Model-based dynamic estimation of water environmental capacity using grid-level simulation and functional zoning: a case study of the Ganjiang River estuary

To assess the water environmental capacity of the Ganjiang River estuary, this study delineated the region into six sub-areas based on designated water functional zones and established 14,600 computational grids. A field-based pollution source survey and cross-sectional water quality monitoring were conducted to characterize pollutant load distribution. A two-dimensional hydrodynamic–water quality coupled model was developed to simulate the transport and transformation of permanganate index (COD_Mn), ammonia nitrogen (NH₃-N), and total phosphorus (TP) under 2020 hydrological conditions. Based on simulation outputs, a dynamic estimation method was proposed by integrating grid-scale pollutant concentrations, region-specific water exchange periods, and functional zone water quality standards. Results showed clear seasonal variation, with peak capacities in June–July (e.g., 4307.50 t/day for COD_Mn, 858.75 t/day for NH₃-N, and 149.00 t/day for TP) and minima in December. Sensitivity analysis indicated that pollutant load was the dominant factor, while dispersion and degradation coefficients had weaker effects. Monte Carlo–based uncertainty analysis confirmed model robustness, with coefficients of variation <0.15 and 95% confidence intervals consistent with observations. Comparison with conventional static methods revealed that static estimates were only 80%–85% of dynamic values, underestimating capacity during high-flow periods and overestimating it in low-flow months. By explicitly considering pollutant input, degradation, and hydrodynamic transport, the proposed approach yields capacity estimates that reflect realistic residual assimilative potential, i.e., the remaining pollutant-bearing capacity after accounting for existing loads. This framework offers a practical tool for total load control, differentiated permitting, and adaptive water quality management in multi-functional estuarine systems.

1 Introduction

Water environmental capacity (WEC) is defined as the maximum pollutant load a water body can assimilate without breaching designated water quality standards (Zhang et al., 2024). It provides a critical basis for pollutant load allocation, discharge permitting, and functional zoning within watersheds. In recent years, accurate estimation of water environmental capacity has gained increasing importance for advancing high-quality regional water resource governance and ensuring compliance with water quality targets and ecological integrity (Chen et al., 2022; Li et al., 2022). This need is particularly critical in hydrologically complex systems such as lakes and river estuaries (Feng and Huang, 2008), where conventional assessment methods often lack the spatial resolution, temporal responsiveness, and process-level detail required to capture dynamic pollutant transport mechanisms (Bu et al., 2020; Dai et al., 2022). Consequently, there is an urgent demand for dynamic estimation approaches that incorporate spatial, temporal, and hydrodynamic process dimensions.

Globally, research on water environmental capacity has evolved from static load estimation to dynamic, process-based simulation (Chen et al., 2024). Early methods were primarily based on the principle of “meeting water quality standards at control sections,” relying on water balance calculations and concentration differentials to estimate assimilative capacity. Notable examples include the Total Maximum Daily Load framework in the United States and Japan’s tiered water quality zoning and load control strategies (Zhang et al., 2018). Since the early 21st century, the integration of water quality models into capacity assessments has become increasingly prevalent (Hu GZ. et al., 2021; Bing et al., 2022). Coupled hydrodynamic–water quality models—such as River and Stream Water Quality Model QUAL2K, MIKE 21 Hydrodynamic and Water Quality Modeling System, Environmental Fluid Dynamics Code, and Water Quality Analysis Simulation Program—enable spatiotemporal simulation of pollutant transport and support scenario-based environmental capacity evaluation (Liu et al., 2018; Wen and Wang, 2023; Azam and Liu, 2025). Concurrently, artificial intelligence algorithms (Alaswad and Almatrafi, 2024; Mahato et al., 2024) and uncertainty analysis techniques have been introduced to enhance the robustness and reliability of model outputs (Shi et al., 2024).

Although research on water environmental capacity in China began relatively late, substantial progress has been made since the early 21st century, particularly in major river basins such as the Yangtze River Economic Belt and the Yellow River Basin. A relatively mature framework for capacity estimation and pollutant load allocation has since been developed. Current research trends include (1) transitioning from static, single-section estimation to dynamic, multi-section, multi-source coordinated control (Mou et al., 2019); (2) expanding from point-source control to integrated management that includes non-point sources (Ma et al., 2023); (3) promoting the integration of environmental capacity with ecological redlines, functional zoning (Wen and Wang, 2023), and emissions trading schemes (Torres-Bejarano et al., 2022). Despite these advancements, several challenges persist. These include strong dependence on monitoring data for model calibration, misalignment between water quality targets and functional zoning, oversimplification in hydrodynamic simulations, and the absence of ecological response mechanisms in capacity outputs (Zhang et al., 2020). In regions with complex river networks—particularly at confluences of inlets and outlets—uncertainties in hydrodynamics can significantly impact simulation accuracy. Thus, there is an urgent need for refined, high-resolution, multi-variable coupled estimation methods (Wu et al., 2022).

This study investigates the Ganjiang River estuary, a representative multi-branch river system. A two-dimensional coupled hydrodynamic–water quality model was developed to simulate pollutant transport, identify major pollution sources, and estimate the environmental capacity for key pollutants. A novel dynamic estimation method was proposed, integrating grid-level static capacity with water exchange periods to evaluate the actual spatiotemporal assimilative capacity of the water body. The study aims to support water quality compliance and ecological safety management in branching river systems while providing a scalable technical framework for dynamic environmental capacity assessment in hydrologically complex regions.

Traditional dynamic WEC estimation methods commonly rely on process-based models to simulate pollutant transport and degradation under varying hydrological conditions. However, such approaches are often constrained by their high model complexity and extensive parameter requirements. In contrast, this study proposes a simplified yet robust framework with three key innovations: (1) high-resolution static WEC estimation across 14,600 computational grid cells, enabling fine-scale spatial differentiation of hydrodynamic conditions; (2) aggregation of grid-level capacities in accordance with legally designated water functional zones to ensure spatial policy alignment; and (3) incorporation of region-specific water exchange periods to derive temporally dynamic capacities that are directly compatible with daily discharge permitting schemes.

2 Materials and methods

2.1 Study area overview

Poyang Lake, the largest freshwater lake in China (Qin et al., 2023), is fed by five major rivers: the Ganjiang, Raohe, Xiushui, Fuhe, and Xinjiang Rivers (Wang et al., 2019). Among these, the Ganjiang River is the largest tributary in the Poyang Lake Basin, with a drainage area of approximately 83,500 km², accounting for about 51% of Jiangxi Province’s total land area (Fu, 2024).

The Ganjiang River estuary, located in the river’s lower reaches near Nanchang, serves as the final discharge segment before the river enters Poyang Lake (Hu H. et al., 2021). Downstream of the Bayi Bridge, the river bifurcates at Yangzizhou—about 2 km from the bridge—into two primary channels: the East River and West River. The West River subsequently splits into the Main Branch and North Branch, while the East River divides into the Middle Branch and South Branch. Consequently, the Ganjiang River discharges into Poyang Lake through four distributaries: the Main Branch, North Branch, Middle Branch, and South Branch (Hu H. et al., 2021).

The estuarine zone includes one hydrological station (Waizhou Station) and five water level stations—Nanchang, Changyi, Jiangbu, Louqian, and Chucha—as shown in Figures 1a–c.

Figure 1

Figure 1. Geographical location of the study area, which illustrate the national location (a), provincial context (b), and the estuarine monitoring stations (c).

Within this region, six national- and provincial-level water quality monitoring sections have been established. The area is functionally zoned for various uses, including industrial water supply, recreation and scenic purposes, fisheries, and drinking water source protection. According to the regional water environmental functional zoning, the Main Branch, North Branch, and Middle Branch are required to meet Class III water quality standards, while the Main Branch and South Branch are subject to Class IV standards (China GB3838, 2002). The spatial distribution of water functional zones and monitoring sections is shown in Figure 2.

Figure 2

Figure 2. Water environmental functional zoning and distribution of water quality monitoring sections.

2.2 Water quality status

Water quality monitoring data from the Ganjiang River estuarine area in 2020 indicate that overall water quality in the study region was relatively good, with pollutant concentrations consistently meeting or exceeding the standards for their designated water functional zones. Specifically, annual mean concentrations of ammonia nitrogen (NH₃-N) and total phosphorus (TP) generally fell within Class II to Class III thresholds, while permanganate index (COD_Mn) levels ranged from Class III to Class IV. Based on comprehensive pollution index assessments, water quality at cross-sections D1, D3, D4, D5, and D6 was classified as “clean” to “slightly polluted” throughout the year. In contrast, cross-section D2 exhibited water quality ranging from “slightly polluted” to “mildly polluted.”

2.3 Pollution source inventory

Pollutants entering the Ganjiang River estuary originate from both point and non-point sources. Point-source pollution primarily includes domestic sewage and industrial effluents discharged from urban areas, while non-point sources consist mainly of domestic wastewater from rural settlements, agricultural runoff, and aquaculture discharges. Pollutant load estimation was conducted in accordance with the Technical Guidelines for National Water Environmental Capacity Assessment (Planning, 2003). Loads from domestic, industrial, agricultural, and aquaculture sources were calculated separately, with specific calculation methods detailed in Equations 1–10.

2.3.1 Domestic pollution load into river

$W_{dps\_1}=W_{dps\_1p}\times {\beta }_2(1)$

$W_{dps\_1}=W_{rur-N}\times {\alpha }_1(2)$

$W_{dps\_2}=\left(W_{dps\_2p}-{\theta }_2\right)\times {\beta }_3(3)$

$W_{dps\_2p}=N_{city}\times {\alpha }_2(4)$

In these equations, $W_{\text{dps}\_1}$ and $W_{\text{dps}\_2}$ represent the rural and urban domestic pollution loads entering the river, respectively. $W_{\text{dps}\_1p}$ and $W_{\text{dps}\_2p}$ are the total pollutant discharges from rural and urban domestic sources. The coefficients ${\beta }_2$ and ${\beta }_3$ represent the fractions of rural and urban sewage entering the river. Values of 0.5 and 0.8 were adopted for ${\beta }_2$ and ${\beta }_3$, respectively, based on field conditions and treatment efficiency. $W_{\text{rur}-N}$ and $N_{\text{city}}$ denote the rural and urban populations, while ${\alpha }_1$ and ${\alpha }_2$ are the corresponding per capita pollutant generation coefficients, as detailed in Table 1. ${\theta }_2$ is the amount of pollutant removed through urban sewage treatment, The values of the above coefficients were determined with reference to relevant standards (Planning, 2003) and literature (Yu et al., 2023).

Table 1

Table 1. Emission coefficient of domestic pollution sources (unit: g/(person·day)).

For this study, daily water generation rates were assumed to be 150 L/(person·day) for urban areas and 90 L/(person·day) for rural areas.

2.3.2 Farmland pollution load into river

$W_{farm}=W_{farm-p}\times {\beta }_4\times {\gamma }_1(5)$

$W_{farm-p}=M\times {\alpha }_3(6)$

In these expressions, $W_{farm}$ represents the farmland pollution load entering the river, and $W_{farm-p}$ is the total pollutant discharge from agricultural land. The coefficient ${\beta }_4$, set to 0.15 in this study, denotes the proportion of farmland-generated pollutants that reach the river. ${\gamma }_1$ The correction factor ${\gamma }_1$, which accounts for variations in fertilizer application intensity, was set to 0.8, based on regional conditions and literature recommendations (Jin et al., 2022). $M$ is the total cultivated area, and ${\alpha }_3$ is the farmland discharge coefficient, as provided in Table 2.

Table 2

Table 2. Farmland discharge coefficient (unit: kg/(acre·year)).

According to regional agricultural practices, the average annual fertilizer application in the study area is approximately 20 kg per mu, justifying the use of a correction factor of 0.8.

2.3.3 Aquaculture pollution load into the river

$W_{poul}=W_{poul-p}\times {\beta }_5(7)$

$W_{poul-p}={\delta }_1\times t\times N_{poul}\times {\alpha }_4+{\delta }_2\times t\times N_{poul}\times {\alpha }_5(8)$

$W_{aqu}=W_{aqu-p}\times {\beta }_6(9)$

Here, $W_{poul}$ is the pollution load from livestock and poultry farming entering the river, and $W_{poul-p}$ is the total discharge from these sources. The coefficient ${\beta }_5$, representing the proportion of pollutants from livestock and poultry farms entering the river, was set at 0.2. In Equation 8, ${\delta }_1$ is the average pollutant content in animal feces (g/day), $t$ is the average feeding period (days), $N_{poul}$ is the number of animals farmed, and ${\alpha }_4$ is the average daily feces production per animal (kg/day). ${\delta }_2$ The second term accounts for urine: ${\delta }_2$ is the average pollutant content in urine (g/kg), and ${\alpha }_5$ is the average daily urine production per animal (kg/day). In Equation 9, $W_{aqu}$ denotes the pollution load from aquaculture entering the river, and $W_{aqu-p}$ is the total discharge from aquaculture. The coefficient ${\beta }_6$, representing the fraction of aquaculture-derived pollutants entering the river, was set at 0.3.

All coefficients were adopted from national technical guidelines and relevant literature (Bui and Pham, 2023).

2.3.4 Industrial pollution load into river

$W_{ind}=\left(W_{ind-p}-{\theta }_1\right)\times {\beta }_1(10)$

In this expression, $W_{ind}$ represents the industrial pollution load entering the river, and $W_{ind-p}$ is the total direct discharge of industrial pollutants. The term ${\theta }_1$ denotes the amount of pollutant removed by wastewater treatment processes. ${\beta }_1$ is the coefficient for industrial pollutant discharge into the river (set to 1.0 in this study, meaning that all industrial pollutants not removed by wastewater treatment plants are assumed to enter the river).

2.4 Hydrodynamic–water quality model construction and calibration

2.4.1 Governing equations

A two-dimensional hydrodynamic–water quality coupled model was employed to simulate hydrodynamics and pollutant transport in the Ganjiang River estuary. The model is based on the shallow water equations and principles of mass and momentum conservation, and it integrates a pollutant mass balance equation to capture transport, diffusion, and transformation processes. This modeling approach is particularly well suited for regions with complex hydraulic behavior and spatially heterogeneous pollutant loads, such as river bifurcations and lake–river transition zones. The governing equations for hydrodynamics are provided in Equations 11–13 (Yang et al., 2019; Yang et al., 2022), while the pollutant transport process, primarily governed by the advection–diffusion equation, is given in Equation 14.

$\frac{\mathrm{\partial }\mathrm{h}}{\mathrm{\partial }t}+\frac{\mathrm{\partial }\left(u\mathrm{h}\right)}{\mathrm{\partial }x}+\frac{\mathrm{\partial }\left(v\mathrm{h}\right)}{\mathrm{\partial }x}=q(11)$

$\frac{\mathrm{\partial }\left(u\mathrm{h}\right)}{\mathrm{\partial }t}+\frac{\mathrm{\partial }\left(u^2\mathrm{h}\right)}{\mathrm{\partial }x}+\frac{\mathrm{\partial }\left(uv\mathrm{h}\right)}{\mathrm{\partial }y}=-g\mathrm{h}\frac{\mathrm{\partial }\eta }{\mathrm{\partial }x}+f_x-\frac{T_{bx}}{\rho }(12)$

$\frac{\mathrm{\partial }\left(V\mathrm{h}\right)}{\mathrm{\partial }t}+\frac{\mathrm{\partial }\left(uv\mathrm{h}\right)}{\mathrm{\partial }x}+\frac{\mathrm{\partial }\left(V^2\mathrm{h}\right)}{\mathrm{\partial }y}=-g\mathrm{h}\frac{\mathrm{\partial }\eta }{\mathrm{\partial }y}+f_y-\frac{T_{by}}{\rho }(13)$

Here, $\mathrm{h}$ represents the water depth (m); $u$ and $v$ denote the flow velocities in the x- and y-directions (m/s), respectively; $\eta$ denotes the water surface elevation (m); $g$ denotes the gravitational acceleration (m/s²); $q$ denotes the external source or sink terms (m/s); $T_{bx}$ and $T_{by}$ denote the bottom shear stress components in the x- and y-directions, respectively (N/m²); $\rho$ denotes the water density (kg/m³); $f_x$ and $f_y$ denote Coriolis and other force terms.

$\frac{\mathrm{\partial }C}{\mathrm{\partial }t}+u\frac{\mathrm{\partial }C}{\mathrm{\partial }x}+v\frac{\mathrm{\partial }C}{\mathrm{\partial }y}=E_x\frac{{\mathrm{\partial }}^{2}C}{\mathrm{\partial }{x}^2}+E_y\frac{{\mathrm{\partial }}^{2}C}{\mathrm{\partial }{y}^2}-KC+S(14)$

where $C$ denotes the pollutant concentration (mg/L); $E_x$ and $E_y$ are the longitudinal and transverse dispersion coefficients (m²/s), respectively; $K$ denotes the first-order decay coefficient (1/s); $S$ denotes external source or sink term (e.g., pollutant inputs).

2.4.2 Model setup

A two-dimensional coupled hydrodynamic–water quality model was constructed for the Ganjiang River estuarine region, integrating both hydrodynamics and pollutant transport processes.

1. Daily inflow discharges and monthly average concentrations of COD_Mn, NH₃-N, and TP measured at Waizhou Station (M1) were used as upstream boundary conditions. Downstream boundary conditions were defined using daily water level data from Changyi Station (M3), Jiangbu Station (M4), Louqian Station (M5), and Chucha Station (M6). The spatial distribution of these monitoring stations is shown in Figure 1. The computational domain was discretized using a combination of triangular, rectangular, and hybrid mesh elements, resulting in a total of 14,600 grid cells.

2. The simulation period extended from 1 January 2020, to 1 January 2021, totaling 8,784 hourly time steps. To maintain numerical stability, the time step varied between 0.01 s and 60 s, with a Courant–Friedrichs–Lewy (CFL) number set to 0.8. Wind forcing was incorporated using daily meteorological data from 2020. A dynamic wind drag coefficient was applied based on wind speed: 0.001255 for speeds below 7 m/s, 0.002425 for speeds exceeding 25 m/s, and linearly interpolated values for intermediate wind speeds.

3. Given the complex morphology of the river network, sediment–flow interactions, and the influence of riparian vegetation—particularly under seasonally variable hydrological conditions (wet, normal, and dry periods)—a spatially and temporally variable Manning’s roughness coefficient was employed. The regional zoning and seasonal roughness assignments are shown in Figure 3.

Figure 3

Figure 3. Spatial distribution of Manning’s roughness coefficients. (a) Wet season. (b) Normal season. (c) Dry season.

2.4.3 Model calibration

The hydrodynamic and water quality modules were calibrated using field data, including water level, flow velocity, discharge, and concentrations of COD_Mn, NH₃-N, and TP.

2.4.3.1 Water level calibration and validation

Model calibration and validation were performed using water level data from Nanchang Station for the years 2020 and 2012, respectively. The results show that the average relative error during calibration was about 5%–18%, while the validation average relative error was within 15% (Figure 4).

Figure 4

Figure 4. Water level calibration (2020) and Validation (2012) at the Nanchang Station. (a) Water level validation. (b) Water level calibration.

2.4.3.2 Flow velocity and discharge calibration

Calibration of flow velocity and discharge was conducted using field measurements from two cross-sections: Changyi–Ganjiang Bridge and Louqian Bridge, collected on 19 November 2020. The comparison between simulated and observed values indicated good overall agreement. The relative error in flow velocity ranged from 5% to 13%, while discharge errors varied between 7% and 16%, satisfying standard calibration precision criteria. Detailed calibration curves are available in a separate study by the authors (Lu et al., 2024).

2.4.3.3 Water quality calibration

Degradation coefficients for COD_Mn, NH₃-N, and TP in the main channel of the Ganjiang River and its four distributaries were determined based on field measurements and literature references, as summarized in Table 3 (China GB3838, 2002). Pollution sources included both point and non-point contributors; point-source locations are shown in Figure 1, while non-point sources (e.g., agricultural runoff) were modeled as spatially uniform inputs driven by rainfall. Using 2020 pollutant inventory data and monthly monitoring records, five representative cross-sections (D2, D3, D4, D5, and D6; see Figure 2 for locations) were selected for model calibration. The comparison between simulated and observed concentrations is shown in Figure 5. For most pollutants at these sections, relative errors were within 20%. Notable exceptions included COD_Mn at D5 (23.7%) and NH₃-N at D4 and D6 (23.4% and 21.3%, respectively). Overall, the model demonstrated satisfactory performance, effectively capturing the spatiotemporal patterns of water quality in the Ganjiang River estuary.

Table 3

Table 3. Degradation and dispersion coefficients for water quality parameters.

Figure 5

Figure 5. Calibration results of water quality parameters.

2.5 Sensitivity and uncertainty analysis

2.5.1 Sensitivity analysis

To evaluate the influence of model parameters on the simulated concentrations of COD_Mn, NH₃-N, and TP, perturbations of ±10% and ±20% were applied to pollutant loads, dispersion coefficients, and degradation coefficients. A relative change rate method was employed to quantify sensitivity, identifying the parameters with the most significant impact on simulation results. The calculation formulas are presented in Equations 15–17.

$SI=\frac{\Delta C/C_0}{\Delta P/P_0}(15)$

$\Delta C/C_0=\frac{C_{pert}-C_{base}}{C_{base}}\times 100\%(16)$

$\Delta P/P_0=\frac{P_{pert}-P_{base}}{P_{base}}\times 100\%(17)$

where $SI$ represents the sensitivity index; $C_{pert}$ is the simulated pollutant concentration under perturbation (calculated as the average of concentrations at monitoring sites D1–D6); $C_{base}$ is the observed mean concentration from the six monitoring sites; $P_{pert}$ and $P_{base}$ denote the perturbed and baseline parameter values, respectively.

2.5.2 Uncertainty analysis

Building upon the sensitivity analysis, the Monte Carlo Simulation (MCS) method was applied to assess the uncertainty of simulated COD_Mn, NH₃-N, and TP concentrations. Specifically, pollutant load, dispersion coefficient, and degradation coefficient were assumed to fluctuate within ±20% of their baseline values, following a uniform distribution. Under this assumption, multiple sets of parameter combinations were generated via random sampling, and the model was run to obtain simulated concentrations for each set. For each pollutant, the mean, standard deviation, coefficient of variation (CV), and 95% confidence intervals were calculated using the average concentration across the six monitoring sites.

2.6 Estimation of water environmental capacity

2.6.1 Static water environmental capacity estimation

As detailed in the hydrodynamic model setup, the study domain was discretized into 14,600 computational grids. These were further grouped into six subregions based on designated water functional zoning (see Figure 2). For each grid cell, daily water level and pollutant concentration outputs were extracted. Using these values and the corresponding water quality thresholds defined for each functional zone, the static environmental capacity was estimated at the grid scale.

The grid-based static capacity was calculated using Equations 18–20, and subsequently aggregated across time and space to obtain the total static capacity for the study area. The applicable water quality standard limits for each pollutant, differentiated by functional zone, are provided in Table 4.

$V_{i,j,t}=S_{i,j}\times \left(H_{i,j,t}-{\mathrm{h}}_{i,j}\right)(18)$

$W_{i,j,t}^{static}=V_{i,j,t}\times \left(C_{std,i}-C_{i,j,t}\right)\times {10}^{-6}(19)$

$W_{total,t}^{static}={\displaystyle \sum _{i=1}^n}{\displaystyle \sum _{j=1}^{m_i}}{W}_{i,j,t}^{static}(20)$

where $S_{i,j}$ denotes the surface area of the j-th grid in the i-th functional zone (m²); $H_{i,j,t}$ denotes the simulated water surface elevation of grid j at time $t$ (m), ${\mathrm{h}}_{i,j}$ denotes the riverbed elevation of grid j in zone i (m); $V_{i,j,t}$ denotes the water volume of grid j at time $t$ (m³); $C_{std,i}$ denotes the standard concentration limit of the pollutant in zone i (mg/L); $C_{i,j,t}$ denotes the simulated concentration in grid j at time $t$ (mg/L); $W_{i,j,t}^{static}$ denotes the static water environmental capacity of grid j at time $t$ (t); $m_i$ denotes the number of grids in functional zone $i$; $n$ denotes the total number of functional zones; $W_{total,t}^{static}$ denotes the total static capacity of the study area at time $t$ (t).

Table 4

Table 4. Standard limits of surface water environmental quality parameters (mg/L).

2.6.2 Dynamic water environmental capacity calculation

From the perspective of physical mechanisms, the static environmental capacity refers to the maximum pollutant load that a water body can accommodate under relatively stationary conditions while meeting the target water quality standards. However, in actual hydrological settings, water bodies are continuously subjected to dynamic exchange and renewal processes, and the static capacity fails to reflect the self-purification processes and system evolution.

To better represent the self-purification potential of water bodies and the turnover rate of pollutants, this study introduces the concept of “water exchange period,” which denotes the time (in days) required for the complete renewal of water within a reservoir. By expressing the water exchange period on a monthly basis, the static capacity is converted into a time-scaled value, resulting in a time-varying dynamic environmental capacity. The calculation method are presented in Equations 21, 22.

$T_{mont\mathrm{h},n}=\frac{V_{aveg,mont\mathrm{h},n}}{{86400\times Q}_{in,mont\mathrm{h},n}}(21)$

$W_n^{dynamic}=W_{total,t}^{static}/T_{mont\mathrm{h},n}(22)$

where $T_{mont\mathrm{h},n}$ denotes the water exchange period (days) of the nth subregion during a given month; $Q_{in,mont\mathrm{h},n}$ denotes the average monthly inflow discharge into subregion $n$ (m³/s); $V_{aveg,mont\mathrm{h},n}$ denotes the average water volume of subregion $n$ (m³); $W_n^{dynamic}$ denotes the dynamic environmental capacity of subregion $n$ for the given month (t/day).

The dynamic environmental capacity derived from Equation 22 is expressed in tons per day (t/d), indicating the amount of pollutant that can be assimilated daily under specific hydrodynamic conditions. A shorter water exchange period implies a faster system renewal rate and stronger self-purification capacity, thereby leading to a greater pollutant-bearing capacity per unit time. This method provides a more realistic representation of the transport and attenuation mechanisms of pollutants and facilitates alignment with discharge permit indicators at a daily temporal resolution.

3 Results and analysis

3.1 Riverine pollutant loads

Based on five primary pollution source categories—namely aquaculture, livestock farming, agricultural runoff, industrial wastewater, and domestic sewage—this study estimated the annual pollutant loads of COD_Mn, NH₃-N, and TP discharged into the Ganjiang River estuary in 2020. The results are presented in Figures 6, 7.

Figure 6

Figure 6. Total annual pollutant loads entering the river in the study area.

Figure 7

Figure 7. Proportional contribution of different sources to pollutant loads in the study area. (a) COD_Mn. (b) NH₃-N. (c) TP.

As illustrated, domestic sewage emerged as the dominant source of pollutant input in the study area. The annual contributions from this source reached 14666.12 t for COD_Mn, 1807.86 t for NH₃-N, and 107.61 t for TP, accounting for 75.15%, 79.92%, and 42.07% of the total pollutant loads, respectively. Livestock farming was identified as the second-largest contributor, particularly for TP, where the load amounted to 97.90 t, representing 38.27% of the total TP input. These findings highlight the significant role of non-point-source pollution from animal husbandry.

Industrial wastewater also contributed notably to TP and NH₃-N pollution, with annual loads of 33.56 t (13.12%) and 95.92 t (4.24%), respectively. In comparison, agricultural runoff accounted for 565.96 t of COD_Mn, 73.52 t of NH₃-N, and 15.71 t of TP, representing 2.90%, 3.25%, and 6.14% of the total inputs. The contribution from aquaculture was minimal across all three pollutants, with each comprising less than 1% of the total annual load.

In summary, the pollutant input structure in the Ganjiang River estuary exhibited a clear hierarchy: domestic sewage was the predominant source, followed by livestock farming, with industrial wastewater and agricultural runoff serving as secondary contributors, and aquaculture having the least impact. These findings underscore the need for targeted pollution control strategies. It is recommended that future water environment management efforts prioritize the construction and upgrading of domestic sewage treatment infrastructure. Simultaneously, non-point-source pollution from agriculture should be addressed through region-specific, seasonally adaptive, and source-oriented measures, in order to effectively enhance the environmental carrying capacity of the Ganjiang River estuary.

3.2 Temporal variations in storage capacity

Using the methodology described previously, the monthly average water storage volumes were computed for each designated functional subregion (Regions I–VI) within the study area for the year 2020. These values were then aggregated to estimate the total monthly average storage capacity of the entire Ganjiang River estuary, as presented in Figure 8.

Figure 8

Figure 8. Monthly average water storage capacity of the study area.

The results reveal distinct seasonal variations in total storage capacity over the year. In January, the total storage volume was approximately 3.98 × 10⁸ m³. A steady increase was observed from February to July, with volumes rising from 2.69 × 10⁸ m³ in February to a peak of 9.03 × 10⁸ m³ in July. This upward trend reflects the enhanced storage capacity associated with elevated inflows and rising water levels during the wet season.

From August to December, the storage volume declined sharply, reaching approximately 2.14 × 10⁸ m³ in December, indicative of reduced inflows and lower water levels during the dry season.

Overall, the annual variation in storage capacity corresponds closely with the regional hydrological regime, exhibiting a typical pattern of expansion during high-flow periods and contraction during low-flow periods.

3.3 Model sensitivity and uncertainty

Figure 9 presents the sensitivity responses of COD_Mn, NH₃-N, and TP concentrations to perturbations of pollution load, dispersion coefficient, and degradation coefficient (±10% and ±20%). As shown, for COD_Mn, a 20% increase in pollutant load led to a significant rise in concentration (+18.7%), while a 20% decrease caused a reduction of −14.2%. Similar positive correlations were observed for NH₃-N and TP. When the dispersion coefficient was increased by 20%, COD_Mn and TP concentrations decreased by −4.7% and −8.0%, respectively. Increasing the degradation coefficient by 20% resulted in decreases of −5.0% for COD_Mn and −5.7% for NH₃-N. Overall, pollutant load exerted the strongest influence on concentration changes, followed by the dispersion coefficient, while the degradation coefficient had the weakest effect.

Figure 9

Figure 9. Sensitivity responses of water quality indicators to perturbations in pollution load, dispersion coefficient, and degradation coefficient. (a) COD_Mn. (b) NH₃-N. (c) TP.

Uncertainty analysis results are summarized in Table 5. For COD_Mn, the simulated mean concentration was 1.95 mg/L, close to the observed mean of 2.18 mg/L. The 95% confidence interval ranged from 1.55 to 2.63 mg/L, with a standard deviation of 0.27 mg/L and a coefficient of variation (CV) of 0.14. This indicates that although absolute fluctuations were relatively large, the relative variability was moderate. For NH₃-N, the simulated mean was 0.24 mg/L (observed value: 0.25 mg/L), with a 95% confidence interval of 0.19–0.30 mg/L, a standard deviation of 0.03 mg/L, and a CV of 0.13. For TP, the simulated mean matched the observed value (0.06 mg/L), with a 95% confidence interval of 0.048–0.072 mg/L, a standard deviation of 0.007 mg/L, and a CV of 0.12, suggesting that the TP predictions were the most robust among the three pollutants.

Table 5

Table 5. Statistical results of uncertainty analysis for simulated pollutant concentrations.

In summary, the uncertainty analysis confirmed that pollutant load was the dominant source of variability for COD_Mn, while NH₃-N and TP exhibited relatively weaker overall responses to parameter perturbations. Since all three pollutants showed CV values below 0.15, the model predictions can be considered highly reliable.

3.4 Environmental capacity estimation and intra-annual variations

To evaluate the environmental carrying capacity of the Ganjiang River estuary for key pollutants—COD_Mn, NH₃-N, and TP—a validated two-dimensional hydrodynamic–water quality coupled model was employed. This model was integrated with daily water storage statistics and the pollutant threshold values assigned to each water functional zone. The study area was divided into six subregions (Regions I–VI), each governed by specific water quality target standards. For each subregion, the monthly environmental capacity was calculated by aggregating three components at the grid-cell level: (i) the simulated pollutant concentration, (ii) the corresponding storage volume, and (iii) the applicable water quality standard. These sub-regional capacities were then summed to obtain the total monthly environmental capacity of the estuary.

Figure 10 presents the monthly average water exchange periods for the study area, while Figures 11–13 depicts the monthly variation in environmental capacity across the six subregions. Among these, Region III consistently exhibited the highest pollutant carrying capacity throughout the year. Peak values were observed during June–July, with the COD_Mn capacity reaching approximately 1,626 t/day, markedly higher than in other subregions. This elevated capacity can be primarily attributed to Region III’s larger water storage volume and comparatively favorable simulated water quality conditions.

Figure 10

Figure 10. Monthly average water exchange period in the study area.

Figure 11

Figure 11. Environmental capacity of COD_Mn in each subregion.

Figure 12

Figure 12. Environmental capacity of NH₃-N in each subregion.

Figure 13

Figure 13. Environmental capacity of TP in each subregion.

In contrast, Regions I and VI demonstrated moderate environmental capacity during the high-flow season (May–August). Region IV, however, consistently showed the lowest environmental capacity across all months.

Figure 14 illustrates the annual trend of total environmental capacity for the entire Ganjiang River estuary. The results reveal distinct seasonal variations in the estuary’s assimilative capacity for all three pollutants, generally following a “low–high–low” pattern that aligns with fluctuations in water storage.

Figure 14

Figure 14. Intra-annual variation in total environmental capacity of the study area.

For COD_Mn, a bimodal distribution was observed, with the lowest capacity recorded in November (228.92 t/day). As inflows increased and storage expanded during the wet season, the capacity rose sharply from February, reaching peaks in June (4307.50 t/day) and July (4266.52 t/day). A marked decline followed during August–December, attributed to diminished hydrodynamic conditions and extended water residence time.

The intra-annual variation for NH₃-N was even more pronounced. The minimum capacity occurred in December (36.66 t/day), while the peak was recorded in June (858.75 t/day)—representing a 23.42-fold increase. A capacity plateau was observed from May to July, whereas other months exhibited substantially lower levels. These results highlight the strong dependence of NH₃-N assimilative capacity on flow dynamics and hydrological conditions.

TP exhibited the most pronounced seasonal fluctuations among the three pollutants. The lowest carrying capacity was recorded in December (6.93 t/day), while the peak occurred in June (149.00 t/day), followed closely by July (148.95 t/day). A sharp decline began in August. Compared to COD_Mn and NH₃-N, the environmental capacity for TP was notably more sensitive to seasonal hydrological variability, underscoring its propensity to accumulate and exceed regulatory thresholds in aquatic systems.

In summary, the June–July period constitutes a critical “buffer window” for pollutant mitigation and discharge management in the Ganjiang River estuary, owing to the peak assimilative capacities during this time. In contrast, the period from December to February is characterized by significantly diminished environmental capacity, thereby presenting an elevated risk of pollutant exceedance. This interval should thus be prioritized for intensive pollution control measures and regulatory interventions.

3.5 Comparison between the conventional static method and the proposed approach

To verify the improvements of the proposed dynamic method for calculating water environmental capacity, the conventional static method was also applied to COD_Mn, NH₃-N, and TP for comparison. The conventional static capacity calculation method was adopted with reference to relevant literature (Zhang et al., 2020), and the results are shown in Figure 14 as “COD_Mn (conventional static),” “NH₃-N (conventional static),” and “TP (conventional static).”

Analysis of Figure 12 indicates that during the high-flow period (June–July), the static method significantly underestimated the carrying capacity; for example, COD_Mn in June was underestimated by approximately 12.9%. In contrast, during the low-flow period (October–December), the static method generally overestimated the capacity, with the maximum overestimation reaching about 9.5%. Compared with this, the dynamic method successfully captured the pronounced seasonal fluctuations. On an annual scale, the capacities estimated by the static method were generally about 80%–85% of those obtained by the dynamic method.

3.6 Spatial distribution characteristics of environmental capacity

To elucidate the spatial distribution patterns of water environmental capacity in the Ganjiang River estuary, the study area was divided into five principal hydraulic sub-branches: the main stem of the Ganjiang River, the primary branch, the northern branch, the central branch, and the southern branch. The monthly proportions of environmental capacity for each pollutant across these sub-branches were calculated and are presented in Figure 15.

Figure 15

Figure 15. Proportion of water environmental capacity for different pollutants across hydraulic sub-branches in the Ganjiang River estuary. (a) COD_Mn. (b) NH₃-N. (c) TP.

As shown in Figure 15a, the primary branch consistently exhibited the highest carrying capacity for COD_Mn, contributing more than 50% of the total environmental capacity throughout the year, with an annual average of 52.11%. This was followed by the Ganjiang main stem, which accounted for 17%–21%. The southern and central branches maintained relatively stable shares of approximately 21% and 8%, respectively. In contrast, the northern branch consistently contributed less than 1%, reflecting its limited hydrological extent and flow conditions. These spatial patterns closely align with the water storage capacity distribution among the subregions, underscoring the dominant role of the primary branch in determining the overall assimilative capacity of the estuary.

As shown in Figure 15b, the spatial distribution of NH₃-N capacity exhibits a more complex pattern compared to COD_Mn. The primary branch continues to serve as the dominant contributor, with an average annual share of 50.68%, followed by the main stem, which contributes approximately 23.33%. In contrast, the southern and central branches demonstrate notable temporal variability, with their capacity shares fluctuating markedly across months. This indicates spatial and seasonal heterogeneity in assimilative performance. The northern branch consistently maintains the lowest contribution. Notably, during the dry season (e.g., December), the main stem shows a significant increase in its share of NH₃-N capacity, suggesting relatively enhanced dilution performance under low-flow conditions.

In Figure 15c, the spatial distribution of TP capacity closely parallels that of COD_Mn. The primary branch exhibits the highest average annual share (approximately 52.63%), followed by the main stem (22.36%) and the southern branch (approximately 16.09%), both of which demonstrate relatively stable performance throughout the year. In contrast, the central branch contributes only minimally to TP capacity. The spatial concentration of TP in the primary and southern branches highlights these regions as priority zones for internal load management and external inflow pollution control measures.

In summary, the spatial distribution of environmental capacity in the Ganjiang River estuary is primarily governed by branch-specific hydrodynamic conditions, baseline water quality, and designated functional zoning. The primary branch and the main stem emerge as the principal capacity-bearing zones and should therefore be prioritized in future water quality regulation and discharge management strategies. The southern branch, given its moderate capacity and TP enrichment, warrants enhanced control of non-point-source pollution. In contrast, the central and northern branches, due to their limited assimilative capacities, should be subject to strict limitations on new discharge outlets and stringent control of pollutant loading to prevent localized exceedances.

4 Discussion

This study employed a two-dimensional hydrodynamic–water quality coupled model to simulate the annual hydrological and water quality dynamics of the Ganjiang River estuary and to quantify the water environmental capacities for COD_Mn, NH₃-N, and TP. The results revealed distinct seasonal fluctuations in environmental capacity, following a typical “low in winter–high in summer–declining in autumn” pattern. For example, COD_Mn capacity peaked at 4307.50 t/day in June, nearly 17 times higher than the lowest value recorded in December (228.92 t/day). Even more pronounced seasonal variations were observed for NH₃-N and TP, highlighting the high sensitivity of the estuary’s self-purification potential to hydrological variability and seasonal inflows. These findings underscore the need for seasonally adaptive pollution control strategies. During the high-flow season (June–July), emission reduction and flow regulation measures can take advantage of the enhanced assimilative capacity to improve water quality management efficiency. Conversely, the low-flow period requires stricter total load control, proactive pollutant reduction, and early warning mechanisms to mitigate the elevated risk of water quality exceedances.

Traditional approaches to estimating water environmental capacity often rely on static assumptions, employing empirical formulas or simplified models to approximate a water body’s pollutant dilution potential while overlooking the effects of spatiotemporal variability (Zhang et al., 2016; Huang et al., 2019). In contrast, this study presents a dynamic, grid-based capacity estimation method grounded in a hydrodynamic–water quality coupled model. By extracting daily simulated water quality data and integrating grid-specific water volumes with water quality standards defined for functional zones, the study establishes a systematic framework for grid-level computation, regional aggregation, and temporal extrapolation, enabling real-time estimation of dynamic environmental capacity.

Moreover, a zonal capacity accounting strategy is proposed, based on water environmental functional zoning. This addresses the discrepancy between heterogeneous functional requirements and the need for standardized capacity assessments. Given that different zones are subject to distinct pollutant concentration thresholds, a uniform calculation across the entire study area would introduce significant bias. Instead, calculating environmental capacity independently within each zone and subsequently aggregating the results provides a more accurate and regulation-aligned method. This approach supports targeted management, offering technical support for differentiated emission permitting and zone-specific compliance assessment. Its effectiveness in handling multi-branch, multi-functional water systems highlights its broad applicability and scalability in complex aquatic environments.

Lastly, the proposed method represents a fundamental departure from traditional “theoretical capacity” estimation approaches. Conventional models frequently neglect actual pollutant inputs and in-stream transformation processes, resulting in idealized maximum carrying capacities that are often significantly overestimated in real-world applications. In contrast, the water quality simulation framework adopted in this study explicitly incorporates external pollutant loads, hydrodynamic transport, and biogeochemical processes, thereby ensuring that the estimated capacity reflects the realistic residual assimilative potential under current pollution conditions. This shift from theoretical to practically available capacity offers a more operationally meaningful basis for setting pollutant load reduction targets and issuing discharge permits.

The Ganjiang River estuary showed pronounced seasonal fluctuations in environmental capacity, with high-flow season peaks exceeding low-flow values by more than tenfold. This pattern highlights the need for seasonally adaptive management. Discharge permitting should shift from annual averages to differentiated seasonal strategies to avoid exceedances in low-flow periods. Industrial and municipal discharges should be synchronized with capacity variation—moderately adjustable during high-flow months but strictly limited from December to February, with staggered and time-specific regulation. Agricultural non-point sources also require targeted control in the high-flow season to prevent enhanced runoff inputs from offsetting assimilative gains. Incorporating dynamic capacity estimation into basin-wide load control and functional zone assessments would facilitate the transition from annual to seasonal management, yielding more robust discharge standards and water quality objectives. These fluctuations are primarily hydrologically driven: increased precipitation and inflows in the high-flow season shorten water exchange periods and enhance dilution, whereas reduced inflows and longer residence times in the dry season suppress self-purification.

This study estimated capacity mainly from pollutant concentration dynamics, effectively capturing spatiotemporal patterns but not ecological responses such as dissolved oxygen or biodiversity. This limits direct linkage to ecosystem health. Future work should integrate ecological modules or multi-indicator frameworks to connect pollutant loads with endpoints such as fish habitat suitability or benthic biodiversity, strengthening ecological interpretability. Long-term monitoring is also needed to assess temporal trends and extend the method into an integrated “capacity–ecological response” tool. Parameter choices (e.g., degradation coefficient β, dispersion coefficient γ) were based on regional data and validated through calibration, but systematic field studies under varying hydrological regimes are required to improve robustness and predictive reliability.

It is worth noting that the proposed approach relies on a high-resolution hydrodynamic–water quality coupled model and presupposes detailed knowledge of pollution sources and functional zoning. As such, its applicability may be constrained in basins with complex hydro-geomorphological conditions or where observational data are limited.

From the perspective of hydrological and geomorphic features, this method is particularly suitable for medium-to-large river networks characterized by clearly delineated functional zones and relatively stable hydrodynamic regimes. In contrast, regions with dense tributary systems, complex confluences, or significant tidal influences often exhibit fluctuating boundary conditions, which may hinder the direct applicability of the method. Previous studies have advocated for a “subsection summation” strategy to evaluate environmental capacity in such complex reaches, aiming to account for heterogeneity in tributary inflows and zoning structure (Zhang et al., 2020). This concept aligns well with the control-section–function-zone coordination framework emphasized in this study, offering valuable insights for extending the method’s applicability.

In addition, the implementation of this approach depends heavily on the availability of in-situ data, including pollutant discharge loads, model calibration parameters, and long-term water quality observations. In data-sparse regions, model development and validation become substantially more challenging. To address this issue, auxiliary techniques such as parameter transfer from neighboring basins and satellite-based retrievals may be utilized as initial estimates (Hu et al., 2024). Previous studies proposed a “sensitivity–adaptivity co-optimization strategy,” in which the model structure and key parameters are iteratively adjusted to achieve optimal simulation accuracy under conditions of structural uncertainty (Burgan, 2021). This provides a robust framework for enhancing the generalizability of modeling efforts in regions lacking comprehensive datasets.

Furthermore, for watersheds exhibiting marked seasonal hydrological shifts (e.g., pronounced wet–dry season reversal) or substantially different pollution compositions compared to the Ganjiang River (e.g., snowmelt-driven basins or those dominated by agricultural non-point sources), additional adjustments are necessary with respect to boundary condition assignment and pollutant loading mechanisms. It is recommended that pollution loading schedules be regionally reconstructed, functional zoning schemes be tailored to local characteristics, and representative seasonal boundary conditions be incorporated to enhance the model’s forecasting capacity and relevance to policy implementation.

Compared with conventional dynamic WEC estimation methods, the proposed approach integrates high-resolution grid-based capacity estimation with functional zoning-based aggregation and further introduces water exchange periods as a time-adjustment mechanism. This allows for direct alignment of the results with permit management practices. For instance, in Functional Zone I, over 3,000 grid cells are involved, and their localized hydrodynamic characteristics are dynamically adjusted based on monthly exchange periods, yielding more realistic representations of daily assimilative capacity. Furthermore, this approach circumvents the need for fully parameterized process models, thereby enhancing both operational simplicity and policy applicability. Its streamlined structure and compatibility with regulatory spatial units make it more feasible for real-world water environment management.

External pollutant inputs in the study area are primarily concentrated in Region I and Region VI, and domestic sewage constitutes the dominant pollution source, accounting for approximately 75% of COD_Mn and 80% of NH₃-N loads. Accordingly, total load control and permitting strategies should incorporate temporal differentiation. For instance, during high-flow seasons when the branch-level capacity increases, daily discharge allowances for major municipal wastewater treatment plants could be moderately relaxed. Conversely, during dry seasons with sharply reduced capacity, stricter discharge caps or temporary storage strategies should be considered to avoid overloading the system. A dynamic linkage between assimilative capacity and discharge permitting can better align pollution control with actual environmental capacity.

In summary, the dynamic capacity assessment framework proposed in this study performs well in river basins where the model is rigorously calibrated, functional zoning is well-defined, and pollutant source data are sufficiently characterized. Looking forward, integrating the subsection-based evaluation logic with parameter self-adaptation mechanisms may facilitate broader application across basins with varying hydrological contexts and pollution regimes, ultimately strengthening its generalizability and decision-making utility.

5 Conclusion

In this study, a two-dimensional hydrodynamic–water quality coupled model was developed for the Ganjiang River estuary. Based on a comprehensive pollution source inventory for 2020, the model was used to simulate the spatiotemporal dynamics of COD_Mn, NH₃-N, and TP, and to propose a dynamic water environmental capacity estimation method that integrates traditional static capacity assessment with water exchange period data. The main conclusions are as follows:

1. The Ganjiang River estuary exhibits pronounced seasonal fluctuations in pollutant carrying capacity. June and July recorded the highest environmental capacities, with COD_Mn capacities of 4307.50 and 4266.52 t/day, NH₃-N capacities of 858.75 and 849.36 t/day, and TP capacities of 149.00 and 148.95 t/day, respectively. In contrast, the lowest capacities occurred in November and December: COD_Mn reached its minimum in November (228.92 t/day), while NH₃-N and TP both dropped to their lowest levels in December, at 36.66 t/day and 6.93 t/day, respectively. These findings underscore the strong influence of hydrological conditions on pollutant assimilation, following a characteristic “wet-season enhancement and dry-season decline” pattern.

2. The spatial distribution of environmental capacity follows the pattern: primary branch > main steam > other tributaries. The average daily environmental capacities over the year were 1677.37 t/d for COD_Mn, 324.02 t/d for NH₃-N, and 55.84 t/d for TP. The primary branch contributed 52.12% and 50.68% of the COD_Mn and NH₃-N capacities, respectively. For TP, the combined contribution from the primary branch and main stem approached 75%, highlighting their dominant role in regional pollutant assimilation.

3. This study introduced a zonal environmental capacity accounting framework based on water environmental functional zoning, integrated with outputs from the hydrodynamic–water quality coupled model. By incorporating the water exchange period at the grid scale, the method overcomes key limitations of traditional capacity estimation approaches, which often neglect pollutant loading intensity, in-stream decay, and hydrodynamic transport, the estimated capacities reflect the realistic residual assimilative potential under actual pollution scenarios.

4. The dynamic differences in environmental capacity provide a scientific basis for watershed water quality management and policy formulation. The high-flow season can serve as a “buffer window” for pollutant reduction and flow regulation, during which total load reduction or discharge scheduling can be appropriately arranged. In contrast, the low-flow season requires stricter discharge permitting and risk warning mechanisms to reduce the likelihood of exceedances. Furthermore, the proposed zonal capacity accounting framework offers technical support for differentiated emission permitting and zone-specific compliance assessment, facilitating the transition from annual-average control to season- and region-specific management, thereby enhancing the scientific rigor and operational feasibility of watershed water quality governance.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

Author contributions

JL: Conceptualization, Formal Analysis, Writing – original draft. YF: Software, Writing – review and editing. LZ: Funding acquisition, Supervision, Writing – review and editing. WH: Data curation, Investigation, Writing – review and editing.

Funding

The authors declare that financial support was received for the research and/or publication of this article. The research was supported by Science and Technology Project of Jiangxi Provincial Water Resources Department (202526YBKT25, 202526YBKT27, 202526YBKT39, and 202426ZDKT08).

Acknowledgements

We would like to thank the reviewers for their helpful comments and suggestions.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The authors declare that no Generative AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fenvs.2025.1676840/full#supplementary-material

References

Alaswad, S. O., and Almatrafi, E. (2024). Artificial intelligence applications in forward osmosis for water treatment: recent developments and research directions. Desalination Water Treat. 319, 100551. doi:10.1016/j.dwt.2024.100551

CrossRef Full Text | Google Scholar

Azam, S., and Liu, S. M. (2025). Role of moisture on gas sorption capacity and kinetics of coal: novel experimental and modeling insights. Chem. Eng. J. 510, 161540. doi:10.1016/j.cej.2025.161540

CrossRef Full Text | Google Scholar

Bing, L., Xi, H. J., Hu, Y. T., Zhe, L., Li, T. H., and Wen, Z. Q. (2022). Impact of local environmental standards on water environmental carrying capacity in large water diversion project: a system dynamics analysis in Shandong, China. J. Clean. Prod. 362, 132423. doi:10.1016/j.jclepro.2022.132423

CrossRef Full Text | Google Scholar

Bu, J. H., Li, C. H., Wang, X., Zhang, Y., and Yang, Z. W. (2020). Assessment and prediction of the water ecological carrying capacity in Changzhou city, China. J. Clean. Prod. 277, 123988. doi:10.1016/j.jclepro.2020.123988

CrossRef Full Text | Google Scholar

Bui, L. T., and Pham, H. T. H. (2023). Linking hydrological, hydraulic and water quality models for river water environmental capacity assessment. Sci. Total Environ. 857, 159490. doi:10.1016/j.scitotenv.2022.159490

PubMed Abstract | CrossRef Full Text | Google Scholar

Burgan, H. I. (2021). Numerical modeling of structural irregularities on unsymmetrical buildings. Teh. Vjesnik-Technical Gaz. 28 (3), 856–861. doi:10.17559/TV-20200328103359

CrossRef Full Text | Google Scholar

Chen, S. Y., He, Y. H., Tan, Q., Hu, K. J., Zhang, T. Y., and Zhang, S. (2022). Comprehensive assessment of water environmental carrying capacity for sustainable watershed development. J. Environ. Manag. 303, 114065. doi:10.1016/j.jenvman.2021.114065

PubMed Abstract | CrossRef Full Text | Google Scholar

Chen, L., Pu, Y., Zhu, K. H., Guo, C. X., Wang, Y. W., and Shen, Z. Y. (2024). Integrating best management practices with dynamic water environmental capacity for optimal watershed management. J. Environ. Manag. 371, 123264. doi:10.1016/j.jenvman.2024.123264

PubMed Abstract | CrossRef Full Text | Google Scholar

China (GB3838-2002) (2002). Environmental quality standards for surface water.

Google Scholar

Dai, D., Sun, M. D., Lv, X. B., Hu, J. X., Zhang, H., Xu, X. Q., et al. (2022). Comprehensive assessment of the water environment carrying capacity based on the spatial system dynamics model, a case study of Yongding River Basin in North China. J. Clean. Prod. 344, 131137. doi:10.1016/j.jclepro.2022.131137

CrossRef Full Text | Google Scholar

Feng, L. H., and Huang, C. F. (2008). A risk assessment model of water shortage based on information diffusion technology and its application in analyzing carrying capacity of water resources. Water Resour. Manag. 22 (5), 621–633. doi:10.1007/s11269-007-9182-z

CrossRef Full Text | Google Scholar

Fu, Y. (2024). Optimising curriculum quality assessment in application-oriented universities: a student-centric approach using entropy and cloud model. Eur. J. Educ. 59 (4), e12801–e12814. doi:10.1111/ejed.12801

CrossRef Full Text | Google Scholar

Hu, G. Z., Zeng, W. H., Yao, R. H., Xie, Y. X., and Liang, S. (2021a). An integrated assessment system for the carrying capacity of the water environment based on system dynamics. J. Environ. Manag. 295, 113045. doi:10.1016/j.jenvman.2021.113045

PubMed Abstract | CrossRef Full Text | Google Scholar

Hu, H., Jin, D. F., Yang, Y. Y., Zhang, J., Ma, C. P., and Qiu, Z. M. (2021b). Distinct profile of bacterial community and antibiotic resistance genes on microplastics in ganjiang River at the watershed level. Environ. Res. 200, 111363. doi:10.1016/j.envres.2021.111363

PubMed Abstract | CrossRef Full Text | Google Scholar

Hu, Y. T., Liu, J. Y., Wang, Y., Liu, G., Song, K. S., Wu, S. H., et al. (2024). The evaluation of small river water pollution caused by tailing spill in the Northeast of China using high-resolution images. Environ. Impact Assess. Rev. 109, 107633. doi:10.1016/j.eiar.2024.107633

CrossRef Full Text | Google Scholar

Huang, J., Pang, Y., Zhang, X. Q., and Tong, Y. F. (2019). Water environmental capacity calculation and allocation of the Taihu Lake Basin in Jiangsu Province based on control unit. Int. J. Environ. Res. Public Health 16 (19), 3774. doi:10.3390/ijerph16193774

PubMed Abstract | CrossRef Full Text | Google Scholar

Jin, H. Y., Chen, W. Q., Zhao, Z. H., Wang, J. J., and Ma, W. C. (2022). New framework for dynamic water environmental capacity estimation integrating the hydro-environmental model and load-duration curve method-a case study in data-scarce luanhe river basin. Int. J. Environ. Res. Public Health 19 (14), 8389. doi:10.3390/ijerph19148389

PubMed Abstract | CrossRef Full Text | Google Scholar

Li, X. X., Li, C. H., Wang, X., Liu, Q., Yi, Y. J., and Zhang, X. L. (2022). A developed method of water pollution control based on environmental capacity and environmental flow in Luanhe River Basin. Water 14 (5), 1–17. doi:10.3390/w14050730

CrossRef Full Text | Google Scholar

Liu, Q. K., Jiang, J. G., Jing, C. W., and Qi, J. G. (2018). Spatial and seasonal dynamics of water environmental capacity in mountainous Rivers of the Southeastern Coast, China. Int. J. Environ. Res. Public Health 15 (1), 99. doi:10.3390/ijerph15010099

PubMed Abstract | CrossRef Full Text | Google Scholar

Lu, J., Fu, Y., Li, X., Zhou, Y., and Zhang, L. (2024). Impact of the comprehensive remediation project on hydrological conditions in the lower reaches of the Ganjiang River. Water Sci. Technol. 89 (9), 2577–2592. doi:10.2166/wst.2024.137

PubMed Abstract | CrossRef Full Text | Google Scholar

Ma, X., Brookes, J., Wang, X. L., Han, Y. C., Ma, J. R., Li, G. Y., et al. (2023). Water quality improvement and existing challenges in the Pearl River Basin, China. J. Water Process Eng. 55, 104184. doi:10.1016/j.jwpe.2023.104184

CrossRef Full Text | Google Scholar

Mahato, J. K., Rawat, S., Gupta, S. K., and Yadav, B. (2024). Adsorptive remediation of coal bed methane produced water (CBMW) using a novel bio-adsorbent and modern enable artificial intelligence modeling. Chem. Eng. Res. and Des. 207, 181–191. doi:10.1016/j.cherd.2024.05.029

CrossRef Full Text | Google Scholar

Mou, S. Y., Yan, J. J., Sha, J. H., Li, S. L., Ma, Y. F., He, G. Y., et al. (2019). Dynamic simulation of the optimal allocation of water resources via the introduction of integrated water environmental policies in Baoding, China. Environ. Sci. Pollut. Res. 26 (26), 27516–27533. doi:10.1007/s11356-019-05537-9

PubMed Abstract | CrossRef Full Text | Google Scholar

Planning, C. A. F. E. (2003). Technical guideline for national water environmental capacity assessment.

Google Scholar

Qin, X. X., Sun, N., Teng, W. X., Zhu, Y. C., Liu, Z. W., Li, W. T., et al. (2023). Spatiotemporal distribution of microplastics in the Ganzhou section of the Ganjiang river: an insight into the source area impact. J. Environ. Chem. Eng. 11 (3), 627–640. doi:10.1016/j.jece.2023.109695

CrossRef Full Text | Google Scholar

Shi, Y., Yang, S. P., Zhang, L. W., Chen, W. W., Fan, Y. J., Lu, L., et al. (2024). Forecasting and advancing water carrying capacity in Henan Province in China: application of 'four determinations with water' in AHP and SD modeling. Sci. Total Environ. 919, 170757. doi:10.1016/j.scitotenv.2024.170757

PubMed Abstract | CrossRef Full Text | Google Scholar

Torres-Bejarano, F. M., Verbel-Escobar, M., and Atencia-Osorio, M. C. (2022). Water quality model-based methodology to assess assimilative capacity of wastewater discharges in rivers. Glob. J. Environ. Sci. Management-Gjesm 8 (4), 449–472. doi:10.22034/gjesm.2022.04.01

CrossRef Full Text | Google Scholar

Wang, W. J., Fraser, J. D., and Chen, J. K. (2019). Distribution and long-term population trends of wintering waterbirds in Poyang Lake, China. Wetlands 39, S125–S135. doi:10.1007/s13157-017-0981-6

CrossRef Full Text | Google Scholar

Wen, Y. H., and Wang, Y. (2023). Numerical simulation of water environment capacity from wet, normal, and dry years: taking the Luo River as an example. Aqua-Water Infrastructure Ecosyst. Soc. 72 (12), 2446–2471. doi:10.2166/aqua.2023.318

CrossRef Full Text | Google Scholar

Wu, L., Chen, Z., Ding, X., Liu, H. Y., and Wang, D. Q. (2022). Research on water environmental capacity accounting of the Yongzhou Section of Xiangjiang River Basin based on the SWAT-EFDC coupling model. J. Water Clim. Change 13 (2), 1106–1122. doi:10.2166/wcc.2021.319

CrossRef Full Text | Google Scholar

Yang, Z. Y., Song, J. X., Cheng, D. D., Xia, J., Li, Q., and Ahamad, M. I. (2019). Comprehensive evaluation and scenario simulation for the water resources carrying capacity in Xi'an city, China. J. Environ. Manag. 230, 221–233. doi:10.1016/j.jenvman.2018.09.085

PubMed Abstract | CrossRef Full Text | Google Scholar

Yang, C., Zheng, L. X., Zhang, Z. Y., and Feng, M. Q. (2022). Study on water quality simulation and dynamic water environment capacity of Helan County section of the third drainage ditch in Ningxia. Front. Ecol. Evol. 10, 1068876. doi:10.3389/fevo.2022.1068876

CrossRef Full Text | Google Scholar

Yu, Q. W., Sun, Z. H., Shen, J. Y., Xu, X., and Chen, X. N. (2023). Interactive allocation of water pollutant initial emission rights in a Basin under total amount control: a leader-follower hierarchical decision model. Int. J. Environ. Res. Public Health 20 (2), 1511. doi:10.3390/ijerph20021511

PubMed Abstract | CrossRef Full Text | Google Scholar

Zhang, Z., Balay, J. W., Bertoldi, K. M., and Macoy, P. O. (2016). Assessment of water capacity and availability from unregulated stream flows based on ecological limits of hydrologic alteration (ELOHA) environmental flow standards. River Res. Appl. 32 (7), 1469–1480. doi:10.1002/rra.2979

CrossRef Full Text | Google Scholar

Zhang, X. Y., Zuo, Q. T., and Yang, Q. X. (2018). Calculation of the water resources dynamic carrying capacity of Tarim River Basin under climate change. Desalination Water Treat. 119, 243–252. doi:10.5004/dwt.2018.22056

CrossRef Full Text | Google Scholar

Zhang, X., Luo, J. G., Zhang, X., and Xie, J. C. (2020). Dynamic simulation of river water environmental capacity based on the subsection summation model. Water Environ. Res. 92 (2), 278–290. doi:10.1002/wer.1242

PubMed Abstract | CrossRef Full Text | Google Scholar

Zhang, X. Q., Chen, P., Dai, S. N., Han, Y. H., and Chang, X. (2024). Urban HydroConnect: harnessing SWMM for dynamic waterway management assessment. Water Resour. Manag. 38 (12), 4473–4498. doi:10.1007/s11269-024-03875-z

CrossRef Full Text | Google Scholar