Water pollution is a globally significant issue. Water environment simulation is an important tool to study water pollution. In order to investigate the impact of hydrological periods, water quality indicators and cross-sections on the water environment capacity, we took the Luo River in Luoyang City as an example, and used MIKE21-coupled hydrodynamic and water quality model and the segmentation method to calculate the water environment capacity of COD, TP, and NH3-N in wet, normal, and dry periods under L1 (Gao Yazhai-Lilou), L2 (Li Lou-Baimasi), and L3 (Baimasi-G207 Highway Bridge) section. The results of the study showed that from the hydrological period, the water environmental capacity was the largest in the wet period due to the dilution effect caused by higher precipitation; from the water quality indexes, the COD concentration was the largest due to industrial pollution, but the higher degradation coefficients result in the largest water environment capacity for COD; from the cross-section, L1 has the longest channel, thus the highest capacity of the water environment was L1. This result can broaden ideas for the application of the MIKE21 model, provide a basis for the calculation of the dynamic water environment capacity and water pollution control in the river.

  • Innovative hydrodynamic water quality model evaluates Luo River pollution across diverse flow scenarios.

  • Dry periods exhibit poorest water quality, while wet periods showcase the best.

  • Identified sections exceeding capacity emphasize pollution management during distinct hydrological conditions.

In recent years, water pollution has emerged as a crucial global concern, presenting potential hazards to human health and economic progress, while also negatively impacting the balance of the ecological-social-system (Tang et al. 2022). The close relationship between water quality issues and hydrological as well as hydrodynamic processes is widely acknowledged. Hence, a comprehensive understanding of hydrological, hydrodynamic, and water quality processes, monitoring water pollutants, along with the establishment of a water environment model for assessing watershed water conditions, holds paramount importance for water management and ecological initiatives (Fatahi Nafchi et al. 2021; Sun et al. 2021). Approximately 85% of urban rivers in China are contaminated. The primary factors include household sewage, industrial wastewater, rainwater, and runoff (Zhao et al. 2019; Fu et al. 2023). River contamination will directly jeopardize human health and impede agricultural and fishery activities (Kumar et al. 2021). Additionally, water pollution can contribute to water scarcity. Thus, conducting water environment simulations for rivers, investigating their spatial and temporal distribution characteristics and evolutionary trends, and analyzing the underlying causes can furnish a theoretical foundation for managing water pollution in the environment and for the utilization and safeguarding of water resources (Żelazny et al. 2023).

Water environment simulation involves gathering hydrological, meteorological, pollution source, and runoff data, and utilizing specific numerical simulation techniques aligned with research objectives to achieve a more precise emulation of actual water flow and pollutant transport (García-Alba et al. 2019). This primarily encompasses hydrodynamic and water quality simulations. The natural water cycle constitutes an intricate interplay of physical and chemical processes influenced by a multitude of factors. Through river water quality monitoring and assessment, coupled with analysis of temporal-spatial pollutant distribution traits and evolutionary trends, a theoretical foundation for river basin water pollution management can be established (Deng et al. 2021). Given the expansive scale of natural rivers, studying them solely through direct measurements proves highly challenging. Recently, the fusion of simulation and measurement has gained prominence as the prevailing approach in contemporary research (Goliatt et al. 2021; Montazeri et al. 2023). Regan et al. (2019) utilized applications of the National Hydrologic Model (NHM) to offer insights for enhancing efficiency in water resources planning and management. Ghunowa et al. (2021) developed a spatial decision support system for Stream Power Index for Network (SPIN) toolbox, which can be used to evaluate the stability of urban rivers at the basin scale. Suriya & Mudgal (2012) took the Daya River Basin as an example to simulate the water environment and established a flood management system for the basin, which provided strategic support for water environment management in the basin. Son et al. (2020) used a combination of water quality indices (WQIs) and pollution indices (PIs) to assess riverine water quality in the Cau River to better understand water conditions in the river could be an option for water use decision-making. Research into river water environment simulation frequently demands substantial experimental data backing, and obtaining such data is restricted by diverse factors, including access permissions, data latency, and delays in analytical outcomes. The procurement of experimental data frequently entails substantial investments of human, material, and financial resources. Therefore, the goal of constructing a mathematical water environment model is to circumvent these challenges and minimize expenditure while attaining optimal economic gains.

Mathematical numerical models utilize fundamental equations to tackle a range of complex problems that are challenging to address through conventional means, making them prevalent in water environment simulation research (Paniconi & Putti 2015). The evolution of fluid dynamics mathematical models has undergone substantial refinement and enhancement, enabling them to accurately depict water body flow conditions across diverse operational circumstances (Ghidossi et al. 2006). By configuring hydraulic parameters and fine-tuning input model data, encompassing factors like rainfall, evaporation, tides, and wind patterns, the modeling process strives to closely emulate real-world scenarios (Michot et al. 2011). The hydrodynamic model employs numerical discretization techniques and governing equations to solve water flow, exploring distinct water pathways in varying settings. It significantly contributes to analyzing fluid motion patterns in rivers across different circumstances owing to its heightened accuracy and practicality (Hartmann et al. 2014). Water quality models are employed to depict the patterns of pollutant dispersion in rivers, frequently utilized in researching water pollution, self-purification capabilities, water quality prediction, pollution control, sewage management, and more (González et al. 2014). The MIKE model, a water environment simulation system devised by the Danish Institute for Water and Environment Research, integrates hydrodynamic and water quality modules to simulate and analyze one-dimensional, two-dimensional, and three-dimensional hydrodynamic and water quality states within a watershed (Ji et al. 2022). This model finds extensive application in water environment simulation.

In this study, the MIKE21 water environment model was employed to simulate hydrodynamic and water quality conditions in the Luo River of Luoyang City, serving as an illustrative case. Through refinement of input model boundary conditions, the simulation results exhibited reduced divergence from monitoring data. Subsequently, analysis of the spatial and temporal distribution of water quality variables was conducted. Optimization of pertinent parameters, including diffusion and degradation coefficients, was carried out, followed by an assessment of the plausibility of simulation outcomes. Ultimately, the water environment's capacity and the extent of reduction or remaining capacity in the Luo River across distinct hydrological periods were computed.

The objective of evaluating water environment capacity across varying hydrological phases is to gauge the water system's capability to absorb pollutants and uphold acceptable water quality benchmarks. Acquiring knowledge of water environment capacity empowers researchers to establish water quality thresholds or standards tailored to distinct hydrological periods. Simulating the spatial and temporal evolution of pollutants provides a scientific basis for the calculation of the water environmental capacity, and provides a new idea of thinking for the calculation of the environmental capacity of the river and the water pollution control of pollutants in different hydrological periods. At the same time, this information facilitates the establishment of pragmatic goals for pollutant reduction and offers direction in formulating strategies for water management.

Study area

The Yiluo River Basin comprises the Yi River and the Luo River, constituting a prototypical dual river system. The Luo River serves as the main stem of the Yiluo River, while the Yi River constitutes its primary tributary. The confluence of these two rivers forms the Yiluo River, unifying the Yi and Luo Rivers as integral components. Situated within the Sanmenxia-Huankou region, the Yiluo River Basin is positioned in the middle section of the Yellow River Basin, spanning longitudes 109°17′ to 113°10′E and latitudes 33°39′ to 34°54′N. Encompassing an expanse of 18,881 km2, the Yiluo River Basin comprises 2.37% of the overall Yellow River Basin area. Within the Yiluo River Basin, the region overlapping with Henan Province spans 15,818 km2. Among this, hilly terrain encompasses 14,236 km2, constituting roughly 90%, while the remaining flatland occupies 10%. The geographical arrangement of the Yiluo River Basin, hydrological stations, monitoring sections, and monitoring points is illustrated in Figure 1.
Figure 1

Map of the river system and hydrological station network in the study area.

Figure 1

Map of the river system and hydrological station network in the study area.

Close modal

Data sources

The climatic data such as daily precipitation, evaporation, wind direction, wind speed, discharge, and velocity of the Yiluo River Basin were obtained from the Yellow River Water Network (http://www.yrcc.gov.cn/). The topographic elevation data were obtained from the geospatial data cloud (https://www.gscloud.cn/) with a spatial resolution of 30 m. Monthly water quality data of the Luo River in Luoyang City, including COD, NH3-N, TP and DO, were obtained from the Luoyang Environmental Monitoring Monthly Report and the Henan Environmental Bulletin.

Model principle

Hydrodynamic model

The hydrodynamic module in MIKE21 utilizes terrain meshing and incorporates factors such as bed roughness, boundary conditions, precipitation, evaporation, wind patterns, ice coverage, eddy viscosity coefficients, and Coriolis force (Wood et al. 2023). These inputs are employed to simulate water levels, flow velocities, and various other elements of the flow field. This methodology finds extensive utility in simulating two-dimensional shallow water flows with open surfaces.

The equations governing the hydrodynamic module are presented as follows (Li et al. 2020):
formula
(1)
formula
(2)
formula
(3)
where Pa is the local atmospheric pressure, t is the study time, x and y are the Cartesian coordinates, η is the surface elevation, d is the still water depth, h=η+d is the total water depth, u and v are the flow velocity components, f is the Coriolis parameter. The formula is , is the angular velocity of the Earth's rotation, is the geographical latitude, is the density of water, , and are the radiation stress components respectively, s is the point source term, and are the source term water flow rate, , , and are different shear stresses in multiple orientations, A is the section area, and is the three horizontal viscous forces, these expressions are:
formula
(4)
formula
(5)
formula
(6)

Water quality model

The basis of the water quality equation is established on the mass balance equation. The two-dimensional water quality equation is (Edwards et al. 2002):
formula
(7)
where is scalar variable concentration, is horizontal diffusion coefficient, t is time, is degradation factor of scalar.
formula
(8)
where c is concentration of ECO Lab state variable, u, v, and w are flow velocity components of the convection term, Dx, Dy and Dz are dispersion coefficient of the diffusion term, Sc is source-sink term, Pc is biochemical reactions of Eco Lab.

Model steps

The selection of the MIKE21 model is attributed to its widespread acknowledgment and extensive utilization in water resource management and environmental modeling endeavors. Its versatility, adaptability, and customizable attributes render it appropriate for a multitude of research objectives and site-specific circumstances. The model's robust numerical algorithms guarantee precise and efficient simulations, even when confronted with intricate and dynamic surroundings. Furthermore, the user-friendly interface and comprehensive post-processing tools simplify setup, visualization, and result analysis procedures. The substantial validation and endorsement from the scientific community additionally bolster its credibility. In summary, the MIKE21 model delivers reliability, adaptability, and computational efficiency, rendering it a suitable option for studying water quality dynamics and guiding decisions in water resource management.

The computational procedures of the MIKE21 water environment model are bifurcated into two segments: the hydrodynamic (HD) model and the water quality (Ecolab) model. The process of simulating the water environment is illustrated in Figure 2.
Figure 2

Water environment simulation process.

Figure 2

Water environment simulation process.

Close modal

Establishment of the hydrodynamic model

Grid terrain

A pivotal preliminary procedure prior to constructing a hydrodynamic model involves generating a mesh topography, wherein the mesh quality significantly influences the model's functionality (Gayathri et al. 2016). Occasionally, a mesh lacking in precision proves inadequate for sustaining accurate model computations, thereby resulting in model dispersion and non-convergence. The simulation employs the triangular unstructured mesh division technique. The mesh generation yields a cumulative node count of 18,861, encompassing 36,501 individual mesh segments. Generally, a minimum angle not less than 30° in the unstructured grid is deemed indicative of superior grid quality. Figures 3 and 4 illustrate the river boundary grid and grid topography interpolation.
Figure 3

River boundary grid.

Figure 3

River boundary grid.

Close modal
Figure 4

Grid topography interpolation.

Figure 4

Grid topography interpolation.

Close modal

Boundary processing

The extensive monthly flow dataset from the Baimasi hydrological station spanning 1960–2020 is adopted as the foundational data for this study. The temporal progression of monthly flow at the Baimasi station is depicted in Figure 5, effectively illustrating the annual variation in Luo River's water levels, encompassing periods of wet, normal, and dry.
Figure 5

Flow distribution from 1960 to 2020 at the Baimasi hydrological station.

Figure 5

Flow distribution from 1960 to 2020 at the Baimasi hydrological station.

Close modal

Based on the multi-year long time series measured monthly flow data from 1960 to 2020 at Baimasi hydrological station, the P-III curve is drawn, and the measured flow data are allocated to determine a total of three calculation conditions, which are P = 10%, P = 50% and P = 90% of the annual allocation flow. The calculation conditions are shown in Table 1.

Table 1

Calculation conditions

Condition settingTypeFlow rate under guarantee (m3/s)Typical representative year
Condition 1 Wet year (P = 10%) 29.63 2010 
Condition 2 Normal water year (P = 50%) 16.25 2013 
Condition 3 Dry year (P = 90%) 6.42 2020 
Condition settingTypeFlow rate under guarantee (m3/s)Typical representative year
Condition 1 Wet year (P = 10%) 29.63 2010 
Condition 2 Normal water year (P = 50%) 16.25 2013 
Condition 3 Dry year (P = 90%) 6.42 2020 
Table 2

Calibration results of model parameters

Status variablesParameterValue
COD Adsorption coefficient (/day) 0.004 
Semi-saturated oxygen concentration (mg/L) 1.9 
20°C degradation rate coefficient (/day) 0.003 
Temperature coefficient of degradation rate 1.07 
NH3-N Nitrogen uptake semi-saturation (mg/L) 0.045 
BOD decay release ammonia ratio (g NH3-N/g BOD) 0.25 
Temperature coefficient 1.092 
COD degradation of ammonia nitrogen ratio (g NH3-N/g COD) 0.25 
DO The sediment requires oxygen (/m2)/day 0.4 
Sediment aerobic temperature coefficient (/day) 1.07 
The number of aerobic semi-saturated sediment in the sediment (mg/L) 
The respiration rate of the plant (/m20.135 
TP Phosphorus ingested by plants (g P/g DO) 0.092 
The phosphorus content of dissolved BOD (g P/g BDO) 0.05 
Adsorption coefficient (/day) 0.0002 
Phosphorus ingested by bacteria (g P/g DO) 0.015 
Status variablesParameterValue
COD Adsorption coefficient (/day) 0.004 
Semi-saturated oxygen concentration (mg/L) 1.9 
20°C degradation rate coefficient (/day) 0.003 
Temperature coefficient of degradation rate 1.07 
NH3-N Nitrogen uptake semi-saturation (mg/L) 0.045 
BOD decay release ammonia ratio (g NH3-N/g BOD) 0.25 
Temperature coefficient 1.092 
COD degradation of ammonia nitrogen ratio (g NH3-N/g COD) 0.25 
DO The sediment requires oxygen (/m2)/day 0.4 
Sediment aerobic temperature coefficient (/day) 1.07 
The number of aerobic semi-saturated sediment in the sediment (mg/L) 
The respiration rate of the plant (/m20.135 
TP Phosphorus ingested by plants (g P/g DO) 0.092 
The phosphorus content of dissolved BOD (g P/g BDO) 0.05 
Adsorption coefficient (/day) 0.0002 
Phosphorus ingested by bacteria (g P/g DO) 0.015 

According to the corresponding guaranteed rate flow, the corresponding typical representative years are selected, and 2010 is chosen to represent the wet year, 2013 represents the normal year, and 2020 represents the dry year. According to the convergence experience of the model, the upstream boundary is given the flow condition, and the downstream boundary is given the water level condition, when the model is more likely to converge.

Parameter setting

The selection of the parameters will directly affect the accuracy of the model simulation. The following parameters are required for model calculations (Table 2).

Time step

The simulations are conducted for three hydrological scenarios, namely, wet year, average year, and dry year. The simulation time step for each scenario is one day, equivalent to 86,400 s, with a total of 365 time-steps representing 1 year of simulation.

CFL

Courant–Friedrichs–Lewy (CFL) is a stability criterion extensively employed in numerical techniques for solving partial differential equations (PDEs), notably within the realm of computational fluid dynamics (CFD). In numerical PDE-solving methodologies, equations are usually discretized onto a grid or mesh, while time is also divided into minute time intervals (Gnedin et al. 2018). The CFL criterion furnishes a prerequisite guaranteeing the stability of the numerical solution. Rooted in the concept that system information or disturbances ought not to disseminate excessively and swiftly in relation to the numerical method's resolution. It establishes a connection between the time step magnitude, spatial discretization, and the utmost speed at which perturbations can advance within the system. To summarize, the CFL criterion serves as a stability criterion, offering a directive for selecting a fitting time step magnitude within numerical methods geared towards solving CFD problems. This ensures the stability and precision of simulations. CFL is an important condition to judge the convergence of results in hydrodynamic calculations. In this study, CFL is 0.8.

Wind field conditions
The wind field conditions are input into the model using a time series file, and the wind field conditions for the three conditions are shown in Figure 6.
Figure 6

Wind field diagram under three hydrological conditions.

Figure 6

Wind field diagram under three hydrological conditions.

Close modal
Manning coefficient
The Manning coefficient, ranging from 22 to 50 m1/3/s, stands as a pivotal parameter for quantifying bottom bed roughness. The roughness of the riverbed significantly impacts water flow dynamics, predominantly due to the inverse relationship between riverbed smoothness and hydraulic resistance. Variations in bottom roughness directly correlate with the simulation precision of the hydrodynamic model. For this study, the Manning coefficient was established via on-site investigations, in conjunction with techniques described in relevant literature for deriving this coefficient. Multiple evaluations were conducted to ascertain its value, as depicted in Figure 7.
Figure 7

Manning coefficient.

Figure 7

Manning coefficient.

Close modal
Dry and wet water depth

Since the simulation area is in the alternating boundary area of dry and wet, it is necessary to set drying depth, flooding depth and wetting depth to avoid distortion in the operation of the model. According to the definition principle of dry and wet water depth, after several debuggings, the above three parameters are set to 0.005, 0.05, and 0.10 m, respectively.

Eddy viscosity

The eddy viscosity is a parameter to solve the problem of Reynolds additional stress. There are three ways to set, respectively, no eddy viscosity, constant eddy viscosity, and Smagorinsky formula method (Terziev et al. 2020). Eddy viscosity of the Smagorinsky formula setting method in the range of 0.25–1.00, after the model parameters for many times, the final determination of the study eddy viscosity to take the value of 0.28, at this time the minimum error of hydrodynamic simulation results.

Model validation

The constructed hydrodynamic model is used to numerically simulate the Luoyang City section of the Luo River. The model is validated by inputting the above-mentioned model parameters and using the 2020 measured data from the Baimasi hydrological stations. The simulation results and water level error are shown in Figures 8 and 9.
Figure 8

2020 Baimasi water level verification.

Figure 8

2020 Baimasi water level verification.

Close modal
Figure 9

Simulated water level error diagram of Baimasi.

Figure 9

Simulated water level error diagram of Baimasi.

Close modal

In the simulation process, the lack of some data led to a not very good fit between the simulated water level and the actual data in the early stage, but as the model continued to simulate the downstream section of the river, the calculated values are more consistent with the actual measured values. The results show that the model has a high fit, the simulated value of water level has a consistent trend with the measured value, and the model simulation results are good. January–July belongs to the non-flood period, the water level basically remains the same or decreases, after July, the water level starts to rise rapidly, which is in line with the actual situation. From the error graph between the simulated and measured values, we can see that the largest relative error is June 30, the error is only 0.09 m, which is within the acceptable error range. The measured average water level at the Baimasi hydrological station is 113.42 m, and the simulated average water level value is 113.28 m, respectively. The relative error is 0.12%, R2 is 0.942, NSE is 0.928. The relative error is less than 1%, and the R2 and NSE are greater than 0.9, which meets the model calculation requirements.

Establishment of the water quality model

Parameter selection and calibration

Degradation coefficient k is an extraordinary quantity that varies with time and temperature and other influencing factors. Different water quality has different degradation capacity of pollutants, and usually, for the convenience of calculation, it is often considered that for pollutants in the same river section, the degradation capacity is certain and set as k value. The degradation coefficient can generally be determined based on the pollutant concentrations at different monitoring points in the simulation area. In the simulation process, the pollutant in the water body following the movement of the flow field has the phenomenon of diffusion, the degree of diffusion is expressed by the modeling of the given diffusion coefficient of this parameter to determine. The diffusion coefficient is finally determined through several model parameter rate determinations: E = 0.45 m2/s, KCOD = 0.18/day,KTP = 0.04/day, KDO = 0.10/day. After several simulations to verify the solution, the following are the results of water quality model parameter rate determination.

Simulated working conditions and model boundary conditions

In this study, the numerical water quality simulation conditions are set according to the three conditions of the hydrodynamic model. Based on the results of parameter rate determination of the above water environment model, the distribution of pollutant concentrations under different hydrological year assurance rate scenarios was analyzed. The measured time series data of Gao Yazhai hydrological station are selected upstream and the measured time series data of Baimasi hydrological station are selected downstream, as shown in Table 3.

Table 3

Setting of different working conditions and boundary conditions

ConditionWet year (10%)Normal water year (50%)Dry year (90%)
Upstream boundary Hydrology Time series of measured flow in Gao Yazhai in 2010 Time series of measured flow in Gao Yazhai in 2013 Time series of measured flow in Gao Yazhai in 2020 
Water quality Time series of water quality concentration in Gao Yazhai in 2020 
Downstream boundary Hydrology Time series of measured water levels in 2010 at Baimasi Time series of measured water levels in 2013 at Baimasi Time series of measured water levels in 2020 at Baimasi 
Water quality Zero gradient boundary 
ConditionWet year (10%)Normal water year (50%)Dry year (90%)
Upstream boundary Hydrology Time series of measured flow in Gao Yazhai in 2010 Time series of measured flow in Gao Yazhai in 2013 Time series of measured flow in Gao Yazhai in 2020 
Water quality Time series of water quality concentration in Gao Yazhai in 2020 
Downstream boundary Hydrology Time series of measured water levels in 2010 at Baimasi Time series of measured water levels in 2013 at Baimasi Time series of measured water levels in 2020 at Baimasi 
Water quality Zero gradient boundary 

Model validation

Through the analysis of observed water level and water quality data, it has been found that the water quality in the entire watershed does not meet the standards. Significant exceedances were observed for pollutants such as TN, COD, TP, and NH3-N. The primary causes of these exceedances are attributed to the rapid industrial growth and population increase in recent years, leading to urban expansion and a corresponding increase in wastewater discharge. However, inadequate construction of urban sewage treatment plants and collection systems, which have not kept pace with population growth, has resulted in a substantial amount of untreated domestic sewage being directly discharged into the river, severely impacting the water environment in some urban areas. In this paper, we select the data of Baimasi water quality monitoring section and Gao Yazhai water quality monitoring section in 2020 for model validation, with several times of parameter rate setting. The numerical simulation parameters of Luoyang section of the Luo River are optimally selected under the conditions of its water quality simulation effect is the best. By comparing the water quality simulation condition of monitoring section with the actual measurement data, we evaluate the accuracy of the water environment model, the validation results are shown in Figure 10. In order to further investigate the accuracy of the model, the mean error (), the coefficient of certainty (R2), and the Nash–Sutcliffe Efficiency coefficient (NSE) are introduced to further discern the accuracy of the model (Zhang et al. 2022). The calculation formulas are as follows:
formula
(9)
formula
(10)
formula
(11)
where T is the total number of observations, denotes the measured concentration, denotes the simulated value, i denotes the number of samples and n denotes the number of monitoring sites. NSE ranges from 0 to 1, and the closer the NSE is to 1, the more accurate the model calculation is. R2 results are the opposite. The closer the result of R2 is to 1 indicates that the model calculations are less precise. The results are shown in Table 4.
Figure 10

Model COD concentration validation graph.

Figure 10

Model COD concentration validation graph.

Close modal
Table 4

Water quality model validation results

SectionIndicatorR2NSE
Gao Yazhai monitoring section COD 2.40 0.89 0.87 
DO 2.66 0.90 0.88 
NH3-N 3.63 0.87 0.85 
TP 1.30 0.92 0.90 
Baimasi monitoring section COD 4.96 0.85 0.83 
DO 2.92 0.91 0.89 
NH3-N 4.14 0.86 0.84 
TP 1.01 0.93 0.90 
SectionIndicatorR2NSE
Gao Yazhai monitoring section COD 2.40 0.89 0.87 
DO 2.66 0.90 0.88 
NH3-N 3.63 0.87 0.85 
TP 1.30 0.92 0.90 
Baimasi monitoring section COD 4.96 0.85 0.83 
DO 2.92 0.91 0.89 
NH3-N 4.14 0.86 0.84 
TP 1.01 0.93 0.90 

Broadly speaking, the model attains commendable outcomes in simulating pollutant concentrations. The deviations in simulated COD concentrations at both monitoring sections remain below 7%, with maximum, minimum, and average errors of 6.84, 0.35, and 3.68%, respectively. Simulation outcomes reveal that pollutant concentrations in each zone fall below actual monitoring data, suggesting the model's propensity to potentially underestimate pollutant concentrations. In certain instances of extreme values and abrupt transitions, simulation outcomes prove subpar, warranting further enhancements and optimization efforts. The statistical analysis depicted in Table 5 showcases the average R2 and NSE values derived through the execution of the water environment simulation model devised in this study. Notably, R2 values in both model sections surpass 0.85, denoting a robust correlation between observed and projected data. Moreover, the NSE values exceed 0.80, further bolstering the claim that the model demonstrates reliability in accurately prognosticating the pollution trends across various state variables.

Table 5

Table of pollutant degradation coefficients in the river section of the study area

River section numberCross-sectionHydrological conditionDegradation coefficient (K)
TPCODNH3-N
L1 Gao Yazhai-Li Lou Wet year 0.1428 0.1517 0.0824 
Normal year 0.1107 0.0632 0.0632 
Dry year 0.0858 0.0387 0.0414 
L2 Li Lou-Baimasi Wet year 0.0905 0.2114 0.0588 
Normal year 0.0839 0.1829 0.0453 
Dry year 0.0531 0.0652 0.0327 
L3 Baimasi-G207 Highway Bridge Wet year 0.1208 0.1452 0.0619 
Normal year 0.1128 0.1085 0.0531 
Dry year 0.0990 0.0648 0.0399 
River section numberCross-sectionHydrological conditionDegradation coefficient (K)
TPCODNH3-N
L1 Gao Yazhai-Li Lou Wet year 0.1428 0.1517 0.0824 
Normal year 0.1107 0.0632 0.0632 
Dry year 0.0858 0.0387 0.0414 
L2 Li Lou-Baimasi Wet year 0.0905 0.2114 0.0588 
Normal year 0.0839 0.1829 0.0453 
Dry year 0.0531 0.0652 0.0327 
L3 Baimasi-G207 Highway Bridge Wet year 0.1208 0.1452 0.0619 
Normal year 0.1128 0.1085 0.0531 
Dry year 0.0990 0.0648 0.0399 

Calculation of the water environment capacity

Regarding the water environmental capacity of rivers, there is a basic consensus that it represents the total amount of pollutants that a region can tolerate under specific environmental indicators. The calculation of water environmental capacity is usually carried out using a segmented method, and the capacity of each segment is accumulated to obtain the overall water environmental capacity (Li et al. 2010). Water environmental capacity is divided into dilution capacity (Edilution) and self-cleaning capacity (Eself-cleaning).
formula
(12)
where S is the water quality standard (mg/L), is the flow rate of the river (m3/s), is the background concentration of the river (mg/L).
formula
(13)
where S is the water quality standard (mg/L), Qt is the flow rate of the river plus the wastewater flow rate (m3/s), l is the length of each river section (m), k is the degradation coefficient, and u is the river flow rate (m/s).
The calculation methods for water quality degradation coefficient mainly include the empirical data backward calculation method, analogy method, and analysis borrowing method. In this study, reference was made to domestic and international research achievements on water quality degradation coefficients. Based on the simulated water quality results of the Luo River in the Luoyang section, the empirical data backward calculation method was selected to calculate the water quality degradation coefficient of the study area river section. The calculation formula is as follows:
formula
(14)
where K is the pollutant degradation coefficient, u is the flow velocity of water (m/s), L is the length of the river section (km), C1 is the pollutant concentration at the upstream calculation cross-section (mg/L), and C2 is the pollutant concentration at the downstream calculation cross-section (mg/L). The results of the degradation coefficient calculation are shown in Table 5.

Spatiotemporal variations of water quality

A monitoring point is selected to monitor the change of pollutants in the upper and middle reaches of the model river, respectively. The location of water quality monitoring sites A1 (3829800.57, 627684.14), A2 (3841072.47, 644828.14), and A3 (3842147.29, 660128.35) are shown in Figure 11.
Figure 11

Location map of monitoring points in the study river section.

Figure 11

Location map of monitoring points in the study river section.

Close modal

Wet year

Pollutant distribution within the river is simulated under a prescribed assurance level of P = 10%. Emphasis is placed on investigating the spatial dispersion of COD, TP, NH3-N, and DO, as illustrated in Figure 12. The range of COD concentrations varies along different river sections: from Gao Yazhai to Li Lou, it spans 16–21 mg/L; from Li Lou to Baimasi, it extends between 19 and 24 mg/L; and from Baimasi to the G207 highway bridge, it fluctuates from 22 to 25 mg/L. TP concentrations display the following ranges: from Gao Yazhai to Li Lou, they fall between 0.08 and 0.12 mg/L; from Li Lou to Baimasi, they span 0.11–0.15 mg/L; and from Baimasi to the G207 highway bridge, they fluctuate between 0.14 and 0.17 mg/L. Generally, TP concentrations in the investigated river sections remain low. This is largely attributed to diminished water velocity at bends, which subsequently affects the mobility and concentration of TP. During wet years, the total phosphorus concentration essentially aligns with the specifications for Class II water quality. The NH3-N concentrations range from 0.51 to 0.65 mg/L from Gao Yazhai to Li Lou, 0.45–0.81 mg/L from Li Lou to Baimasi, and 0.53–1.12 mg/L from Baimasi to the G207 highway bridge. These values categorize the water body as Class III.
Figure 12

Distribution of each pollution index during the wet year.

Figure 12

Distribution of each pollution index during the wet year.

Close modal
Figure 13 illustrates the temporal evolution of pollutant concentrations at three monitoring points – A1, A2, and A3 – under wet year conditions. As depicted in the figure, pollutant concentrations at these points exhibit ranges between categories II and IV. Specifically, COD concentrations at A1 range from 16 to 21 mg/L, while A2 and A3 consistently exceed 20 mg/L. Notably, COD levels are notably lower at the A1 monitoring point compared to A2 and A3. TP and DO concentrations at all three monitoring points fall within the III to IV range. Furthermore, the spatial distribution of NH3-N pollution indicates superior water quality during the flood season spanning May to August, a trend consistent with actual conditions.
Figure 13

Pollutant concentration at each monitoring point during the wet year.

Figure 13

Pollutant concentration at each monitoring point during the wet year.

Close modal

Normal year

Pollutant distribution within the river is simulated under the condition of a P = 50% guarantee rate, as depicted in Figure 14. COD concentrations range from 18 to 22 mg/L from Gao Yazhai to Li Lou, 21–25 mg/L from Li Lou to Baimasi, and 23–25 mg/L from Baimasi to the G207 highway bridge. The hydrodynamic conditions during a normal water year are not as favorable as those during wet years, which subsequently impacts COD degradation. This condition primarily accounts for the elevated COD concentration downstream. COD concentrations within this region generally satisfy the criteria for Class IV water quality. The collective TP concentration ranges from 0.12 to 0.18 mg/L. It notably falls short of the water quality observed under wet year conditions. NH3-N concentrations range from 0.71 to 1.2 mg/L from Gao Yazhai to Li Lou, 0.83–1.61 mg/L from Li Lou to Baimasi, and between 0.75 and 1.52 mg/L from Baimasi to the G207 highway bridge, categorizing the concentrations as class IV. NH3-N displays patterns corresponding to water flow. DO concentrations vary between 4.01 and 8.27 mg/L, exhibiting a decrease compared to the conditions of the P = 10% guarantee rate. This reduction is attributed to inadequate river hydrodynamics, reduced flow rates, and further declines in reoxygenation rates.
Figure 14

Distribution of each pollution indicator during the normal year.

Figure 14

Distribution of each pollution indicator during the normal year.

Close modal
Figure 15 illustrates the temporal evolution of pollutant concentrations at each monitoring point under normal water year conditions. As depicted in the figure, pollutant concentrations at the three monitoring points, during a normal year's flow conditions, primarily fall between categories III and IV. Specifically, COD concentrations at A1 range from 18 to 22 mg/L, while A2 and A3 consistently exceed 22 mg/L. This aligns with the findings of the wet year analysis. TP concentration within the water body exhibits an increasing trend as fluid travels downstream. Urban river sections experience impacts from industrial and domestic wastewater, coupled with diminished hydrodynamic conditions, resulting in TP concentrations largely conforming to Class IV water standards. Similar to the pattern observed during abundant water years, NH3-N concentration also follows a rising-falling trend, and NH3-N levels in the water are notably low during the flood season. Industrial effluents primarily influence NH3-N content, leading to NH3-N concentrations corresponding to either class III or class IV water standards at each monitoring point. Changes in DO concentration are relatively subtle. During the wet period, DO concentration does not significantly differ from that of the dry period. Additionally, a slight increase is observed in comparison to the wet year, falling within class IV standards.
Figure 15

Pollutant concentration at each monitoring point during the normal year.

Figure 15

Pollutant concentration at each monitoring point during the normal year.

Close modal

Dry year

The distribution of pollutants in the river is simulated under a guaranteed rate of P = 90%, as depicted in Figure 16. The dry year is characterized by the poorest hydrodynamic conditions, low incoming water flow, and reduced precipitation, making it highly unfavorable for pollutant deposition and diffusion. COD concentrations range from 24 to 26 mg/L in the Gao Yazhai section to Li Lou section, 25–28 mg/L from Li Lou section to Baimasi section, and 27–29 mg/L from Baimasi section to G207 highway bridge section. The concentration of TP ranges from 0.16 to 0.21 mg/L, adhering to Class IV water standards. NH3-N concentration from the Gao Yazhai section to the Li Lou section ranges between 0.62 and 1.24 mg/L, from the Li Lou section to the Baimasi section ranges between 0.72 and 1.32 mg/L, and from the Baimasi section to the G207 highway bridge section ranges between 0.76 and 1.45 mg/L, all corresponding to Class IV water standards. DO concentration ranges from 7.01 to 9.83 mg/L.
Figure 16

Distribution of each pollution index during the dry year.

Figure 16

Distribution of each pollution index during the dry year.

Close modal
Figure 17 illustrates the temporal evolution of pollutant concentrations at each monitoring point during a dry year. Specifically, the COD concentration at A1 ranges from 24 to 26 mg/L, while A2 and A3 display concentrations of 24–30 mg/L. The efficiency of COD degradation significantly diminishes during dry periods, when hydrodynamic conditions are at their most unfavorable. Under such circumstances, such as the dry periods within a dry year, the discharges from sewage treatment plants exert a more substantial impact on the river's water quality. The concentration of TP exhibits an increasing trend downstream, influenced by the flow of fluid. Coupled with the impact of industrial and domestic outfall wastewater within the urban river section, TP concentrations primarily correspond to Class IV water standards. Similar to observations during wet years, the NH3-H concentration pattern demonstrates an increase followed by a decrease. The concentration of NH3-H in the water body is notably low during the flood season. NH3-H concentration is closely related to industrial wastewater, and concentrations at each monitoring point predominantly adhere to Class IV standards.
Figure 17

Pollutant concentration at each monitoring point during the dry year.

Figure 17

Pollutant concentration at each monitoring point during the dry year.

Close modal

In summary, from the perspective of hydrological year, the overall water environment quality is ranked from good to bad in the order of wet year, normal year, and dry year. The large amount of precipitation in wet year has a diluting effect on pollutants, and good hydrodynamic conditions and faster flow rates increase the self-purification capacity of the river, thus leading to the best water quality in that hydrological year. From the point of view of the concentration of water quality indicators, the overall water quality concentration in descending order is COD, NH3-N, TP, and DO. Because there are a large number of wastewater treatment plants in the downstream, such as Luonan Wastewater Treatment Plant, Jiandong Wastewater Treatment Plant, Luoning Wastewater Treatment Plant, and a large number of sewage outfalls, such as the Luolonglu outfall, the Huarun Power industrial outfall, which discharges a large amount of industrial wastewater, leading to an increase in the concentration of COD. Farmland surface source pollution and residential sewage discharge lead to increased concentrations of NH3-N and TP. In terms of sections, the overall water quality of the sections is ranked from good to bad in the order of A1, A2, and A3. The water quality of section is poorer in the downstream due to the presence of a large number of wastewater treatment plants and outfalls in the downstream, and the deterioration of water quality in the downstream is aggravated by the cumulative effect of water quality in the upstream to that in the downstream.

Spatiotemporal variations of water environment capacity

In this paper, the typical representative factors COD, NH3-N, and TP are selected as the control indexes of the water environment capacity of Luoyang city section of the Luo River. We examined the water environment capacity of river sections under three hydrological periods: wet, normal, and dry. The one-dimensional steady-state model is used to calculate the water environment capacity. The calculation results are shown in Tables 68.

Table 6

COD water environment capacity accounting results of the study area river section

River section numberCross-sectionCalculation typeCOD water environment capacity
Wet periodNormal periodDry period
L1 Gao Yazhai-Li Lou Dilution capacity 2,874.614 1,880.779 738.132 
Self-cleaning capacity 3,942.566 2,438.966 1,301.870 
Environmental capacity 6,817.180 4,319.746 2,040.002 
L2 Li Lou-Baimasi Dilution capacity 1,846.793 564.621 176.887 
Self-cleaning capacity 1,384.567 1,091.814 361.241 
Environmental capacity 3,231.360 1,656.434 538.127 
L3 Baimasi -G207 Highway Bridge Dilution capacity 2,311.425 1,008.067 395.064 
Self-cleaning capacity 2,326.899 1,587.613 681.432 
Environmental capacity 4,638.323 2,595.679 1,076.496 
River section numberCross-sectionCalculation typeCOD water environment capacity
Wet periodNormal periodDry period
L1 Gao Yazhai-Li Lou Dilution capacity 2,874.614 1,880.779 738.132 
Self-cleaning capacity 3,942.566 2,438.966 1,301.870 
Environmental capacity 6,817.180 4,319.746 2,040.002 
L2 Li Lou-Baimasi Dilution capacity 1,846.793 564.621 176.887 
Self-cleaning capacity 1,384.567 1,091.814 361.241 
Environmental capacity 3,231.360 1,656.434 538.127 
L3 Baimasi -G207 Highway Bridge Dilution capacity 2,311.425 1,008.067 395.064 
Self-cleaning capacity 2,326.899 1,587.613 681.432 
Environmental capacity 4,638.323 2,595.679 1,076.496 
Table 7

Accounting results of TP water environment capacity in the study area river section

River section numberCross-sectionCalculation typeTP water environment capacity
Wet periodNormal periodDry period
L1 Gao Yazhai-Li Lou Dilution capacity 363.727 204.965 80.542 
Self-cleaning capacity 49.620 35.045 21.007 
Environmental capacity 413.347 240.010 101.549 
L2 Li Lou-Baimasi Dilution capacity 262.822 135.354 51.486 
Self-cleaning capacity 7.975 6.747 4.817 
Environmental capacity 270.797 142.101 56.302 
L3 Baimasi -G207 Highway Bridge Dilution capacity 307.408 171.449 66.779 
Self-cleaning capacity 25.913 21.991 11.087 
Environmental capacity 333.321 193.440 77.866 
River section numberCross-sectionCalculation typeTP water environment capacity
Wet periodNormal periodDry period
L1 Gao Yazhai-Li Lou Dilution capacity 363.727 204.965 80.542 
Self-cleaning capacity 49.620 35.045 21.007 
Environmental capacity 413.347 240.010 101.549 
L2 Li Lou-Baimasi Dilution capacity 262.822 135.354 51.486 
Self-cleaning capacity 7.975 6.747 4.817 
Environmental capacity 270.797 142.101 56.302 
L3 Baimasi -G207 Highway Bridge Dilution capacity 307.408 171.449 66.779 
Self-cleaning capacity 25.913 21.991 11.087 
Environmental capacity 333.321 193.440 77.866 
Table 8

Accounting results of NH3-N water environment capacity of the river section in the study area

River section numberCross-sectionCalculation typeNH3-N water environment capacity
Wet periodNormal periodDry period
L1 Gao Yazhai-Li Lou Dilution capacity 1,131.073 620.051 238.568 
Self-cleaning capacity 162.920 102.024 61.404 
Environmental capacity 1,293.993 722.076 299.971 
L2 Li Lou-Baimasi Dilution capacity 553.803 265.552 95.835 
Self-cleaning capacity 39.877 29.813 12.135 
Environmental capacity 593.680 295.365 107.970 
L3 Baimasi-G207 Highway Bridge Dilution capacity 835.398 462.783 177.906 
Self-cleaning capacity 122.734 61.341 36.221 
Environmental capacity 958.132 524.123 214.127 
River section numberCross-sectionCalculation typeNH3-N water environment capacity
Wet periodNormal periodDry period
L1 Gao Yazhai-Li Lou Dilution capacity 1,131.073 620.051 238.568 
Self-cleaning capacity 162.920 102.024 61.404 
Environmental capacity 1,293.993 722.076 299.971 
L2 Li Lou-Baimasi Dilution capacity 553.803 265.552 95.835 
Self-cleaning capacity 39.877 29.813 12.135 
Environmental capacity 593.680 295.365 107.970 
L3 Baimasi-G207 Highway Bridge Dilution capacity 835.398 462.783 177.906 
Self-cleaning capacity 122.734 61.341 36.221 
Environmental capacity 958.132 524.123 214.127 

Water environment capacity variations in different hydrological periods

The dynamic changes of water environment capacity in different hydrological periods can be seen from the above three tables. The water environment capacity of COD is the largest, NH3-N is the second largest and TP is the smallest in the section of Luoyang city, where the water environment capacity of COD is 26,913.35 kg/d, NH3-N is 5,009.44 kg/d, and TP is 1,828.73 kg/d. The water environment capacity of the three pollution indicators is the largest in the wet period, the second largest in the normal water period and the smallest in the dry period, in which the water environment capacity of COD is 14,686.86; 8,571.86; and 3,654.63 kg/d in the wet, normal, and dry water periods, respectively, the water environment capacity of NH3-N is 2,845.80; 1,541.56; and 622.07 kg/d in wet, normal, and dry water periods, respectively. The water environment capacity of TP is 1,017.47; 575.55; and 235.72 kg/d in wet, normal, and dry water periods, respectively. This is mainly because COD, NH3-N and TP in the Luoyang City section of the Luo River mainly originate from point source pollution. The poor hydrodynamic conditions of the river during the dry period and the decrease in flow rate, coupled with the growth of industrial and urban residents' domestic water demand, result in the growth of urban industrial and domestic wastewater discharge. It can be seen that there is a non-uniformity in the water environment capacity of the Luoyang city section of the Luo River in time, and calculating it only with a certain guarantee rate will inevitably lead to the waste of water environment capacity, which will in turn restrict the economic development of the region. Studying the water environment capacity of the river section from the perspective of different hydrological periods is more conducive to socio-economic development and regional water resources protection.

Water environment capacity variations in different river sections

For COD, the longest L1 river section corresponds to the largest COD water environment capacity, which is 6,817.18; 4,319.75; and 2,040.00 t/a in the wet, normal, and dry periods, respectively. L3 is the second longest, which corresponds to the smaller COD water environment capacity, which is 4,638.32; 2,595.68; and 1,076.50 t/a in the wet, normal, and dry periods, respectively. L2 is the shortest river section and has the smallest water environment capacity, which is 3,231.36; 1,656.43; and 538.13 t/a in wet, normal, and dry periods, respectively, indicating that the longer the length of the river, the larger the water environment capacity of COD. Of course, the size of the water environmental capacity should also have a great relationship with the water quality objectives, but the water quality objectives on the river section of this study are all class I. Therefore, the difference of the water environmental capacity of this study is mainly reflected in the length of the river.

For NH3-N, under the conditions of different hydrological periods, the water environmental capacity of each river section of the Luoyang city section of the Luo River is the largest during the wet period, slightly smaller than that during the normal period, and the worst flow conditions during the dry period lead to the smallest water environmental capacity. The water environmental capacity of L1 is the largest, and the water environmental capacity during the wet, normal, and dry periods are 1,293,99, 722.08, and 299.97 t/a, respectively. The corresponding water environmental capacity of L3 is the second shortest and corresponds to a smaller water environmental capacity of NH3-N, which is 958.13, 524.12, 214.13 t/a during the wet, normal, and dry periods, respectively. L2 is the shortest and has the smallest water environmental capacity, which is 93.68, 295.36, 107.97 t/a, further indicating that the longer the length of the river, the greater the NH3-N water environmental capacity.

For TP, the flow conditions of each section of the Luo River Luoyang city are the best during the wet period resulting in the largest water environment capacity, followed by the second in the normal period and the smallest in the dry period. L1 has the largest water environment capacity, which is 413.35, 240.01, 101.55 t/a in the wet, normal, and dry periods, respectively. L3 has the second largest, corresponding to the smallest water environment capacity of TP, which is 333.32, 193.44, 77.87 t/a in the wet, normal, and dry water periods, respectively. L2 was the shortest and had the smallest water environment capacity, with the TP water environment capacity in wet, normal, dry periods being 270.80; 142.10; and 56.30 t/a, respectively.

Water environment management

In summary, in terms of hydrological periods, the capacity of the water environment is ranked from largest to smallest as the wet period, the normal period, and the dry period. The higher precipitation is in the wet period, which has a higher dilution capacity for pollutants, and better hydrodynamic conditions, with higher flow rates and better pollutant transport performance. From the point of view of water quality indicators, the capacity of the water environment from large to small is in the order of COD, NH3-N, TP. The degradation coefficient of COD is larger compared to NH3-N and TP, and affected by the discharge of industrial wastewater, the flow of COD discharged into the river is higher, which leads to the larger water environment capacity for COD. From the point of view of the river section, the capacity of the water environment is ranked from large to small as L1, L3, L2. The length of the river section plays a key role in the size of the capacity of the water environment, and the longer the river section, the larger the capacity of the water environment.

Considering the field conditions of the Luoyang urban section of the Luo River, the following control suggestions are proposed. Based on the characteristics of pollutants in the river channel of the Luoyang urban section, targeted cultivation of microbial communities is recommended for pollutant decomposition. These microbial communities can then be transplanted into the river channel to ensure favorable growth conditions. Measures such as artificial aeration and oxygenation should be implemented to support the survival of the microbial communities, significantly increasing their survival time and degradation effectiveness. It is important to note that throughout the entire stage of applying water ecological restoration techniques, priority should be given to selecting diverse and abundant plant species that can better complement each other in order to improve water quality through degradation effects. During the low-water period, pollutant discharge restrictions should be implemented, while water flow control measures should be taken during the high-water period. In the Luoyang urban section of the Luo River, a portion of COD, NH3-N, and TP pollutants originates from non-point source pollution. During the high-water period with increased rainfall and higher flow rates, non-point source pollutants are more likely to enter the river through stormwater runoff, leading to a decrease in water environmental capacity. During the low-water period, when water flow is reduced and industrial water supply to the riverside industrial enterprises in the Luoyang section decreases, it is recommended for the city of Luoyang to establish a pollution prevention and control system. This system should include implementing temporary shutdowns and production restrictions for key pollution source enterprises, as well as strengthening supervision to ensure compliance with pollutant emissions standards.

In order to explore the impact of hydrological periods, water quality indicators, and river sections on the water environment capacity, this paper took the Luo River in Luoyang city as an example, we used MIKE21-coupled hydrodynamic and water quality model and the segmentation method to calculate the water environment capacity of COD, TP, and NH3-N in wet, normal, and dry periods under L1, L2, and L3 sections. The following conclusions were shown.

During the hydrodynamic model process, the average value of the measured water level is 113.42 m and the average value of the simulated water level is 113.28 m, with a relative error of 0.12%. R2 is 0.942 and NSE is 0.928. During the water quality model process, the R2 reached more than 0.85 and the mean value of NSE of the two monitoring sections reached above 0.80. Thus, the model is reliable and can accurately predict the migration change of each water quality index.

From the perspective of hydrological year, the water quality is best in the wet year, which is due to the fact that the increase of precipitation has a diluting effect on the pollutants, the faster flow rate can also promote the water purification capacity, and the largest water environment capacity during the wet period. In terms of the concentration of water quality indexes, the concentration of COD is the largest, but the higher degradation coefficients result in the largest water environment capacity for COD. Therefore, the government needs to strengthen the management of COD, and control the discharge of industrial wastewater upstream and the domestic sewage discharge of coastal residents. In terms of river sections, the water quality in the downstream is worse than the upstream due to the diffusion and accumulation of pollutants from the upstream to the downstream, and the presence of a large number of wastewater treatment plants in the downstream exacerbates the deterioration of water quality. Additionally, the longer river section results in the largest water environment capacity in L1.

This paper can achieve the organic linkage between the management of the water quality objectives and the control of the total amount of pollution and provide scientific guidance to the improvement of water quality in the Luo River. This study also broadens new ideas for the application of MIKE 21 and promoting refined management of river water environments in wet, normal, and dry periods. However, there are some shortcomings in this paper. For example, this paper only selected the main pollutant exceedance factors COD, DO, TP, NH3-N. There are still some important pollutants not considered, such as biochemical oxygen demand and Escherichia coli. Thus, further research is needed. The hydrodynamic water quality coupling model and water environment capacity model can be considered in the future with a three-dimensional model, so that the vertical flow and pollution dispersion are also taken into account.

This work was supported by the North China University of Water Resources and Electric Power Innovation Ability Improvement Project for Postgraduates [grant number NCWUYC-2023006] and [grant number NCWUYC-2023027].

All authors contributed to the study conception and design. Writing and editing by Y.W. Preliminary data collection by Y.W. All authors read and approved the final manuscript.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.

Deng
C.
,
Liu
L.
,
Li
H.
,
Peng
D.
,
Wu
Y.
,
Xia
H.
,
Zhang
Z.
&
Zhu
Q.
2021
A data-driven framework for spatiotemporal characteristics, complexity dynamics, and environmental risk evaluation of river water quality
.
Science of The Total Environment
785
,
147134
.
Edwards
P.
,
Clarke
M.
,
Diguiseppi
C.
,
Pratap
S.
,
Roberts
I.
&
Wentz
R.
2002
Identification of randomized controlled trials in systematic reviews: Accuracy and reliability of screening records
.
Statistics in Medicine
21
(
11
),
1635
1640
.
Fatahi Nafchi
R.
,
Yaghoobi
P.
,
Reaisi Vanani
H.
,
Ostad-Ali-Askari
K.
,
Nouri
J.
&
Maghsoudlou
B.
2021
Eco-hydrologic stability zonation of dams and power plants using the combined models of SMCE and CEQUALW2
.
Applied Water Science
11
(
7
),
109
.
Fu
Y.
,
He
G.
,
Zhao
S.
&
Jiang
H.
2023
Research on watershed water ecological management and compensation based on evolutionary game
.
AQUA – Water Infrastructure, Ecosystems and Society
72
(
7
),
1164
1183
.
García-Alba
J.
,
Bárcena
J. F.
,
Ugarteburu
C.
&
García
A.
2019
Artificial neural networks as emulators of process-based models to analyse bathing water quality in estuaries
.
Water Research
150
,
283
295
.
Gayathri
R.
,
Murty
P. L. N.
,
Bhaskaran
P. K.
&
Srinivasa Kumar
T.
2016
A numerical study of hypothetical storm surge and coastal inundation for AILA cyclone in the Bay of Bengal
.
Environmental Fluid Mechanics
16
,
429
452
.
Ghidossi
R.
,
Veyret
D.
&
Moulin
P.
2006
Computational fluid dynamics applied to membranes: State of the art and opportunities
.
Chemical Engineering and Processing: Process Intensification
45
(
6
),
437
454
.
Ghunowa
K.
,
Macvicar
B. J.
&
Ashmore
P.
2021
Stream power index for networks (SPIN) toolbox for decision support in urbanizing watersheds
.
Environmental Modelling & Software
144
,
105185
.
Gnedin
N. Y.
,
Semenov
V. A.
&
Kravtsov
A. V.
2018
Enforcing the Courant–Friedrichs–Lewy condition in explicitly conservative local time stepping schemes
.
Journal of Computational Physics
359
,
93
105
.
Goliatt
L.
,
Sulaiman
S. O.
,
Khedher
K. M.
,
Farooque
A. A.
&
Yaseen
Z. M.
2021
Estimation of natural streams longitudinal dispersion coefficient using hybrid evolutionary machine learning model
.
Engineering Applications of Computational Fluid Mechanics
15
(
1
),
1298
1320
.
González
S. O.
,
Almeida
C. A.
,
Calderón
M.
,
Mallea
M. A.
&
González
P.
2014
Assessment of the water self-purification capacity on a river affected by organic pollution: Application of chemometrics in spatial and temporal variations
.
Environmental Science and Pollution Research
21
,
10583
10593
.
Hartmann
A.
,
Goldscheider
N.
,
Wagener
T.
,
Lange
J.
&
Weiler
M.
2014
Karst water resources in a changing world: Review of hydrological modeling approaches
.
Reviews of Geophysics
52
(
3
),
218
242
.
Kumar
P.
,
Lai
S. H.
,
Mohd
N. S.
,
Kamal
M. R.
,
Ahmed
A. N.
,
Sherif
M.
,
Sefelnasr
A.
&
El-Shafie
A.
2021
Enhancement of nitrogen prediction accuracy through a new hybrid model using ant colony optimization and an Elman neural network
.
Engineering Applications of Computational Fluid Mechanics
15
(
1
),
1843
1867
.
Li
Y.
,
Qiu
R.
,
Yang
Z.
,
Li
C.
&
Yu
J.
2010
Parameter determination to calculate water environmental capacity in Zhangweinan Canal Sub-basin in China
.
Journal of Environmental Sciences
22
(
6
),
904
907
.
Li
X.
,
Huang
M.
&
Wang
R.
2020
Numerical simulation of Donghu Lake hydrodynamics and water quality based on remote sensing and MIKE 21
.
ISPRS International Journal of Geo-Information
9
(
2
),
94
.
Michot
B.
,
Meselhe
E. A.
,
Rivera-Monroy
V. H.
,
Coronado-Molina
C.
&
Twilley
R. R.
2011
A tidal creek water budget: Estimation of groundwater discharge and overland flow using hydrologic modeling in the Southern Everglades
.
Estuarine, Coastal and Shelf Science
93
(
4
),
438
448
.
Montazeri
A.
,
Mazaheri
M.
,
Morid
S.
&
Mosaddeghi
M. R.
2023
Effects of upstream activities of Tigris-Euphrates River Basin on water and soil resources of Shatt al-Arab Border River
.
Science of The Total Environment
858
,
159751
.
Paniconi
C.
&
Putti
M.
2015
Physically based modeling in catchment hydrology at 50: Survey and outlook
.
Water Resources Research
51
(
9
),
7090
7129
.
Regan
R. S.
,
Juracek
K. E.
,
Hay
L. E.
,
Markstrom
S. L.
,
Viger
R. J.
,
Driscoll
J. M.
,
Lafontaine
J. H.
&
Norton
P. A.
2019
The US geological survey national hydrologic model infrastructure: Rationale, description, and application of a watershed-scale model for the conterminous United States
.
Environmental Modelling & Software
111
,
192
203
.
Son
C. T.
,
Giang
N. T. H.
,
Thao
T. P.
,
Nui
N. H.
,
Lam
N. T.
&
Cong
V. H.
2020
Assessment of Cau River water quality assessment using a combination of water quality and pollution indices
.
Journal of Water Supply: Research and Technology – AQUA
69
(
2
),
160
172
.
Sun
K.
,
Rajabtabar
M.
,
Samadi
S.
,
Rezaie-Balf
M.
,
Ghaemi
A.
,
Band
S. S.
&
Mosavi
A.
2021
An integrated machine learning, noise suppression, and population-based algorithm to improve total dissolved solids prediction
.
Engineering Applications of Computational Fluid Mechanics
15
(
1
),
251
271
.
Suriya
S.
&
Mudgal
B. V.
2012
Impact of urbanization on flooding: The Thirusoolam sub watershed–A case study
.
Journal of Hydrology
412
,
210
219
.
Tang
W.
,
Pei
Y.
,
Zheng
H.
,
Zhao
Y.
,
Shu
L.
&
Zhang
H.
2022
Twenty years of China's water pollution control: Experiences and challenges
.
Chemosphere
295
,
133875
.
Terziev
M.
,
Tezdogan
T.
&
Incecik
A.
2020
Application of eddy-viscosity turbulence models to problems in ship hydrodynamics
.
Ships and Offshore Structures
15
(
5
),
511
534
.
Wood
M.
,
Haigh
I. D.
,
Le
Q. Q.
,
Nguyen
H. N.
,
Tran
H. B.
,
Darby
S. E.
,
Marsh
R.
,
Skliris
N.
,
Hirschi
J. M.
,
Nicholls
R. J.
&
Bloemendaal
N.
2023
Climate-induced storminess forces major increases in future storm surge hazard in the South China Sea region
.
Natural Hazards and Earth System Sciences
23
(
7
),
2475
2504
.
Zhang
X.
,
Zhao
D.
,
Duan
B.
&
Qiao
W.
2022
A novel groundwater burial depth prediction model – based on the combined VMD-WSD-ELMAN model
.
Environmental Science and Pollution Research
29
(
50
),
76310
76320
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).