6	5.6	4
K2	mealy sand	6.0	6	4
K7	mealy sand	6.0	5.7	2.8
K9	sand	5.8	5.62	4
K21	mealy sand	5.3	4.15	2.8
K22	mealy sand	5.0	4.5	2.8

.

Based on the geological and hydrological borehole data, the numerical model of groundwater flow and solute transport were constructed and verified by the water level monitoring data of 7 observation wells in the study area. The fitting error between the simulated groundwater level and the measured value should be less than 10% of the water level change during the fitting calculation period.

In the case of small water level change (less than 5 m), the water level fitting error should generally be less than 0.5 m. Based on the upper and lower limits of the groundwater level in the simulation of water and salt migration in the calculated zone, and the principle that the area less than the buried depth line of the mining water level is mined to the buried depth line of the water level, the area greater than the buried depth line of the water level is not mined, and the maximum decline rate of confined water is not more than 1 m/a, the underground water flow numerical model is used to calculate the amount of groundwater recoverable resources under different water level constraints. Then, subdivide the recoverable amount of groundwater with different salinity based on the spatial distribution characteristics of groundwater [28,29].

2.3. Optimum Collocation Model

(1)

Objective Function

(1) The economic benefits are most effectively represented by the water supply benefits for water users in the irrigation area.

$$max{f}_1\left(x\right)={\sum }_{i=1}^I{\sum }_{j=1}^J{\sum }_{k=1}^K\left(b_{ij}^k-c_{ij}^k\right){(x}_{ij}^k{+Q_j^k)\alpha }_{ij}^k{\beta }_{ij}^k$$

(3)

In the formula, the net economic benefit is the net benefit per unit of water use multiplied by the unit of water amount, and the cost and benefit of water use for reservoir water and groundwater water are the same. $f_1\left(x\right)$ is the economic benefits of the irrigation area. $b_{ij}^k$ is the water supply efficiency for water user j from water source i in irrigation area k. $c_{ij}^k$ is the water usage cost for water user j from water source i in irrigation area k. ${\alpha }_i^k$ is the water supply order coefficient for water source i in irrigation area k. $x_{ij}^k$ is the supply amount of water for water user j in irrigation district k. $Q_j^k$ represents the amount of water resources provided by the reservoir water source to water user j in sub-district k.

$${\alpha }_{ij}^k={\displaystyle \frac{1+n_{imax}^k-n_i^k}{{\sum }_{i=1}^I(1+n_{imax}^k-n_i^k)}}$$

(4)

$${\beta }_{ij}^k={\displaystyle \frac{1+m_{i max}^k-m_i^k}{{\sum }_{i=1}^I(1+m_{i max}^k-m_i^k)}}$$

(5)

$n_i^k$ is the sequence number of water source i supplying water to irrigation area k. $n_{i max}^k$ k is the maximum value of the supply sequence number. $m_i^k$ is the sequence number of water use for water user j in sub-area k. $m_{i max}^k$ is the maximum value of the water use sequence number. The smaller the supply sequence number of the water source, the greater its supply priority coefficient value, indicating a higher priority for supply [26,32].

(2) The ecological benefits are represented by the minimum excessive exploitation of groundwater in the irrigation area.

$$min{f}_2\left(x\right)={\sum }_{k=1}^K{\sum }_{j=1}^J(N_j^k-{\sum }_{i=1}^I{x}_{ij}^k-Q_j^k)$$

(6)

In the formula, the ecological benefit is the minimum water shortage, that is, the water demand decreases the water supply. $f_2\left(x\right)$ is the social benefits, $N_j^k$ is the irrigation water amount for water users j in irrigation district k.

(2)

Constraints

(1) The supply capacity of the reservoir water is constrained. The sum of irrigation water for all water users in the k irrigation area cannot exceed the reservoir capacity.

$${\sum }_{k=1}^K{\sum }_{j=1}^J{Q}_j^k\le Q$$

(7)

(2) Water users in the irrigation area are subject to water demand constraints. The water demand for crop j in irrigation area k exceeds the irrigation water supply.

$$N_{\min j}^k>{\sum }_{i=1}^I{x}_{ij}^k+Q_j^k$$

(8)

$N_{\min j}^k$ is the current irrigation water amount for crop j in irrigation district k.

(3) Groundwater supply capacity constraints. In order to prevent soil salinization and avoid the formation of water level drop funnels, the burial depth of groundwater should be controlled below the critical depth for salt return and above the maximum allowable extraction depth, namely,

$${\sum }_{j=1}^I{x}_{ij}^k\le H_{\max i}^k$$

(9)

The extract is as follows: $H_{min i}^k$ is the exploitable water volume of the i water source in the k irrigation area when the groundwater level is greater than the critical burial depth. $H_{\max i}^k$ is the exploitable water volume of the i water source in the k irrigation area when the groundwater level is less than the maximum burial depth.

(4) Configure the water TDS constraint. After blending the slightly saline underground water with the reservoir water, the TDS of the blended water falls within the range of the ecological salinity threshold.

$$d_{minj}^k\le {\displaystyle \frac{{2.6Q}_j^k+{\sum }_{i=4}^{i=6}{d}_{ij}^k{x}_{ij}^k}{Q_j^k+x_{ij}^k}}\le d_{maxj}^k$$

(10)

In the formula, $d_{\min j}^k$ represents the minimum value of TDS of the irrigation water configuration for crop j in irrigation area k. $d_{\max j}^k$ represents the maximum value of TDS of the irrigation water configuration for crop j in irrigation area k.

$$x_{ij}^k\ge 0$$

(11)

2.4. Water Source Division

Agricultural water use in Yingjisha County consists of two types of irrigation water sources: reservoir water and groundwater. Groundwater is classified based on different salinity levels, while reservoir water is a public water resource that requires coordinated allocation within the county. Water users include cotton farmers, wheat and corn farmers, and other crop farmers, as detailed in

Table 4. Water source classification table.

Unconfined Water						Confined Water			Reservoir Water
i = 1	i = 2	i = 3	i = 4	i = 5	i = 6	i = 7	i = 8	i = 9	i = 10
<1 g/L	1–2 g/L	2–3 g/L	3–4 g/L	4–5 g/L	>5 g/L	<1 g/L	1–2 g/L	2–3 g/L	2.6 g/L

.

2.5. Solving Model

The PSO algorithm is a novel swarm intelligence evolutionary algorithm with advantages such as fast convergence speed, high search efficiency, and a simple iterative process. It has been widely applied in fields such as function optimization. The modified particle swarm optimization (PSO) algorithm demonstrates enhanced precision in identifying optimal solutions [20]. In this study, we employed a linear weighting method to transform the multi-objective problem into a single-objective formulation, assigning weights of 0.3 and 0.7 to economic and ecological objectives, respectively. The improved PSO algorithm was subsequently applied to solve the water resources zoning optimization model under multiple constraints. Using MATLAB(R2022b) to write improved PSO algorithm code to solve the optimal allocation of water resources. In the optimal configuration model, the rate of TDS threshold parameter will lead to the change in reservoir water distribution for diluting underground saltwater in the optimal configuration result. The change in groundwater level simulated by water and salt migration in the vadicle zone will lead to the change in underground available water supply, and the greater the buried water level, the greater the underground available water supply [25].

3. Results

3.1. Spatial Distribution Characteristics of Multiple Water Sources

The anticlinal structure caused significant differences in the spatial distribution of groundwater, and the diving TDS concentration ranged from 590 mg/L to 6352 mg/L, with an average value of 3396.2 mg/L. There was a significant difference in diving TDS between the south and north of the anticlinal, and the gradual increase in TDS from southwest to northeast was consistent with the direction of groundwater runoff. The concentration of TDS in confined water in the first layer is between 633 mg/L and 3860 mg/L, with an average value of 1447 mg/L. The difference in TDS under the influence of anticline is significant. Compared with the range and average value of diving TDS, the TDS gradually increases from southwest to northeast, which is consistent with the direction of groundwater runoff. The TDS concentration of the second layer confined water ranges from 755 mg/L to 984 mg/L, with an average value of 893 mg/L and a coefficient of variation of 0.09. The groundwater concentration has little change, and the TDS in the whole region is less than 1 g. The distribution of water quality in Yingjisha County is shown in Figure 2.

3.2. Irrigation District Division

According to the spatial distribution characteristics of groundwater, the research area is divided into 3 irrigation areas, as shown in Figure 3. In the southern irrigation area, the gradient of dive salinity from southwest to northeast is large, and freshwater resources are short, so it is necessary to dilute low salinity water with high salinity dive reservoir water, which is mainly used for cotton irrigation. The northern irrigation area has high diving salinity and the cotton planting area accounts for most of the county, so it is preferred to irrigate cotton with underground saline water. The irrigated area of cotton in the western irrigation area accounts for 1.72%, and the water with high salinity is less, and the agricultural irrigation mainly uses freshwater.

Based on agricultural water use data provided by the Yingjisha County Water Resources Bureau, the irrigation water consumption for each crop in the three irrigation districts was analyzed. The agricultural irrigation water consumption for each crop from September 2021 to August 2022 is detailed in

Table 5. Total crop irrigation water volume from September 2021 to August 2022.

	South Irrigation Area/104 m3	Northern Irrigation Area/104 m3	Western Irrigation Area/104 m3
Cotton	911.96	1698.09	310.39
Wheat and Corn	3321.32	7238.19	12,663.72
Other crops	3300.87	2340.77	5073.04

.

3.3. Output of Supplying Water

The exploitable groundwater volumes of different water quality in each irrigation district are shown in Figure 4. This work ensures that the groundwater level remains within an appropriate range, preventing land salinization due to excessive water levels while also avoiding impacts on plant root growth caused by excessively low water levels. The proportion of underground diving is 43.1%, and the proportion of underground saltwater is 56.9%. The proportion of groundwater freshwater in the first layer is 93.1%, and the proportion of groundwater saltwater is 6.9%. The TDS of the second layer of confined water in the whole area is less than 1 g.

3.4. Optimal Configuration Result

The optimized allocation results are presented in

Table 6. Optimal allocation of water resources results unit: 10⁴ m³

Area		Unconfined Water Freshwater			Unconfined Water Saltwater			Confined Water Freshwater			Water of Reservoir	Total
		i = 1	i = 2	i = 3	i = 4	i = 5	i = 6	i = 7	i = 8	i = 9
k = 1	j = 1	2	69	12	294	285	4	4	34	81	101	886
	j = 2	1	27	90	17	0	0	39	287	484	1800	2745
	j = 3	16	71	17	12	0	0	12	1193	666	1270	3257
k = 2	j = 1	0	0	1	4	424	24	76	33	8	1105	1675
	j = 2	0	0	4	9	0	0	24	233	60	5180	5510
	j = 3	0	0	0	4	0	0	63	330	18	1536	1951
k = 3	j = 1	30	26	16	1	1	0	14	42	0	96	226
	j = 2	169	212	45	2	0	0	184	13	0	10,610	11,235
	j = 3	67	53	23	83	0	0	803	124	0	3661	4814
Total	Total	285	458	208	426	710	28	1219	2289	1317	25,359	32,299

. The reservoir water withdrawal reached 253.59 million m³, while the total groundwater extraction amounted to 69.40 million m³. The water shortage was recorded at 45.59 million m³, corresponding to a shortage rate of 12%, with an economic benefit of 190 million CNY. The intensity of groundwater exploitation has decreased significantly: the overall groundwater exploitation level decreased from 123% to 74%, and the utilization rate of fresh groundwater declined from 127% to 64%. Additionally, saline groundwater extraction reached 11.64 million m³, with its utilization rate increasing from 0% to 83%. This allocation scheme not only ensures water supply security but also effectively alleviates regional groundwater overexploitation and enhances the efficient utilization of saline water resources.

In the southern, northern, and western irrigated areas, the amount of groundwater recoverable is 36%, 36%, and 5%, respectively, and the water consumption of cotton in the southern, northern, and western irrigated areas accounts for 12%, 15%, and 1% of the water consumption of the irrigated areas, respectively, so it is affected by the amount of underground saline water and cotton planting area. Under maximum extraction water levels, the utilization rates of mildly saline groundwater in the southern, northern, and western irrigation districts are 78%, 42%, and 10%, respectively. After optimization, the water shortage rate is reduced to 12%, generating an economic benefit of 1.9 billion yuan. The phreatic aquifer extraction amounts to 211.5 million m³, of which 116.4 million m³ (55.0%) is saline groundwater with TDS above 3 g/L. The confined aquifer extraction totals 482.5 million m³, while reservoir water extraction reaches 2535.9 million m³, 0.935 million m³ less than the total supply.

In the optimized scheme, freshwater groundwater extraction is reduced to 577.6 million m³, lowering the freshwater utilization rate from 127% to 64%. The water supply reliability is shown in

Table 7. Water supply reliability unit: 10⁴ m³

Current Water Quantity	p = 50%	p = 75%	p = 90%	Optimized Guarantee Rate = 88%
36,858	18,429	27,644	33,172.2	32,299.0

, with an optimized supply guarantee rate of 88%.

4. Discussion

4.1. Analysis of Mining Structure

The Comparison of underground freshwater and saltwater production under the configuration scheme is shown in Figure 5. Zhao Bin proposed the principles of rational water resource allocation [11,25]. Based on these principles, this study sets supply and consumption coefficients, considering that the exploitation of mildly saline groundwater is influenced by both water quality allocation and cotton irrigation water demand.

Cotton irrigation accounts for 12.1% of the total irrigation water in the southern irrigation district and 15.06% in the northern irrigation district. Together, these two districts make up 91.5% of the total cotton planting area in Yingjisha County. Due to the larger cotton planting areas and higher groundwater mineralization in these two districts, the mixed water quality of reservoir water and groundwater source four results in a TDS exceeding 2.6 g/L, which is only suitable for cotton irrigation. Therefore, the southern and northern irrigation districts prioritize the utilization of saline water, with cotton cultivation accounting for the majority of saline water consumption in their allocation structure. In contrast, the western irrigation district has limited saline water resources and a smaller cotton cultivation area, resulting in lower supply priority and the lowest utilization rate. These findings demonstrate effective control over the multi-source water supply sequence, validating the rationality of the optimized allocation model for saline groundwater resources.

In contrast, the western irrigation district has less mildly saline groundwater and a smaller cotton planting area, with a lower supply priority, leading to the least utilization of mildly saline groundwater. The model’s designed supply and consumption priority order aligns with the actual supply structure of mildly saline water utilization, demonstrating that the optimized water resource allocation model for differentiated water quality irrigation is reasonable.

4.2. Agricultural Irrigation Water Shortage

Figure 6 illustrates the comparison between water users’ demand and the allocated water volume in the optimized scheme. The northern and western irrigation districts account for 85.9% of the total water deficit, which can be attributed to two main factors: (1) the western district’s extensive cultivation of wheat, corn, and other crops with high freshwater demands, and (2) the northern district’s inherent scarcity of freshwater resources. In contrast, the southern district exhibits the lowest water deficit due to its minimal crop cultivation area and correspondingly lower water requirements. Notably, wheat and corn in the western and northern districts experience higher water deficits compared to other crops. This distribution pattern results from the optimization model’s parameter settings, where water allocation coefficients for equity and supply benefit were designed to prioritize economic crops such as cotton [20,25]. The observed allocation outcomes demonstrate the model’s effectiveness in optimizing saline water utilization while maintaining balanced water distribution among competing demands.

The utilization of saline groundwater for cotton irrigation represents the primary approach for optimal allocation of saline water resources. The optimization results demonstrate a remarkable increase in saline water utilization from 0% to 83%. Notably, the extraction of saline groundwater in the northeastern region has contributed to a decline in phreatic water levels, thereby mitigating soil secondary salinization caused by the shallow water table of highly mineralized water in the phreatic aquifer. The potential of underground saline water is developed for agricultural irrigation of salt-tolerant crops, and the overall irrigation water volume is increased, which ensures the normal water requirement of crops in arid areas with water shortage and meets the requirements of sustainable agricultural economic development.

4.3. Water Resources Change Trend Analysis

In the water balance calculation, the natural groundwater recharge is 93.33 million m³, while the optimized groundwater extraction under the proposed scheme is 69.40 million m³, leaving 23.93 million m³ of groundwater unused. This effectively addresses the issue of continuous groundwater level decline and aligns with the concept of sustainable groundwater utilization. Of the total natural groundwater recharge, 80.77 million m³ is fresh groundwater, while 12.55 million m³ is saline groundwater. Under the recharge-discharge balance condition, adding the remaining 11.37 million m³ of fresh groundwater to the optimized irrigation allocation would reduce the water shortage rate to 5.9%.

Table 8. Table of groundwater utilization degree.

Area	Actual Production/(104 m3)	Natural Groundwater Recharge/(104 m3)	Utilization Degree	Project Recovery/(104 m3)	Optimal Utilization Degree	Drawdown/ (104 m3)
Southern	5818	4683	124%	3717	79%	36%
Northern	2880	2457	117%	1315	53%	54%
Western	2791	2193	127%	1908	87%	32%
Total	11,489	9333	123%	6940	74%	40%

presents the utilization of groundwater resources before and after optimization. Before optimization, the actual groundwater extraction in the southern irrigation district was 58.18 million m³, exceeding the maximum exploitable volume of 46.82 million m³, resulting in a groundwater utilization rate of 124%. After optimization, the utilization rate was reduced to 79%, representing a 36% reduction in groundwater extraction. This greatly reduces the exploitation rate of groundwater, indicating that the optimal allocation model of fresh–saltwater resources is advanced in the sustainable development of agricultural economy [27].

4.4. Comparative Analysis of Water Resources Consumption

A comparison of the actual water usage in Yingjisha County from 2021 to 2022 with the optimized water allocation for each type of water source is shown in Figure 7. Groundwater depth exerts a significant influence on crop growth. Within a certain range, saline water irrigation can stimulate plant growth without causing significant yield reduction or even enhance productivity, while simultaneously improving water use efficiency [13,14,30]. This study adopts 6 m as the maximum groundwater level threshold for crop growth. However, field investigations reveal that most areas have already exceeded this critical threshold. Continued over-extraction would further exacerbate agroecological degradation.

A comparison between the actual water usage in the baseline year and the optimized allocation results is presented in Figure 6. The western and northern irrigation districts demonstrate sufficient reservoir storage capacity, strong regulation capability, and high dependence on reservoir water. In contrast, the southern district exhibits the lowest actual water consumption but possesses the greatest exploitable groundwater potential, primarily due to its smaller cultivation area (accounting for only 14.9% of the total). This suggests that agricultural development priorities should shift toward the southern district.

Notably, the actual water usage across all three irrigation districts shows strong agreement with the optimized allocation results, confirming the high efficiency of the proposed irrigation system optimization.

4.5. Deficiency and Prospect

(1) The economic benefit objective function in the optimization model incorporates both water supply priority coefficients and water use equity coefficients. These coefficients dynamically influence economic outcomes based on variations in water supply and utilization sequences. Consequently, while the economic benefit primarily serves to regulate water allocation priorities, there remains a need for further refinement to enhance the precision of economic benefit modeling.

(2) The water level constraint in the optimization model fails to limit the minimum mining level, because the water distribution salinity constraint restricts the full mining of water under the minimum water level constraint. Although the optimization results cannot restrict the minimum mining water level, the underground saltwater used for cotton irrigation has reached the expected goal. Future research needs to further explore the constraint conditions of the minimum extraction water to improve the model parameter calibration.

5. Conclusions

(1) The water resource optimization model demonstrates robust consistency between the designed water supply/utilization sequence and the actual groundwater allocation order in the optimization results. This alignment ensures preferential utilization of saline water for salt-tolerant cotton cultivation while strictly satisfying irrigation water quality constraints, thereby verifying both the rationality and advanced nature of the proposed optimization framework.

(2) Through integrated freshwater–saline water optimization allocation, this study successfully exploits the utilization potential of saline groundwater, significantly improving the efficiency of groundwater-based irrigation systems. The southern irrigation district exhibits the greatest exploitation potential, suggesting that future agricultural development should strategically prioritize this region.

(3) The optimization results achieve substantial utilization of saline water irrigation potential, effectively mitigating the risk of secondary soil salinization caused by excessively shallow water tables and providing notable ecological protection. Future implementation of water-saving irrigation techniques combined with rational reduction in irrigated areas could potentially eliminate water shortages, thereby promoting sustainable agricultural economic development.