Water Ecological Security Pattern Based on Hydrological Regulation Service: A Case Study of the Upper Hanjiang River

Abstract

Water ecological problems involve flood, drought, water pollution, destruction of water habitat and the tense relationship between humans and water. Water ecological problems are the main problems in the development of countries all over the world. In terms of ecological protection, China has put forward the ecological red line policy. Water ecology is an important component of the ecosystem, and the delineation of the water ecological red line is an important basis for ecological protection. Based on ecosystem services, this paper tries to determine the red line of the water ecology space and tries to solve various water problems comprehensively. Based on the ecosystem services accounting method, the SWAT (soil and water assessment tool) model was used to simulate the spatial–temporal dynamic quantities of water purification and rainwater infiltration services in the upper reaches of the Hanjiang River. The basin was divided into 106 sub-basins and 1790 HRUs (hydrological response units). Water quality data taken from 8 sites were used to verify the simulation results, and the verification results have high reliability. The grid scale spatialization of water quality and rainwater infiltration is realized based on HRU. The seasonal characteristics of hydrological regulation and control services were analyzed, the red line of hydrological regulation and control in the upper reaches of the Hanjiang River was defined, and the dynamic characteristics of water ecological red line were analyzed. According to the research results, the water ecological protection strategy of the basin is proposed. The prevention and control of water pollution should be emphasized in spring and summer, the prevention and control of rain flood infiltration in autumn and winter, and the normal monitoring and management should be adopted in the regulation and storage. The results of this study can provide scientific reference for water resources management and conservation policy making.

1. Introduction

Watershed ecosystem is the basic unit of inland ecosystem with unique ecological, economic and geographical pattern [1]. It is a relatively closed, well-defined whole that maintains an exchange of matter, energy, and information with the outside world [2]. The basin ecosystem provides a variety of ecological products and services for the basin residents. From the perspective of ecosystem services, water, as a necessity for the survival of organisms, not only provides supply and regulation services such as water conservation, fresh water supply, hydrological regulation and water quality for human beings [3,4]. In 2011, the Chinese government and the State Council first put forward an ecological protection strategy, clarifying the ecological red line and strictly implementing ecological protection policy guidance. The ecological red line refers to the space boundaries and management limits that need to be strictly protected in the aspects of natural ecological service functions, environmental quality and safety, and utilization of natural resources so as to safeguard national and regional ecological security, sustainable economic and social development, and people’s health. It is another “lifeline” that has been mentioned at the national level in China after the “1.8 billion mu of arable land red line” [5]. On 15 August 2023, the Blue Book of China’s Ecological Protection Red Line (2023) was publicly released. The total area of the ecological protection red line in this plan is about 3.19 million square kilometers, of which the land ecological protection red line area is about 3.04 million square kilometers (accounting for more than 30% of China’s land area), and the Marine ecological protection red line area is about 150,000 km² [6]. Through the delineation of ecological protection red lines—the most important areas of China’s natural ecosystem—the most unique natural landscape, the most essential natural heritage, and the most biodiversity are fully protected. China’s ecological protection red line policy is effective in protecting ecological resources, restoring the ecosystem, and promoting regional ecological civilization construction [7].

However, China is still faced with multiple and complex water ecological problems: floods, droughts, water pollution, loss of aquatic habitats, and increasing estrangement between people and water [8]. In addition, there are many important water source protection works, especially for the water source of China’s South-to-North Water Diversion Project, these areas have high requirements for water quantity and quality [9]. Therefore, on the basis of the ecological red line, some scholars proposed to define the boundary of a relatively complete water ecosystem and define the bottom line to maintain its safety and health, namely the water ecological space red line [10]. The basic work involves a comprehensive solution of water problems by ecological methods. Therefore, the delineation method of water ecological space red line has become a new problem. From the perspective of ecosystem services, the spatial red line of water ecology includes four aspects—water resource supply, hydrological regulation, aquatic biological support, and cultural services—among which hydrological regulation is more susceptible to the impact of human activities.

From the end of the 19th century to the beginning of the 20th century, European and American countries began to pay attention to the comprehensive management of river basin water environments, but early basin management was limited to a single goal, such as flood control, water supply, or shipping [11]. In the 1950s, river basin water pollution control and treatment gradually became an important part of river basin management. After the 1990s, it mainly aimed at the coordinated development of the river basin and emphasized the comprehensive management of the whole river basin [12]. At present, the relevant research abroad mainly focuses on the hydrological protection of disaster areas in the basin. For example, the hydrology of the Amazon Red Plain and its implications for aquatic conservation are primarily concerned with key hydrological characteristics, such as the magnitude, duration, frequency, and predictability of runoff and inundation in aquatic habitats. Remote sensing method, measurement method and modeling method of flood plain hydrology are the main methods of this kind of research [13]. US researchers employed a coupled economic-agro-ecological-hydrological modeling framework to assess the cascading impacts of climate change mitigation policies on agriculture and the resulting water quality synergies. Their analysis of a wetland conservation policy revealed that it reduced carbon emissions in the United States by approximately 50%, primarily through increased energy costs and the price of nitrogen fertilizer production. Additionally, it led to a reduction of approximately 15% in the amount of fertilizer applied to corn production in the Mississippi River basin. Simultaneously, it mitigated nitrate leaching in other areas of the basin, enhanced water quality, and achieved potential ecological effects [14]. For example, Turkey’s water policy emphasizes large-scale, centralized, technical, and supply-oriented water projects to address water allocation issues but ignores pressing issues such as unsustainable water use, impact on rural livelihoods, and institutional weaknesses in the water sector [15]. Based on the situation of water ecology abroad, it can be seen that the coordinated development of river basins in Europe and the United States started earlier, that the water environment management laws of river basins are relatively sound, and that obvious results have been achieved [16]. However, in some water-scarce areas, water resource distribution and ecological environment pressure are still faced [17]. However, the ecological red line policy is unique to China.

At present, China has accumulated considerable methodological experience in the identification and demarcation of ecological red lines. A large number of studies have taken the accurate identification of ecological red lines as the main research goal. For example, the MaxEnt model is used to predict the scope of potential high-value ecological protection land through sample training, and the contribution of common influencing factors in ecological planning is quantified [18]. In order to ensure the stability of ecological red lines, some researchers put forward a method of delineating ecological red lines based on a Bayesian network model. This method considers both ecological suitability conditions and land dynamic changes, and the model is applied in Ezhou City near the middle reaches of the Yangtze River [19]. Another study took Chishui River in Guizhou Province, China as an example. Based on the ecological red line theory and method, this study evaluated the sensitivity of karst rocky desertification, soil and water loss, importance of soil and water conservation and species diversity, searched for the most sensitive ecological area and the most important ecological function area, and carried out superposition analysis on the Chishui River liquor-production- and development-restricted area and Moutai water conservation area. The ecological red line boundary was finally determined [20]. These cases have drawn ecological red lines in the river basin, effectively improving the quality of the water ecological environment. Therefore, the management method of demarcating the water ecological red line has strong applicability in China’s policy management environment.

Drawing from this context, the study aims to establish the ecological red line for water in the upper Hanjiang River basin, ensuring the long-term sustainable development of the regional ecological environment. Leveraging the temporal and spatial quantification of ecosystem service functions, the ecological red line for the upper Hanjiang River basin in Shaanxi Province was delineated. This method is infrequently utilized in current national spatial planning, primarily due to the complexities involved in data acquisition, processing, and model verification. Considering the unique realities of the upper reaches of the Hanjiang River, this study was conducted from three aspects: water purification service, regulation and storage service, and infiltration service. The selection of these ecosystem service types is more targeted towards the water ecological red line, offering a more comprehensive approach than the conventional focus on mountains, rivers, forests, fields, and lakes in national spatial planning.

In summary, there is relatively little research on water purification services in China, most of which focuses on carbon sequestration, water conservation, and soil and water conservation. The evaluation of water purification service first involves the spatial simulation of water quality. At present, water quality simulation methods mostly use water ecological model and hydrodynamic mechanism to simulate the spatial distribution of water quality in watershed [21]. The flood control and storage function of rivers, lakes, and wetlands can reduce the threat of floods to the downstream water system and the loss caused by flooding. The regulation and storage function of the lake wetland ecosystem in the middle and lower reaches of the river will directly affect the ecological security and social and economic development of the area along the river. At present, there are more studies on flood control and storage functions of river basins [22] but few studies on regional rainwater and flood infiltration capacity, which mainly focus on urban rainwater and flood resource utilization. Improve the infiltration capacity of urban rainwater and flood by improving rainwater and flood utilization technology. At the watershed scale, there is a certain relationship between rainwater infiltration and runoff, and the rainwater infiltration capacity—including soil properties, land use types, and vegetation cover—also affects the regional hydrological regulation function.

The upper reaches of the Han River are an important water source for the middle route of the South-to-North Water Diversion Project. Some studies have found that the hydrological regulation services of the sub-basins of the Hanjiang River system are the strongest in the whole Yangtze River basin [23]. However, it is responsible for the allocation of water resources in the water-scarce areas in the north, and the protection of its water ecological environment determines the long-term benefits of the middle route of the South-to-North Water Diversion Project. Hence, this study aims to identify the water ecological red line area in the Hanjiang River basin in Shaanxi province by focusing on water purification, rain storage, and rainwater infiltration services using spatial classification and superposition, ultimately providing a scientific foundation for water source protection planning.

2. Materials and Methods

2.1. Study Site

The Han River, the largest tributary of the Yangtze River in China, originates in Ningqiang County, Hanzhong City, Shaanxi Province. It flows through Shaanxi and Hubei provinces and ultimately joins the Yangtze River at Longwang Temple, Hankou, Wuhan City [24,25]. With a total length of 1577 km [26], it boasts the largest drainage area within the Yangtze River system [27] (Figure 1).

In this study, the Hanjiang River Basin in Shaanxi Province was selected as the study area, encompassing a geographical range of 106–109.2° E and 31.9–34.2° N. The terrain in this region is complex and diverse, primarily mountainous and hilly, with intervening valley basins. The Qinling Mountains stretch from east to west, while the Daba Mountains run from north to south. The main streams of the Hanjiang River and its tributaries are deeply embedded in the bedrock, creating numerous canyons, rapids, and waterfalls. The terrain is expansive and undulating, with high and steep mountains, narrow valleys, and swift currents. The area belongs to the subtropical monsoon climate zone, experiencing four distinct seasons and characterized by concurrent periods of rain and heat. Summers are hot and rainy, while winters are mild with less rainfall. The annual precipitation is relatively abundant, favoring agricultural production and vegetation growth. Rainfall in this region is primarily concentrated in summer, especially from July to September, with frequent heavy rains that can easily trigger natural disasters such as flash floods and mudslides. Winters bring less precipitation, and the climate is relatively dry. The soil types are diverse, mainly consisting of yellow brown soil, brown soil, and lime soil. Vegetation coverage is high, dominated by forest and shrubland, and the region is rich in biodiversity, harboring many rare wildlife resources [28].

The study area is not only an important water source for the Middle route of China’s South-to-North Water Diversion Project but also a water intake point for the “YinHanJiWei” Project in Shaanxi Province and a major infrastructure construction project in the Guanzhong-Tianshui Economic Zone planning approved and promulgated by the State Council [29]. The total area affected by water is 14,000 km², benefiting 14.11 million people [30,31]. Therefore, it is particularly important to delineate the red line of water ecology space in the upper Hanjiang River basin in Shaanxi province and formulate its protection policy.

2.2. Methods

2.2.1. SWAT (Soil and Water Assessment Tool) Model Construction

The Soil and Water Assessment Tool, known as the SWAT model, was developed in 1994 by Dr. Jeff Arnold of the Agricultural Research Center of the United States Department of Agriculture (USDA) [32]. The model is a distributed basin hydrological model based on GIS (Geographic Information System), which is mainly used to predict the long-term effects of land management on water, sediment and chemicals in large basins with complex and variable soil types, land use patterns and management practices [33]. The basin is divided into subbasins and further subdivided into HRUs within each sub-basin, each of which has the same range of land use, soil type and slope. The SWAT model can simulate a variety of hydrophysical and chemical processes, including water quantity, water quality, soil erosion, plant growth, nitrogen and phosphorus, and pesticide migration and transformation. The model is widely used in hydrology, water resources management, soil erosion, agricultural chemical loss, water quality evaluation and other fields [34]. Modeling the impacts of different land use and management options on water resources and soil erosion to provide a scientific basis for decision makers.

This study adopts SWAT model to simulate the water quality of the study area. The water quality simulation principle of SWAT model is mainly based on the conceptual model of physical process, which can simulate the change process of various water quality parameters in the basin [35]. This includes the transport and conversion of nutrients (such as nitrogen and phosphorus), pesticides, heavy metals, and other pollutants. The model evaluated the impact of non-point-source pollution on water quality by simulating the complete cycle of nutrients such as nitrogen and phosphorus and the degradation process of pesticides in the hydrologic response unit [36]. Figure 2 shows the loop diagram of N and P simulation in SWAT model. The migration and transformation of nutrients such as nitrogen and phosphorus depend on the transformation process that the compounds undergo in the soil environment. The SWAT model monitored five different nitrogen pools in soil: inorganic nitrogen pools (NH⁴⁺ and NO³⁻) and organic nitrogen pools (new organic nitrogen, active organic nitrogen, and stable organic nitrogen) (Figure 2). Phosphorus in mineral soil mainly exists in three forms: organic phosphorus in humus, insoluble mineral phosphorus, and phosphorus available to plants in soil solution. In oxygen-containing water, nitrogen is transformed step by step, from organic nitrogen to ammonia nitrogen to nitrogen, nitrite nitrogen, and finally nitric nitrogen. Organic nitrogen, ammonia nitrogen, nitrite nitrogen and nitric nitrogen are removed from the channel by sedimentation and sediment adsorption. The phosphorus cycle is similar to the nitrogen cycle. After the algal body dies, the phosphorus in the body is converted into organophosphorus. The organophosphorus is then mineralized into dissolved phosphorus that can be absorbed by algae. Organophosphorus may also be removed from rivers by sedimentation.

The operation of SWAT model requires a variety of input data, including topographic data (such as DEM), land use data, soil data, and weather data (such as daily precipitation, air temperature, relative humidity, solar radiation, wind speed, etc.). The input data in this study area are shown in

Table 1. SWAT model input data list.

Data	Data Source	Data Attribute
DEM (Digital Elevation Model)	China geospatial data cloud (https://www.gscloud.cn/)	Raster data; The spatial resolution is 90 m × 90 m
Soil type and soil attribute data	World soil database (HWSD)	Raster data; Spatial resolution 1000 m × 1000 m; Archive
LUCC data for 2015	Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences	Raster data; Spatial resolution is 30 m; Data sheet
Meteorological data (temperature, precipitation and wind speed)	National Center for Meteorological Data Science (http://data.cma.cn/)	Daily meteorological data (unit: °C, mm, m/s)

.

2.2.2. Verification of Model

The study area was divided into 106 subbasins and 1790 HRU (hydrologic response units) using the threshold recommended by the model. After the successful construction of the model, it is necessary to use the measured hydrological or water quality data for calibration and verification. Water quality data were obtained from eight stations in the upper reaches of the Hanjiang River provided by Shaanxi Provincial Environmental Protection Department (Figure 3).

The sensitivity analysis, calibration, uncertainty analysis and result verification of the model parameters are mainly implemented by SWAT-CUP2012 software. The verification methods in SWAT-CUP software mainly include para sol (parameter solving), PSO (particle swarm optimization), GLUE (generalized likelihood uncertainty estimation), Sufi-2 (sequence uncertainty fitting version 2), and other algorithms [37]. The software is easy to operate, and after iterating repeatedly to reduce the value range of the selected parameters, the best simulation effect of the parameters is finally obtained. This method greatly improves the efficiency of model parameter sensitivity analysis and verification.

SWAT-CUP sensitivity analysis is an important tool to improve the calibration efficiency and accuracy of SWAT model. Through sensitivity analysis, the parameters that have the most significant impact on the model output can be identified, so as to optimize the setting of model parameters and make the model more in line with the actual situation. In this study, SWAT-CUP software is used to analyze the sensitivity of 8 parameters that have a great influence on the model. Since runoff and water quality data are mainly obtained, sensitivity parameters related to runoff channels are selected (

Table 2. Sensitivity parameter calibration.

ID	Parameter Name	Parameter Definition	Min	Max	Rating Value
1	r__CN2.mgt	Runoff curve	−0.2	0.2	−0.08
2	v__ALPHA_BF.gw	Base flow regression coefficient	0.0	1.0	0.7
4	v__GWQMN.gw	Basal flow level threshold	0.0	2.0	1.0
5	v__CH_N2.rte	Manning coefficient of main channel	0.0	0.3	0.19
6	v__CH_K2.rte	Effective water conductivity of the main channel	0	3500	2000
7	SURLAG	Lag coefficient of surface runoff	−0.8	0.8	0.5
8	LAT_TTIME	Running time of soil flow	0	100	5

). We continuously adjust the parameters to get better verification results, and eventually build the final model with the parameters that achieve the best simulation results. After adjusting the model parameters, the model is run, the boundary conditions are adjusted, and the output results are compared. The final values of the parameters in the model are shown in Table 2.

Based on monthly N and P concentration data collected from four water quality stations in the main stream of the Hanjiang River from 2012 to 2019, the study designated 2012–2013 as the warm-up period, 2014–2016 as the normal rate period, and 2017–2019 as the verification period. The output results of the SWAT model were analyzed, calibrated, and verified. In the SUFI-2 algorithm, the p-factor and r-factor are used to quantitatively evaluate the uncertainty of the model. The standard range of the p-factor value is between 0 and 100%, while the standard value range of the r-factor is between 0 and infinity. The closer the p-factor is to 1 and the closer the r-factor is to 0, the better the simulation effect is. The validation results were evaluated using the R² and Nash–Sutcliffe efficiency coefficient (ENS) to assess the simulation performance of the SWAT model. R² indicates the consistency of the trend between simulated and measured values, with values closer to 1 indicating a higher consistency in trend between simulated and measured values. The ENS efficiency coefficient represents the degree of deviation between measured and simulated values, with values closer to 1 indicating that the simulated values are closer to the measured values [38]. It is generally believed that when the value of the Nash–Sutcliffe efficiency coefficient (ENS) is greater than 0.5, the simulated values of the model have a good correlation with the observed values. The results show that the simulated values at the four stations fit well with the measured values of N and P concentrations (

Table 3. The evaluation index of TN (total nitrogen) and TP (total phosphorus) simulation results of each station.

	Rate Periodicity				Validation Period
	p-Factor	r-Factor	R2	NS	p-Factor	r-Factor	R2	NS
	TN (total nitrogen)
Nanliudu	0.59	0.00	0.73	0.68	0.69	0.00	0.70	0.67
Laojunguan	0.66	0.00	0.67	0.65	0.68	0.00	0.69	0.65
Liejinba	0.60	0.00	0.65	0.65	0.70	0.00	0.68	0.66
Xiaogangqiao	0.62	0.00	0.70	0.67	0.69	0.00	0.72	0.68
	TP (total phosphorus)
Nanliudu	0.69	0.00	0.68	0.66	0.69	0.00	0.69	0.67
Laojunguan	0.58	0.00	0.87	0.86	0.58	0.00	0.85	0.74
Liejinba	0.65	0.00	0.85	0.84	0.65	0.00	0.80	0.77
Xiaogangqiao	0.77	0.00	0.71	0.68	0.77	0.00	0.77	0.69

).

2.2.3. Representation Method of Water Purification Service in River Basin

The water purification service level is usually characterized by the output or retention of total nitrogen per unit area of ecosystem. In this paper, the reciprocal of total nitrogen output per unit area of ecosystem (kg/hm²) is selected as the scale of water purification service, that is, the smaller the total nitrogen output per unit area of ecosystem, the higher the level of water purification service.

Q = 1/N

(1)

N is the total nitrogen output of the ecosystem per unit area, and Q is the water purification service of the basin.

2.2.4. Methods for Obtaining Service of Storage Regulation and Rainwater Infiltration

Flood control and storage service plays an important role in regional ecological security. At the same time, the flood control and storage area is also the supply area of main aquatic products. Ecosystems with the capacity to regulate and store flood water include lakes, reservoirs, ponds, and swamps. The service capacity of flood control and storage mainly depends on the capacity of lakes and reservoirs. The larger the volume, the stronger the flood control and storage capacity. To evaluate flood regulation and water storage services in the study area, the authors utilized Landsat-TM image data, integrated the global reservoir and dam dataset, and employed Google Earth on the ArcGIS platform. This allowed them to interpret the reservoir morphology of the upper reaches of the Han River and subsequently create a reservoir distribution map for the upper reaches of the Hanjiang River. The water storage data of reservoirs in the basin from 1999 to 2018 were obtained through the Water Resources Bulletin of the upper reaches of the Hanjiang River.

The highly suitable infiltration area of hydrogeological conditions such as soil type and soil texture is an important part of the regulation of groundwater level. In this paper, the relationship among infiltration, precipitation and surface runoff in SWAT model is used to realize the spatialization of rainwater infiltration service in watershed (Figure 4). In the SWAT model, rainwater infiltration is precipitation minus surface runoff. This model can simulate surface runoff and obtain spatial data of rainwater infiltration.

2.2.5. Red Line Acquisition Method of Watershed Ecohydrological Regulation

Based on the spatial distribution of water quality purification, rain flood regulation, and storage and flood infiltration services, this paper classifies the three services into safe, relatively safe, and unsafe areas according to the environmental quality standards of surface water. The three ecological services have equal importance and weight. The lakes, reservoirs and other areas with water storage function in the basin are directly used as the protection areas for hydrological regulation, and for flood infiltration services. The three security modes are superimposed, and the maximum value is obtained by disjunctive operation. Finally, the water ecological security pattern of the upper reaches of the Hanjiang River was established. The safe, safer, and unsafe areas of rainwater infiltration service can be obtained according to the percentage of infiltration in precipitation as the standard of safety classification. Since the essence of the ecological protection red line is the fundamental threshold for ecological and environmental security, unsafe areas within the water purification service zone, hydrological regulation protection area, and rainwater infiltration service zone can be overlaid to determine the ecological and hydrological regulation flow red line area in this region (Figure 5).

3. Results

3.1. Analysis of Water Quality Simulation Results

Based on the SWAT model, the spatial distribution of total nitrogen and total phosphorus on the HRU scale was obtained (Figure 6). From the simulation results, TN content is obviously higher than TP content. The highest total phosphorus content per unit pixel was 4.646 mg/L in summer, followed by spring, and the lowest in winter. The highest total nitrogen content per unit pixel was 32.954 mg/L in summer, followed by that in spring and the lowest in winter. Spatially, TN concentration is more dispersed than TP concentration in the study area. The spatial distribution of N/P content also showed obvious seasonality. The areas with higher total phosphorus content in summer were mainly distributed in the western part of the study area; the areas with higher total phosphorus content in spring were mainly in the central part of the study area; the total phosphorus content decreased significantly in autumn, mainly distributed in the southeastern part of the basin; and the total phosphorus content decreased to the lowest in winter. The total nitrogen content is spatially most dispersed in spring, mainly in the central part of the basin; the areas with high total nitrogen content in summer are mainly in the northern part of the basin; the areas with high nitrogen content in autumn are mainly distributed in the southeastern part of the basin; and the high nitrogen content in winter is mainly in the northern part of the basin.

According to the spatial distribution data of TN/TP, the pixel percentage under each water quality standard from January to December was extracted. The distribution diagram of pixel percentage under different water quality standards was obtained each month (Figure 7). In Figure 8, the dark red represents the relatively high pixel proportion, the green represents the relatively low pixel proportion, and the yellow represents nearly 50%.

According to the provisions of the “Surface Water Environmental Quality Standards” (GB3838-2002), China’s surface water is divided into five categories according to its use: Class I water quality: good water quality. The ground water only needs to be disinfected, and the surface water can be supplied for drinking after simple purification treatment (such as filtration) and disinfection; Class II water quality: water quality is slightly polluted. After routine purification treatment (such as flocculation, precipitation, filtration, disinfection, etc.), the water quality can be used for drinking; Class III water quality: suitable for secondary protection areas, general fish protection areas, and swimming areas of centralized drinking water sources; Class IV water quality: suitable for general industrial protection areas and recreational water areas where the human body is not in direct contact; Class V Water quality: suitable for agricultural water use areas and general landscape waters. Water bodies exceeding the five water quality standards have basically no use function.

As evident from Figure 7, the water quality is superior in January, February, November, and December, with a higher proportion of pixels meeting the Class 1 standard, exceeding 90%. Conversely, in July, the water quality is relatively poor, characterized by a higher percentage of pixels failing to meet the standard. Specifically, pixels with N content and P content did not meet the standard account for 54.86% and 44.75%, respectively. This observation underscores that the water quality in the basin exhibits distinct seasonal variations, being inferior in summer and superior in winter.

3.2. Water Purification Service and Safety Classification

A Water quality model was used to simulate the migration process of non-point source pollution and the influence of land use type on non-point source pollution. According to the water environmental functions and water quality objectives of rivers and lakes, the scope and protection level of buffer zones are determined.

On the basis of the water quality spatial simulation results, the water quality service spatial distribution map in the study area was obtained according to the water quality service formula (Formula (1)). The spatial security pattern of water quality services on a seasonal scale was obtained by dividing the safety levels of water quality services at seasonal scale according to surface water environmental quality standards (Figure 8).

According to the environmental quality standards of surface water, a water quality below III is poor and cannot be used as a drinking water source. Therefore, water quality areas above Grade III are classified as safe areas, those from Grade III to Grade V are relatively safe areas, and water quality areas above Grade V are classified as unsafe areas. Therefore, the unsafe areas of TN and TP overlap as water quality service areas.

As can be seen from Figure 8, the area of unsafe water quality is large in spring and summer, significantly reduced in autumn, and there is basically no water quality problem in winter. Water quality insecurity areas are mainly in the southern and central regions in spring, mainly in the northern, western and southeastern regions in summer, and mainly in the southeastern region in autumn. Therefore, in terms of water quality monitoring, the areas that need attention during different seasons are also constantly changing.

3.3. Flood Regulation and Storage Service in the Upper Reaches of Hanjiang River

Through image interpretation and water resources bulletin data, the spatial distribution of reservoirs and lakes in the basin of the study area (Figure 9) and the temporal change of water demand in the reservoir (Figure 10) were obtained. Due to the slow change in reservoir scope, the nearest reservoir distribution is divided into the flood control area and the storage service protection area.

The large lakes and reservoirs in the study area include Yinghu Lake, Hanjiang Reservoir, Hanshuiyuan National Wetland Park and Hanjiang Three Gorges. Located in the main stream of the Hanjiang River, Yinghu was formed when the Han River was impinged by the construction of Ankang Hydropower Station in the 1990s. It is the largest reservoir lake in Shaanxi Province, and its inflow accounts for more than 60% of the total inflow of Danjiangkou Reservoir. Hanjiang Reservoir is located 1.5 km away from Shiquan County, Shaanxi Province. It was formed in 1973 with the completion of the Shiquan Hydropower Station. The reservoir area spans three county boundaries, with a water surface area of 0.14 × 10,000 hectares and a storage capacity of 470 million m³. Hanshuiyuan National Wetland Park is located in Hanzhong City, Shaanxi Province and is a national wetland park located in the upper reaches of the Hanjiang River. It is an important water supply area and collection area of the Middle route of the national South-to-North Water Diversion project. The Three Gorges of the Hanjiang River are located in Shiquan County, Ankang City, Shaanxi Province, with a total length of 33.9 km. These reservoirs and lakes play an important role in regulating water flow and storing water for irrigation.

The water storage capacity of the reservoirs in the study area increased significantly after 2014, and in 2017, the water storage capacity reached the highest value in 20 years. This also shows that the reservoir storage service in the basin is gradually increasing.

3.4. Rain and Flood Infiltration Service in the Upper Hanjiang River Basin and Safety Classification

According to the method in Section 2.2.4, the SWAT model is used to obtain the spatial distribution map of rainwater infiltration service volume (Figure 11). It can be seen from the results that the amount of rainwater infiltration is large in spring and summer but relatively small in autumn and winter. The region with high infiltration in spring and summer has a large distribution area, and the northern region has the largest infiltration amount, while the region with high infiltration amount in autumn is significantly reduced. In winter, except for the northern region, the infiltration amount in other regions is basically 0.

According to the following rules, 10% of the maximum rainfall in the basin is taken as the safety standard, and the average rainfall on the scale of HRU in the basin is taken as the relatively safe standard. If the rainfall is less than this value, it is an unsafe area. From the perspective of spatial security pattern of rainwater infiltration services in four seasons, the vulnerable periods of rainwater infiltration services are mainly in autumn and winter, and the unsafe areas account for 54.33% and 79.22%, respectively.

3.5. Safety Pattern Identification of Hydrological Regulation Service

According to the method in Section 2.2.5, unsafe areas of three types of services are superimposed to obtain the red line range of water ecological space (Figure 12). As can be seen from the figure, the three types of services show obvious changes in season and space. The unsafe range of water purification is large in spring and summer, and the water purification service area in spring and summer accounts for more than 50% of the total area, and there are still water purification service areas in the southwest of the basin in autumn. Hydrological regulation and storage services changed little. The amount of rainwater infiltration is larger in the unsafe area in autumn and winter, which enlarges the ecological red line range of rainwater infiltration. Therefore, the protection of water pollution and rainwater infiltration in key areas should pay attention to its time rule. In the spring and summer, managers need to focus on water quality protection, and in the fall and winter, because of the decline in rainwater infiltration, they need to pay attention to rain and flood damage.

4. Discussion

4.1. Research Objective and Feasibility Analysis

The main research objective of this paper is to explore the method of water ecological spatial red line demarcation based on ecosystem service functions, and to conduct an empirical study on the upper reaches of the Hanjiang River Basin. By selecting the hydrologic regulation service branch, we verified the feasibility of the water ecological spatial red line framework system proposed by Yu Kongjian. As hydrological regulation services are more vulnerable and difficult to recover than other ecosystem services, such as supply services, security services, and cultural services, redlining them is particularly critical [39,40,41]. This study delimits the ecological red line according to the “Ecological Service Law” and provides a new idea and method case for the research in related fields. Compared to the previous ecological red line delineation methods, the results of this study are more targeted and practical. However, the process and method in this study are mainly applicable to large and medium-sized basins, and their applicability to small basins needs to be further verified [42,43].

4.2. Analysis of Research Results

Using the SWAT model, we calculate the spatial and temporal distribution of water purification services, hydrological storage, and rainwater infiltration services in the study area. The results show that the water purification service and rainwater infiltration service have obvious seasonal characteristics, in which the weak period of water purification service is mainly in spring and summer, while the weak period of rainwater infiltration service is in autumn and winter. Through communication with peer experts, we believe that the poor water quality in spring and summer is mainly related to the climatic characteristics of summer rainstorm, and the enhancement of human activities in summer also affects the water quality of the basin to a certain extent. The seasonal variation of rainwater infiltration service function is mainly related to soil properties and vegetation defoliation.

4.3. Comparison with Previous Studies

In recent years, quantitative studies on water quality services have gradually increased, but most of them use the InVEST model. For example, the InVEST model was used to quantitatively assess the characteristics of soil conservation and water purification services and their temporal and spatial changes in the ecological zones of the Luanhe River Basin in 2005, 2010, and 2015. Moran’s index was used to analyze the mechanism of action of the two services and the spatio-temporal differentiation of trade-offs and synergies, and the effects of climate and land use change on the two services were discussed [44]. However, there is no detailed spatial and seasonal analysis. In addition, some studies use the InVEST-Nutrient Delivery Ratio (NDR) model to simulate the time change trend of water purification service functions in the watershed from 1975 to 2020. The effects of land use, landscape pattern and other natural and social economic factors on water purification services were investigated by GIS spatial statistical analysis, landscape pattern analysis and geographic detector. The spatial distribution characteristics of water quality in 2020 were analyzed, but the seasonal variation of water quality was not analyzed [45]. In summary, in recent years, most studies on the simulation of water quality services have adopted the InVEST model [46,47]. Although the model is simple to use, its results rely more on land use and soil data and have been less studied at seasonal scales. In contrast, the simulation results of the SWAT model are relatively more accurate, which can realize the data on the HRU scale, and the time can also be accurate to the month. This makes the SWAT model have a higher advantage in simulating the water quality of the basin.

5. Conclusions

In summary, delineating ecological red lines is an effective method of ecological protection planning in China, and it is more accurate and scientific to determine the area of ecological red lines based on the quantitative classification results of ecosystem service functions [48]. Taking hydrological regulation as an example, this study quantified the spatial and temporal distribution of hydrological regulation services in the upper reaches of the Han River based on the SWAT model and obtained the red line of hydrological regulation services in the basin. This provides a scientific approach to delineate the water ecological spatial red line based on ecosystem services. At the same time, we also found that there are obvious seasonal changes in the water ecosystem service functions, so the management should be adjusted according to the seasonal changes in the basin ecological protection process.

In future studies, we can further explore how to better integrate seasonal factors into the delineation of ecosystem service function red lines. In addition, you can try to apply the SWAT model to other types of ecosystem services to verify its universal applicability. At the same time, we also need to pay attention to the uncertainties and limitations of the model, such as the inaccuracy of data input and the sensitivity of model parameters, in order to better understand and evaluate the results of this study.