Evaluation of Water Security in a Water Source Area from the Perspective of Nonpoint Source Pollution

Abstract

Water security is a basic requirement of a region’s residents and also an important point of discussion worldwide. The middle route of the south-to-north water diversion project (MR-SNWDP) represents the most extensive inter-basin water allocation scheme globally. It is the major water resource for the Beijing–Tianjin–Hebei region, and its security is of great significance. In this study, 28 indicators including society, nature, and economy were selected from the water sources of the MR-SNWDP from 2000 to 2017. According to the Drivers-Pressures-States-Impact-Response (DPSIR) framework principle, the entropy weight method was used for weight calculation, and the comprehensive evaluation method was used for evaluating the water security of the water sources of the MR-SNWDP. This study showed that the total loss of nonpoint source pollution (NPSP) in the water source showed a trend of slow growth, except in 2007. Over the past 18 years, the proportion of pollution from three NPSP sources, livestock, and poultry (LP) breeding industry, planting industry, and living sources, were 44.56%, 40.33%, and 15.11%, respectively. The main driving force of water security in all the areas of the water source was the total net income per capita of farmers. The main pressure was the amount of LP breeding and the amount of fertilizer application. The largest impact indicators were NPSP gray water footprint and soil erosion area, and water conservancy investment was the most effective response measure. Overall, the state of the water source safety was relatively stable, showing an overall upward trend, and it had remained at Grade III except for in 2005, 2006, and 2011. The state of water safety in all areas except Shiyan City was relatively stable, where the state of water safety had fluctuated greatly. Based on the assessment findings, implications for policy and decision-making suggestions for sustainable management of the water sources of the MR-SNWDP resources are put forward. Agricultural cultivation in water source areas should reduce the application of chemical fertilizers and accelerate the promotion of agricultural intensification. Water source areas should minimize retail livestock and poultry farming and promote ecological agriculture. The government should increase investment in water conservancy and return farmland to forests and grasslands, and at the same time strengthen the education of farmers’ awareness of environmental protection. The evaluation system of this study combined indicators such as the impact of agricultural nonpoint source pollution on water bodies, which is innovative and provides a reference for the water safety evaluation system.

1. Introduction

Water is the source of life. With socio-economic development, changes in climatic conditions, and the blind and transitional development of water resources by human beings, the security of water resources systems is facing unprecedented challenges [1,2]. The security of water resources in the water source area guarantees the safety of the residents living in the area. Eutrophication of aquatic ecosystems poses the most significant challenge to hydrological environmental quality [3,4] Since the 1990s, few river basins in Africa, Asia, or Latin America have remained unaffected by declining water quality [5,6]. Sources polluting water environments are mainly divided into point source pollution and nonpoint source pollution. Point source contamination denotes identifiable pollution origins characterized by stationary emission locations, including industrial effluents and municipal wastewater, which are directly introduced into aquatic systems through designated discharge points [7]. The government has taken significant steps in controlling point source pollution and has improved the standard system as well as the laws and regulations for the prevention and control of industrial water pollution [8]. However, even with the full control of industrial point source pollution, the problem of water environment pollution in the river basin has not completely improved [9]. In comparison to point source pollution, nonpoint pollution is difficult to control and prevent because of its randomicity, intermittence, hysteresis, and complexity [10]. Studies have shown that nonpoint source pollution has exceeded point source pollution and has become the main source of pollutants in water bodies [11]. Nonpoint source pollution (NPSP) mainly comes from the large-scale use of fertilizers and pesticides in agricultural production, such as the unreasonable discharge of manure from the rural farming industry and the random disposal of rural household waste [12]. It is estimated that 60% of nitrate pollutants and 25% of phosphorus-containing substances in British waters come from agricultural activities [13], and the response time of nonpoint source pollution at the basin scale can take as long as several decades; in the United States, agricultural production has become the first source of river pollution, and agricultural NPSP accounts for 2/3 of the pollution load [14]; in the Netherlands, total nitrogen (TN) and total phosphorus(TP) from agricultural NPSP account for 60% and 40% of total water pollution, respectively [15]. China’s inaugural nationwide assessment of contaminant origins revealed that NPSP contributes more than half of the TN and TP loading to aquatic systems. Projections indicate an exacerbation of water quality degradation in subsequent decades, inevitably amplifying risks to public health, ecological integrity, and long-term socio-economic sustainability [16]. Therefore, NPSP is an important consideration in the process of assessing water resources security.

Water security evaluation is a systematic evaluation of the overall health of the aquatic ecosystem from the perspective of resource utilization, environmental protection, and disaster avoidance through a multidimensional and comprehensive perspective [17]. It provides an important basis for monitoring and optimizing the sustainable use of water resources in the river basin. The complexity of the system must be considered in the study of water resources security in the river basin. A comprehensive evaluation of the water resources framework must be conducted, encompassing ecological, socio-economic, and environmental dimensions, to address the critical constraints hindering water quality enhancement and long-term water resource sustainability [18,19]. Climate change leads to changes in precipitation patterns and an increase in extreme weather events, directly affecting the distribution of water resources and the stability of water ecosystems. Social factors, such as population growth and urbanization, accelerate water demand, while environmental awareness and water use habits affect water use efficiency. Industrial development patterns and water pricing mechanisms determine the intensity of water use, and a rough economy often leads to over-exploitation and pollution. Industrial wastewater, agricultural surface pollution, and ecological damage exacerbate the deterioration of water quality, and the decline in biodiversity weakens the self-purification capacity of water bodies. These factors are intertwined and together threaten water ecological security, and systematic management is urgently needed to achieve sustainable development. In 1993, the United Nations proposed and developed the Driver–Pressure–State–Impact–Response (DPSIR) model in order to comprehensively analyze and describe environmental issues and their relationship with social development [20], combining the advantages of the Pressure–State–Response (PSR) and the Driver–State–Response (DSR) models. The DPSIR model describes a causal chain between the origin and consequence of an environmental problem [21,22]. This cause–effect chain indicates that social, economic, and population development act on the environment as a long-term driving force, thus putting pressure on the environment, causing changes in the state of the ecological environment, and causing damage to the ecological environment such as water resources [23,24]. These impacts prompt humans to respond to changes in the state of the ecological environment, and the response measures act on a composite system of society, economy, and population or directly on environmental pressure, state, and impact. This model studies the interaction between humans and environmental systems from the perspective of system analysis. It is a conceptual model of the evaluation index system widely used in environmental systems and a general framework for organizing environmental state information. The DPSIR model provides a systematic framework for the study of water ecological security, which can comprehensively integrate natural hydrological processes and socio-economic factors, and clearly reveal the causal chain from drivers (e.g., population growth) to pressures (e.g., pollution discharges), states (e.g., deterioration of water quality), impacts (e.g., ecological degradation), and then responses (e.g., governance policies). Unlike simpler models (e.g., Pressure–State–Response), DPSIR’s inclusion of drivers and impacts provides a more nuanced understanding of root causes and consequences. The model is suitable for interdisciplinary, integrated water resources management and provides scientific basis for long-term monitoring and precise management of water ecological security. Therefore, in this study, we used the DPSIR framework to systematically evaluate the security states of water resources in MR-SNWDP from the perspective of NPSP, so as to provide a scientific basis for sustainable use of water resources in the basin.

2. Materials and Methods

2.1. Study Area

The research area of this study is the water source area of MR-SNWDP, including Hanzhong, Ankang, Shangluo, Shiyan, and five counties of Nanyang city (Figure 1) (31°25′–34°12′ N, 106°5′–112°50′ E). The MR-SNWDP is a strategic water transfer project in China, which aims to solve the serious problem of a shortage of water resources in Beijing–Tianjin–Hebei in north China. The main rivers in the water source area are the Han and the Danjiang. As the largest tributary of the Yangtze River, the Hanjiang River Basin has an annual precipitation of 800–1200 mm, which is rich in water resources; at the same time, the basin has a large proportion of hills and mountains, with good vegetation cover, which is conducive to water conservation. The total per capita of water resources in the water source area is 3741 m³/person, which is nearly 70% higher than the national per capita level (2200 m³/person) [25]. The area not only acts as a water source for local use but also for parts of North China. Therefore, the need to secure the water resource is highly significant.

2.2. Selection of Assessment Indicators

The framework of the DPSIR model can be used for the analysis of human activity and environmental causality. The DPSIR model divides the complex interrelationship between environmental change and human behavior into five processes: driving force, pressure, state, impact, and response according to causal logic. The driving force is a potential factor that changes the water environment. Modernization, characterized by population growth, urbanization, and economic development are the main causes of environmental pollution [26]. In this study, regional GDP, total agricultural output, urbanization rate, and population density, per capita GDP, and total net income per farmer were selected as the indicators of driving force. The pressure indicator refers to the pressure caused by human social production activities on the environment. In order to obtain higher economic benefits, human investment in production activities has gradually increased. As the pressure index for this study, 10,000 Yuan of GDP water consumption, chemical fertilizer use, pesticide use, mulch use, amount of livestock and poultry (LP) breeding, and multiple crop index were selected. States indicators refer to the state in which the natural environment is affected by stress. The states indicators for this study were selected as per capita water resource, the total population, total annual water consumption, water quality up to the standard rate, total nitrogen from planting sources into the river, total nitrogen from living sources into the river, and total nitrogen from livestock manure into the river. Impact indicators refer to the effects of environmental conditions on the natural environment. This study selected soil erosion area, flood disaster losses, drought losses, total nonpoint source pollution load, and forest cover rate. Response indicators refer to measures taken by people to promote sustainable development of the water environment. The indicators selected in this study were water-saving irrigation area, total investment in water conservancy, water and soil loss area treatment rate, and 10,000 Yuan GDP water consumption rate of decline (

Table 1. Details of DPSIR components, indicators, and index number.

Component	Indicators	Index Number
Driver	GDP (yuan)	X1
	Total agricultural output (yuan)	X2
	Population density (person/km2)	X3
	Urbanization rate (%)	X4
	Per capita GDP (yuan)	X5
	Total net income per capita of farmers (yuan)	X6
Press	Ten-thousand-yuan GDP water consumption	X7
	Pesticide use (kg/ha)	X8
	Mulch use (kg/ha)	X9
	Chemical fertilizer use (kg/ha)	X10
	Number of livestock and poultry farming (head)	X11
	Multiple crop index (%)	X12
State	Per capita annual water resources (m3)	X13
	The total population (million people)	X14
	Total annual water consumption (m3)	X15
	Water quality up to the standard rate (%)	X16
	Total nitrogen from planting sources into the river (kg)	X17
	Total nitrogen from living sources into the river (kg)	X18
	Total nitrogen from livestock manure into the river (kg)	X19
Impacts	Soil erosion area (ha)	X20
	Flood disaster loss (yuan)	X21
	Drought loss (yuan)	X22
	Gray water footprint of nonpoint source pollution (m3)	X23
	Forest cover rate (%)	X24
Responses	Water-saving irrigation area (ha)	X25
	Total investment in water conservancy (yuan)	X26
	Soil erosion control area (ha)	X27
	Ten-thousand-yuan GDP water consumption rate of decline (%)	X28

).

2.3. Data Sources

The data of each indicator in this study were derived from the City Statistical Yearbook of Shiyan, Nanyang, Shangluo, Ankang, and Hanzhong, the Animal Husbandry Department of the Agricultural Ministry in China, and the Water Resources Bulletin for the five cities. Furthermore, information regarding aggregate water resource availability was obtained from the official Water Resources Bulletins of the five urban centers.

2.4. Analytical Methods

Agricultural nonpoint source pollution directly affects water quality and safety [27]. Several studies have shown that total nitrogen (TN) is the largest influencing factor of nonpoint source pollution on water bodies [28]. Therefore, NPSP, in this paper, only calculates the total nitrogen produced by planting industries, LP breeding industries, and living sources. Table 1 lists the specific indicators (Xi represents index number, i = 1, 2, 3 … 28). Here, TN from planting sources, living sources, and livestock manure into the river (kg) were calculated by referring to other research.

The TN calculation formula for the inflow river generated by the planting industry is [29]:

$$\mathrm{F}\mathrm{T}\mathrm{N}=\mathrm{N}\mathrm{F}\ast \mathsf{\alpha }$$

Here, FTN (kg) refers to the TN of fertilizer discharged into the river; NF (kg) is the amount of pure nitrogen fertilizer; and α the coefficient for pollution discharged into the river (

Table 2. Pollution production coefficient and river entry coefficient of various pollution sources.

	Swine	Cattle	Sheep	Poultry	Fertilizer	Rural Domestic
TN (g/head·day)	12.4	32.15	4.13	0.05	-	10
Coefficient of TN into river (%)	5.25	5.68	5.3	8.47	12	10

).

The total amount of LP in the area within the breeding industry is calculated as swine equivalent. For this purpose, one head of swine amounts to 0.074 head of total cattle, 0.79 head of sheep, and 16.7 heads of poultry. The TN calculation formula for the inflow river generated by the LP breeding industry is [30,31,32]:

$$\mathrm{L}\mathrm{T}\mathrm{N}=\mathsf{\Sigma }{\mathrm{N}}_{\mathrm{j}}{\mathrm{D}}_{\mathrm{i}\mathrm{j}}{\mathrm{T}}_{\mathrm{j}}\mathsf{\alpha }$$

Here, LTN_i (kg) refers to the TN of LP discharged into the river; D_ij (kg·head⁻¹·day⁻¹) represents the TN discharge coefficient of LP_j (Table 2); T_j (day) refers to the average annual feeding time for LP_j (swine have a growth period of 199 days a year, 365 days for cattle, 365 days for sheep, and 210 days for poultry) [32,33].

The TN calculation formula for the inflow river generated by the daily activities of the local residents is as follows [34,35]:

$$\mathrm{R}\mathrm{T}\mathrm{N}=365\mathrm{P}\mathrm{C}\mathsf{\alpha }$$

Here, RTN (kg) refers to the TN of waste discharged by local residents into the river; P represents the number of rural people (persons); and C represents the total amount of nitrogen (g/person·day) of domestic garbage produced by one person a day.

The cumulative gray water footprint (GWF) represents the quantity of freshwater necessary to assimilate anthropogenic pollutant discharges within a standardized hydrological unit [36]. The GWF from NPSP calculation formula is as follows [37]:

$$\mathrm{G}\mathrm{W}\mathrm{F}={\displaystyle \frac{\mathrm{F}\mathrm{T}\mathrm{N}+\mathrm{L}\mathrm{T}\mathrm{N}+\mathrm{R}\mathrm{T}\mathrm{N}}{{\mathrm{C}}_{\mathrm{m}\mathrm{a}\mathrm{x}}-{\mathrm{C}}_{\mathrm{n}\mathrm{a}\mathrm{t}}}}$$

Here, GWF denotes gray water footprint (m³); C _max denotes the maximum permissible concentration threshold for drinking water quality criteria; and C_nat represents the natural background concentration of receiving water body (C_nat = 0) [38,39,40].

2.5. Determination of Weights

The main methods for determining weights include entropy weight method (EWM), analytic hierarchy process (AHP), and principal component analysis (PCA). Compared to subjective methods involving AHP, which require expert judgments, the EWM objectively derives weights from data variability, minimizing bias. Unlike PCA, EWM preserves original indicators without dimensionality reduction. In this study, weight indicators were calculated by using the EWM to reduce subjective errors. Specific calculation steps are as follows [41,42]:

(1)

Standardization of raw data:

$$\mathrm{Positive} \mathrm{indicator}\text{:} x_{ij}={\displaystyle \frac{a_{ij}-\mathrm{Min}\left\{a_{ij}\right\}}{\mathrm{Max}\left\{a_{ij}\right\}-\mathrm{Min}\left\{a_{ij}\right\}}}$$

$$\mathrm{Negative} \mathrm{indicator}\text{:} x_{ij}={\displaystyle \frac{\mathrm{Min}\left\{a_{ij}\right\}-a_{ij}}{\mathrm{Max}\left\{a_{ij}\right\}-\mathrm{Min}\left\{a_{ij}\right\}}}$$

In the formula, x_ij is the value of the water resources security assessment index, and a_ij represents the initial value of the index, where i = 1,2, …, m, j = 1,2, …, n, m is the number of evaluation indicators, and n is the number of evaluation years.

(2)

Normalization of raw data: $P_{ij}={\displaystyle \frac{x_{ij}}{\sum _{j=1}^n{x}_{ij}}}$

(3)

Calculate the index entropy: $e_i={\displaystyle \frac{-\sum _{j=1}^n{(P}_{ij}ln{P}_{ij})}{\ln n}}$

(4)

Calculate indicator weight: $W_i={\displaystyle \frac{1-e_i}{\sum _{\mathrm{i}=1}^{\mathrm{n}}{(1-\mathrm{e}}_{\mathrm{i}})}}$

(5)

Water security index: $WSI=\sum _{\mathrm{i}=1}^{\mathrm{n}}{W}_i{x}_i$

The water security index rating of the water environment is shown in

Table 3. Evaluation grades standards of water security.

Water Security Index	Security Level	Description
0.8–1	I	It indicates that water resources are highly sustainable.
0.6–0.8	II	It demonstrates that water resource allocation satisfies fundamental criteria for environmental sustainability.
0.4–0.6	III	It indicates that water resource security has reached a state of dynamic equilibrium, though this stability remains vulnerable to disruption.
0.2–0.4	IV	It indicates that while water resources are sufficient to fulfill fundamental societal demands, they fall short of supporting enduring sustainability.
0–0.2	V	It indicates that water resources cannot meet the requirements of social development

.

3. Result and Discussion

3.1. Variations of NPS Pollution Emissions in SNWDP

Figure 2 displays the quantity of TN from NPS emissions in SNWDP cities from 2000 to 2017 and the structure of NPSP. Overall, the total loss of NPSP in water source areas had shown a slow increase except for in 2007. The total loss of TN from nonpoint source pollution in 2017 was 1.50 billion tons, showing an increase of 72.45% compared to the year 2000. The pollution from the plantation industry and the source of daily life did not fluctuate significantly, showing a slow growth trend. LP pollution fluctuated significantly; it decreased sharply in 2007. From 2000 to 2003 and 2007, the largest proportion of groundwater pollution from water sources was the planting industry, accounting for 39.08–43.28%. In 2004–2006 and 2008–2017, LP breeding industry pollution became the main source of ground pollution from water sources, accounting for 39.42–49.13%. The pollution caused by living sources was relatively small. As the total load of nonpoint source pollution rises, the proportion of pollution from domestic sources continues to decrease, accounting for 11.79–20% of the total pollution.

In the past 18 years, the average pollution production of the water source area in the planting industries, LP breeding industries, and living sources accounted for 40.33%, 15.11%, and 44.56%, respectively. Living source pollution has not shown significant fluctuations; it has shown a steady increase with the increase in population. The water source planting and LP breeding industries have many influencing factors, which have a greater impact on NPSP.

The amount of pollution produced by the water source planting industry has been increasing slowly. The planting area of crops was slowly decreasing, while the amount of fertilizer, mulch, and pesticides were gradually increasing. It can be seen that the pollution load per unit of cultivated land in water source areas was increasing. The load of fertilizer application at the water source was affected by many aspects such as nature, economy, politics, technology, and tradition. Many Chinese farmers firmly believe that the increased use of fertilizers yields better crop growth. With the economic growth and the purchasing power of the farmers gradually increasing, the amount of chemical fertilizer used by the farmers also increased. Some scholars have shown that crop planting alone cannot meet the household expenditures of farmers, and the primary income of Chinese farming families is from family members who have migrated to other parts of the country in search of work. Therefore, the total crop sown area in water sources has been decreasing in recent years. From a regional perspective, there were obvious differences in the amount of fertilizers applied in areas which are more mountainous and flatter. Due to the topographical factors in Shangluo, mountainous areas have underdeveloped transportation, and economically backward farmers living in the area have lower purchasing power. Therefore, the amount of fertilizer applied per unit area was much lower than in other areas. The flat area of Nanyang City has a higher fertilizer application per unit area than other areas.

Variations in the pollution of the aquaculture industry were affected by multiple factors. In the past 18 years, the amount of groundwater pollution from water sources decreased only in 2007. Affected by avian influenza, LP breeding has decreased sharply, resulting in a lower proportion of LP pollution in 2007. In addition to the influenza virus, the main factors influencing the breeding of LP were terrain, traffic, economic level, and breeding mode. In the economically poor areas, with many mountains and underdeveloped transportation, LP breeding was less than those in other areas, such as Shangluo and Ankang. The better economic areas (Nanyang) tend to have large-scale intensive LP farming, and the number of LP farming was higher than in other areas of water sources.

3.2. Overall Safety of Water in SNWDP

The weight of each indicator calculated by the entropy weight method in each region is shown in Figure 3. There was a difference in the weight of various indicators between the whole water source area and each region, and the weight of each indicator was between 0.008 and 0.072. The driving force and response index of water resources security had the highest weights per farmer’s per capita net income (X6) and water conservancy investment (X26) (Figure 3f). With the economic development and the gradual increase in farmers’ income, water resources security in the region was facing greater pressure and impact, and increasing water conservancy investment was the main response measure to improve regional water security. The pressure, states, and impact indicators of water resources security in water source areas and between regions were different. The highest indicators of pressure and states in water source areas, Shangluo and Nanyang, were livestock breeding volume (X9) and breeding loss TN (X19) (Figure 3c,e). It can be seen that with economic growth and the increase in farmers’ income, people’s demand for meat had gradually increased, and pollution from LP breeding had placed greater pressure on the entire water source in Shangluo and Nanyang. The highest indicators of pressure and state in Ankang, Hanzhong, and Shiyan were fertilizer application (X10) and TN loss from planting (X17) (Figure 3a,b,d). Fertilizer application in Ankang, Hanzhong, and Shiyan has the greatest pressure on the water environment. The most significant indicator of water source areas in Ankang, Hanzhong, and Nanyang was the gray water footprint of NPSP (X23), with coefficients of 0.045, 0.047, and 0.056, respectively, which indicates that NPSP in these three areas was the main factor affecting the water environment. NPSP flows into water bodies through surface runoff, causing significant impacts on the water ecological security of most areas in the water source area. The area with the largest impact index in the Shangluo and Shiyan areas was the soil erosion area (X20), which indicates that soil erosion in these two areas poses a certain threat to water resources security.

From Table 3, it can be seen that, in recent years, the water safety of Shiyan and Shangluo was more stable than other regions. The water resources in Ankang, Hanzhong, Nanyang, and the whole water source are in an equilibrium phase, but this balance can easily change. Over the past 18 years, water security in cities has changed differently. Water safety in Ankang and Nanyang was more stable than in other cities, with water safety remaining at Grade III in 89% and 78% over the years (Grade III indicates that water resource security has reached a state of dynamic equilibrium, though this stability remains vulnerable to disruption). The state of water security in Shiyan City fluctuates significantly, and it has been in the state of Grade IV for 8 years, with a change interval of 0.33–0.64 (Grade IV indicates that while water resources are sufficient to fulfill fundamental societal demands, they fall short of supporting enduring sustainability). Both Hanzhong and Shangluo have been in the state of Grade IV for 6 years. The overall safety states of the water source area were relatively stable, except in 2005, 2006, and 2011, where it remained at Grade III. In general, the water safety of water sources has shown an upward trend. With the passage of time, the economy of various cities has continued to grow, and the environment has also improved.

3.3. Suggestion

Through research and analysis of water safety in the water source area of MR-SNWDP from 2000 to 2017, the development of the water source area should be improved from the following points:

(1)

From the perspective of the planting industry: This study showed that the fertilizer application rate in most areas of the water source area exceeded the national standard of 225 kg/ha [43], and there are differences in fertilizer application levels in different regions. Water source soil fertilizer content detection and management information system should be established throughout the water source area to monitor the current status of fertilizer application in the water source area in real time. To enforce compliance, village-level accountability mechanisms are critical: local collectives should maintain audited fertilizer logs, with incentives (e.g., subsidies) for villages meeting reduction targets. These measures align with the observed link between intensive management and improved fertilizer efficiency [44]. To ensure long-term impact, policies should adopt adaptive management, with biennial reviews of water quality data and farmer adoption rates.

(2)

From the perspective of the LP breeding industry: Studies have shown that due to the free-range farming of LP by farmers, LP manure was discharged in the field or concentrated on the roadside, and the manure was likely to cause nonpoint source pollution after being flushed by rain and into the river, affecting the safety of the river water quality. For large-scale farms, mandatory manure treatment regulations should be enforced, requiring covered storage facilities and regular inspections, with penalties for non-compliance. For small-scale, free-range farmers, eco-agricultural incentives are key: policymakers could subsidize transitions to circular systems like ‘rice-fish-duck symbiosis’ or ‘pig-biogas-orchard’ models [45,46], which repurpose manure as resources while reducing runoff risks. Local governments should prioritize watershed zoning, prohibiting free-range LP breeding near rivers (<500 m) and designating centralized manure collection points in vulnerable areas. Public awareness campaigns can highlight manure’s economic value (e.g., as organic fertilizer) to encourage voluntary adoption.

(3)

Water investment perspective: This study showed that water investment was the most powerful response from humans. The government should increase the construction of sewage treatment plants, so as to improve the operation quality of sewage treatment plants and the sewage and wastewater treatment rates and treatment levels; increase the control of soil erosion to reduce the loss of nutrients during rainwater erosion; with the support of construction, actively construct ponds and dams to ensure the ecological safety of the water environment in the source area; upgrade the village’s infrastructure and public service facilities, and improve the management of domestic sources of pollution in the village.

(4)

Ecological environment perspective: Increase the intensity of returning farmland to forest and grassland. Returning farmland to forest can effectively improve the level of ecological security of the water environment in the water source area. This study shows that in 2017, the forest coverage rate in Nanyang area was 39.52%, while the forest coverage rate in other areas was between 58.25 and 65.60%. Therefore, Nanyang City should be the focus of improving the overall forest coverage of the water source area.

(5)

Farmers’ perspective: Increase publicity and education. Water source farmers have unreasonable treatment methods for agricultural production activities, such as excessive application of chemical fertilizers and random discharge of animal manure. Therefore, the government should educate the farmers on the harm of popular science NPSP to reduce the source of pollution.

4. Conclusions

In this study, we analyzed the spatial and temporal characteristics of NPSP in various cities in the water source of MR-SNWDP from 2000 to 2017. In addition, the overall water security of the water source was evaluated according to five aspects: driving to force brought by the society, pressure caused by human life on the environment, state of water environment in the water source, human response to poor water environment, and impact of multiple factors of water resources on the water source. The following are the conclusions of this study:

(1) First, we evaluated the amount of nonpoint source pollution in the water source area based on the output coefficient model. The results indicated that the total loss of NPSP from water sources showed a slow growth trend except in 2007. Over the past 18 years, the proportion of water pollution from the three NPSP sources—livestock and poultry (LP) breeding industry, planting industry—and living sources has been 44.56%, 40.33%, and 15.11%, respectively. (2) NPSP threatened the water ecological security of the water source area. The largest impact indicator in the water source areas of Ankang, Hanzhong, and Nanyang is the grey water footprint of NPSP (X23), with coefficients of 0.045, 0.047, and 0.056, respectively. (3) The state of water security of the water source was evaluated from 28 indicators including nature, economy, and environment. The overall security state of the water source area was generally on the rise, and it remained at level III except for in 2005, 2006, and 2011. The state of water security in all areas except Shiyan City was relatively stable. The state of water security changes in Shiyan City was more volatile than in the other areas, of which the city was in a state of Grade IV for eight years. (4) In order to maintain water security, we suggest that the government should change the production and operation models of the plantation and LP breeding industries, increase investment in water conservancy, return farmland to forestry, and promote education. This study provides a theoretical basis for the study of water security evaluation and some control measures to reduce NPSP.

This study evaluated the water ecological security of water sources and provided a reference for the prevention and control of ecological security of water sources. While the EWM provides an objective, data-driven approach for water security assessment, it has notable limitations. The entropy weight method, while advantageous for its objectivity in water security evaluation, faces key limitations: (1) its data-dependency may overlook critical but low-variability parameters; (2) inability to incorporate expert knowledge may yield ecologically unrealistic weights. These constraints can be addressed through future research focusing on: (1) hybrid approaches integrating entropy weights with expert-based methods (e.g., AHP or Delphi), and (2) machine learning-enhanced frameworks to improve sensitivity to complex, non-linear relationships in water systems. Additionally, establishing standardized indicator selection protocols could enhance methodological robustness across different hydrological contexts.