# Research on Water Quality Simulation and Water Environmental Capacity in Lushui River Based on WASP Model

^{1}

^{2}

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Materials and Methods

#### 2.1. Case Study

^{2}, with 2278 km

^{2}located in Jiangxi Province and 3397 km

^{2}in Hunan Province. There are two sources of the Lushui River: south and north, with the south as the primary source. Lushui originates from the south of Qianla Mountain in Pingxiang City, Jiangxi Province, and enters Hunan Province from Jinyushi. It joins the northern source at Shuangjiangkou, and Xiangjiang River at Lukou Town (Lukou District). In Hunan, the water flows through Liuyang City and Zhuzhou City. In this study, Lushui River Basin in Zhuzhou City (Hunan Province) is selected as the research area (shown in Figure 1). The mainstream of Lushui River is 82 km long in Zhuzhou City, including 68 km in Liling City and 14 km in Lukou District. The Lushui River system is developed with many tributaries, which flow through twenty-three towns in Liling City, Lukou District, and Youxian County in Zhuzhou, with a total drainage area of 2788 km

^{2}. It also includes a total of six monitoring sections on the mainstream of Lushui River from upstream to downstream.

#### 2.2. Determination of the COD, AN and TP

^{3}) and ammonium (NH

^{4+}) in natural waters. The common methods for quantitative analysis of ammonia nitrogen in natural water are the indophenol blue (IPB) spectrophotometric methods and ophthalaldehyde (OPA) fluorometric methods [9]. The IPB spectrophotometric methods are based on the Berthelot reaction [10].

#### 2.3. Model Description and Application

_{x}, u

_{y}and u

_{z}are the longitudinal, lateral, and vertical adjective velocities (m/day), respectively; E

**, E**

_{x}**, and E**

_{y}**are the longitudinal, lateral, and vertical diffusion coefficients (m**

_{z}^{2}/day), respectively; S

_{L}is the direct and diffuse loading rate (g/m

^{3}day); S

_{B}is the pollution load of the boundary rate (including upstream, downstream, benthic and atmospheric) (g/m

^{3}day); and S

**is the total kinetic transformation rate (g/m**

_{K}^{3}day).

#### 2.4. Model Input

#### 2.4.1. Basic Information

#### 2.4.2. Segmentation and Division of Control Units in the Study Area

- Collection of primary information data of natural geographical features in the Lushui area, including digital elevation (DEM), topographic maps, tributaries, small watershed boundaries, county and city administrative boundaries, environmental function areas, urban built-up areas, companies, drinking water sources, etc.;
- Processing of geographic information data based on ArcGIS, and the above primary geographic information data were overlaid and analyzed. Afterward, a preliminary sketch of the control unit was obtained;
- Division of the control units: after understanding and processing the collected information, the preliminary control unit sketch was fine-tuned according to the control unit’s principles of watershed division and clean boundary. Extraction and confirmation were performed upon the boundaries of different water catchment units and control units on different environmental functions in the same administrative region, to achieve a one-to-one correspondence between pollution source water quality and strong operability;
- Naming of the control units: in the absence of nationally unified naming standards and norms, naming was undertaken as a process of trying to understand and conform to the actual situation of each control unit. Therefore, the Lushui River was discretized into six control unit bodies. Each control unit body was given detailed geometric attributes. Some data were obtained from monitoring points during segmentation and generalization. In addition, other geometry parameters were obtained by the formula on WASP 8 User’s Guide. The measured river geometries and water velocities data were used to obtain the value of the exponent and coefficient of velocities and the depth of each segment, as shown in Equations (2), (3), (4), (5) and (6) [12]:B = B
_{mult}× Q^{bexp}$$\mathrm{V}{\mathrm{m}}_{}=\frac{{D}_{m}^{2/3}}{n}.{S}_{f}^{1/2}$$Q_{m}= V_{m}D_{m}B_{m}V_{mult}× d_{mult}× b_{mult}= 1v_{exp}+ d_{exp}+ b_{exp}= 1_{m}is velocity [m/sec]; B_{m}is width in m; b_{mult,}V_{mult}and d_{mult}are empirical coefficients; b_{exp}, v_{exp}and d_{exp}are empirical exponents; and Q_{m}is flow. The flow functions in WASP require the relationships of some hydraulic parameters such as depth and velocity; and the coefficients of the width are obtained internally from Equations 5 and 6. However, in the flow function, the condition of the hydraulic depth exponent, along with width B_{m}and depth D_{m}under average flow function, are required. We used Manning’s equation to calculate velocity v_{m}, and the average flow Q_{m}, from width and depth. The segmentation and the division of the control unit in Lushui River are shown in Figure 2 and Table 1.

#### 2.4.3. Pollution Load

_{t}is the total pollution loads; W

_{p}is the pollution load of the point source; W

_{np}is the pollution load of the non-point source; t is the time; C

_{p}(t) is the pollutant concentration of the point source at time t; and C

_{np}(t) is the pollutant concentration of the non-point source at time t. Due to the lack of continuous data obtainable from the river water quality monitoring point, the integral Equation (7) should be changed into the discrete Equation (8):

_{i}is the concentration from monitoring data in the ith month (mg/L); Q

_{i}is the average flow in the ith month (m

^{3}/s); and ∆t is the period of the ith month. W

_{i}is the average total pollution loads in the ith month (kg/d): α is the conversion factor and its calculation formula is shown in Equation (11):

#### 2.4.4. Discrete Exchange

^{2}/sec). During data input, these points are interpolated by the dispersion coefficients based on the dispersion function interpolation option selected. The river longitudinal dispersion coefficient represents the area (m

^{2}/s) of the pollutants in the river. The pollutants are longitudinally dispersed along the river direction per second. The integral formula and empirical estimation formula could determine it. However, the integral formula is suitable for river hydrological data and river section data, therefore, for this research, we used the empirical formula for estimation. To date, the well-known empirical formulas mainly include formulas established by Fischer [16], Mcquivey-Keefer [17], Iwasa [18], and Seo [17]. According to the completeness of the collected data, Mcquivey-Keefer appears as the most straightforward and most feasible to calculate, and its calculation is noted in the formula (Equation (12)):

_{x}is the longitudinal dispersion coefficient in m

^{2}/s, Q is the flow of the river in m

^{3}/s, B is the average water surface width in m, J is the hydraulic slope %, g is the acceleration of gravity m/s

^{2}, H is the average water depth in m, and u is the average flow velocity of the segment m/s. According to the Mcquivey-Keefer formula, the longitudinal dispersion coefficient corresponding to each river section is calculated, as shown in Table 4.

#### 2.4.5. Initial and Boundary Conditions

#### 2.5. Sensitivity Analysis and Calibration of Parameters

#### 2.5.1. Sensitivity Analysis of Parameters

- Local sensitivity analysis:

- Determine the analysis method:

_{S}) [23] is as follows:

_{S}means sensitivity index, X is the parameter value, Y is the simulation value, ΔY is the relative change of the pollution caused by the change of parameter, and ΔX is the relative change of parameter value. Table 6 shows the grading standard and level of the sensitivity.

- 2.
- Determine the initial value of the parameter:

- 3.
- Determine the range of change (−50%, +50%):

- Sensitivity analysis based on orthogonal design method:

_{j}= max {K

_{1j}, K

_{2j}, K

_{3j}…} − min {K

_{1j}, K

_{2j}, K

_{3j}…}

_{j}is the range value of the j-th parameter, and K

_{ij}is the average value of the simulation results of the parameter j at the i level.

- Determination of the parameter level: according to the value range of each parameter, five parameter levels were set uniformly, namely 80%, 90%, 100%, 110% and 120% of the initial value of each parameter;
- Determination of the orthogonal design scheme: this orthogonal design was completed using the same level orthogonal table. The orthogonal table was the source of the orthogonal design method, which forms as follows:

_{n}(m

^{k})

_{4}(2

^{3}), L

_{8}(2

^{7}), L

_{9}(3

^{4}), L

_{16}(4

^{5}), L

_{25}(5

^{6}), etc. Additionally, according to the local sensitivity result and the selection principle of the orthogonal table, six factors and five levels were used to complete the orthogonal table, and arsenic was five factors and five levels. Therefore, the selection of L

_{25}(5

^{6}) orthogonal table met this requirement. On this basis, the various levels of each parameter were numbered 1, 2, 3, 4, and 5 respectively.

- 3.
- Performance of range analysis: this orthogonal design used the L
_{25}(5^{6}) orthogonal table for design, so each indicator required a total of 25 tests; that is, COD, AN, and TP all need to run the model for 25 simulations. During the simulation, the following steps occurred: collection of the output concentration values of COD, AN, and TP, inputting of the corresponding positions into the orthogonal table, calculation of the range value of each parameter according to the range analysis method, and finally, obtaining the range analysis results of each parameter. The orthogonal design method is a qualitative analysis, which mainly determines the model’s critical parameters by comparing the relative magnitude of each parameter range, so the smaller range value does not affect the results of the analysis model. Therefore, the orthogonal design method was consistent with the local sensitivity analysis results. Results were determined with both analyses where the key parameters affecting the simulated output concentration were nitrification rate constant at 20 °C (K_{12}), nitrification temperature coefficient (Θ_{12}), dissolved organic phosphorus mineralization rate constant at 20 °C (K_{83}), dissolved organic phosphorus mineralization temperature coefficient (Θ_{83}), COD decay rate constant at 20 °C (K_{d}), and COD decay rate temperature correction coefficient (Θ_{d}). - 4.
- The analysis of theoretical and optimal parameters: in parameter sensitivity analysis, each initial parameter value is not the fixed value of the parameter. Consequently, the primary purpose of the combined analysis is to find the parameter combination that is closest to the monitored value as the initial combination of parameter rate timing, by improving the efficiency of parameter calibration. For example, the annual average values of the monitoring concentrations of COD, AN, and TP in Segment 5 in 2019 were 13.33 mg/L, 0.18 mg/L, and 0.0525 mg/L, respectively. The concentrations were compared with the output simulated value, and the closer monitoring value was the corresponding parameter combination. At that point it was considered to be the theoretical optimal parameter combination. After comparison, the theoretical optimal combination of COD, AN, and TP was determined, as shown in Table 7.

#### 2.5.2. Calibration of Parameters

#### 2.5.3. Model Verification and Model Accuracy Evaluation

- Model verification:

_{M}is the measured value of each water quality index; and C

_{S}is the simulated value of each water quality index. We used the basis of general performance ratings to evaluate the model [29], as shown in Table 8.

- Model Accuracy Evaluation:

^{2}) determines the goodness of fit between simulated and observed data. The range value of R

^{2}was from 0 to 1. If the value of R

^{2}was close to 1, the model simulation fit well with the monitored data. Table 8 shows the general performance ratings of the coefficient of determination (R

^{2}).

#### 2.6. Water Environment Capacity

- Water quality aims were to determine the base level of the water environmental management conditions of the Lushui River. Due to the water environmental assessment conditions of Zhuzhou City, the goal of present environmental work in Zhuzhou to upgrade the water quality of the Lushui River to Grade II standard. Water quality standards [30,31] are shown in Table 9.
- To obtain various pollutant environmental capacities and simulate the pollution loads of Lushui River, we determined the water quality requirements for water quality control in the Lushui River basin as the aims of the simulation, and adjusted the input pollution loads of each indicator, such as COD, TP and AN.
- The water quality of the Lushui River was set to Grade II to obtain the water environmental capacity of the Lushui River, and we also considered the pollutants discharged into the Lushui River from both shores.

## 3. Results

#### 3.1. Parameter Sensitivity Analysis

- For AN: only Θ
_{12}was high sensitivity, K_{12}was medium sensitivity, and the sensitivity of the remaining parameters were small to negligible; - For COD: Θ
_{d}was high sensitivity, K_{d}was medium sensitivity, and the sensitivity of the remaining parameters were small to negligible; - For TP: Θ
_{12}was very high sensitivity, Θ_{83}was high sensitivity, K_{83}was medium sensitivity, and the sensitivity of the other parameters was small to negligible.

_{12}, which is a sensitive parameter for ammonia nitrogen and total phosphorus. This caused the four reaction systems of the EUTRO module [8] (that is, the N cycle, P cycle, and DO balance, and algae growth system) to interact with each other.

#### 3.2. Model Calibration

_{12}, Θ

_{12}, K

_{83}, Θ

_{83}, K

_{d}and Θ

_{d}were calibrated, and the remaining parameters were used as the default values. The Table 11 shows the model parameter calibration results.

#### 3.3. Simulation Results

#### 3.4. Water Environment Capacity

## 4. Discussion

#### 4.1. Generalization and Polltion Load

#### 4.2. Parameter Sensitivity Analysis

_{12}, Θ

_{12}, K

_{83}, Θ

_{83}, K

_{d}and Θ

_{d})were identified as medium to highly influential. The sensitivity index of all these parameters were above 0.05, indicating that they are the key parameters affecting the simulated output concentration, according to the model parameter sensitivity analysis (Table 6 and Table 10). The high sensitivity of these parameters was also reported by Guo et al. [25].

#### 4.3. Simulation Results and Model Verification

#### 4.4. Model Verification and Model Accuracy Evaluation

^{2}) sets the fit between monitored and simulated data, and its range value is from 0 to 1. If R

^{2}was close to 1, the model simulation fit satisfactorily with measured data. It can be seen from Figure 4 and Table 15 that the determination coefficients R

^{2}were all above 0.85 and the Index of Agreement (r) were all above 0.90, indicating that the simulated values were in good agreement with the monitored values and had a reasonable correlation. Similar findings relating to the overall effectiveness of water quality simulation studies in other domestic basins were reported by Guo et al. [25].

#### 4.5. Water Environmental Capacity

## 5. Conclusions

_{12}had high sensitivity, K

_{12}had medium sensitivity, and the sensitivity of the remaining parameters were small to negligible. Additionally, for the parameter sensitivity of COD, Θ

_{d}had high sensitivity, K

_{d}had medium sensitivity, and the sensitivity of the remaining parameters were small to negligible. Furthermore, for total phosphorus, Θ

_{12}had very high sensitivity, Θ

_{83}had high sensitivity, K

_{83}had medium sensitivity, and the sensitivity of the other parameters were small to negligible.

^{2}being above 0.85, and all of the correlation coefficients r being above 0.90. Therefore, this information indicates that the WASP model is applicable and reliable.

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Acknowledgments

## Conflicts of Interest

## References

- Fonseca, B.M.; De Mendonça-Galvão, L.; Padovesi-Fonseca, C. Nutrient baselines of Cerrado low-order streams: Comparing natural and impacted sites in Central Brazil. Environ. Monit. Assess.
**2014**, 186, 19. [Google Scholar] [CrossRef] - Brooks, B.W.; Riley, T.M.; Taylor, R.D. Water quality of effluent-dominated ecosystems: Ecotoxicological, hydrological, and management considerations. Hydrobiologia
**2006**, 556, 365–379. [Google Scholar] [CrossRef] - Zhang, M.L.; Shen, Y.M.; Guo, Y. Development and application of an eutrophication water quality model for river networks. Hydrodynamics
**2008**, 20, 719–726. [Google Scholar] [CrossRef] - Ellina, G.; Papaschinopoulos, G.; Papadopoulos, B.K. Research of fuzzy implications via fuzzy linear regression in data analysis for a fuzzy model. J. Comput. Methods Sci. Eng.
**2020**, 20, 879–888. [Google Scholar] [CrossRef] - Hu, X. Research on Water Quality Management of Xiangtan Section of Yangtze River Based on WASP Model; Xiangtan University: Xiangtan, China, 2013. (In Chinese) [Google Scholar]
- Lin, J.; Huang, J.-L.; Du, P.-F.; Tu, Z.-S.; Li, Q.-S. Local sensitivity and stability analysis of parameters in urban rainfall runoff hydrological simulation. Environ. Sci.
**2010**, 31, 2023–2028. [Google Scholar] - Moriasi, D.N.; Gitau, M.W.; Pai, N.; Daggupati, P. Hydrologic and Water Quality Models: Performance Measures and Evaluation Criteria. Trans. ASABE
**2015**, 58, 1763–1785. [Google Scholar] - American Public Health Association; American Water Works Association. Standard Methods for the Examination of Water and Wastewater, 14th ed.; Method 508; American Public Health Association (APHA): Washington, DC, USA, 1975; p. 550. [Google Scholar]
- Ma, L.; Adornato, R.H.; Byrne, D. Yuan Determination of nanomolar levels of nutrients in seawater. Trends Anal. Chem.
**2014**, 60, 1–15. [Google Scholar] [CrossRef] - Elser, J.J. Phosphorus: A limiting nutrient for humanity? Curr. Opin. Biotechnol.
**2012**, 23, 833–838. [Google Scholar] [CrossRef] - H Šraj, L.O.; Almeida, M.I.G.S.; Swearer, S.E.; Kolev, S.D.; McKelvie, I.D. Analytical challenges and advantages of using flow-based methodologies for ammonia determination in estuarine and marine waters. Trends Anal. Chem.
**2014**, 59, 83–92. [Google Scholar] [CrossRef] - Standard. Surface Water Environmental Quality Standard (GB3838-2002); Ministry of Environmental Protection of the People’s Republic of China: Beijing, China, 2002. (In Chinese)
- Seo, I.W.; Baek, K.O. Estimation of the Longitudinal Dispersion Coefficient Using the Velocity Profile in Natural Streams. J. Hydraul. Eng.
**2004**, 130, 227–236. [Google Scholar] [CrossRef] - Wijaya, D.S.; Juwan, I. Identification and Calculation of Pollutant Load in Ciwaringin Watershed, Indonesia: Domestic Sector and Faculty of Civil Engineering and Planning; Insttitut Teknologi Nasional Bandung: West Java, Indonesia, 2017. [Google Scholar]
- Xiao, K.X.; Wei, Y.; Ke-feng, L. Estimation and discussion of non-point source pollution loads with new Flux method in Danjiangkou Reservoir area, China. 2016. Available online: https://www.researchgate.net/publication/317291492_Estimation_and_discussion_of_non-point_source_pollution_loads_with_new_Flux_method_in_Danjiangkou_Reservoir_area_China (accessed on 5 October 2021).
- Gu, L.; Hua, Z. Research progress in determining the longitudinal dispersion coefficient of natural rivers. Adv. Sci. Technol. Water Resour. Hydropower
**2007**, 27, 85–89. [Google Scholar] - Fischer, H.B. Discussion of Simple Method for Predicting Dispersion in Streams. Keefer. J. Environ. Eng. Div.
**1975**, 101, 453–455. [Google Scholar] [CrossRef] - Mcquivey, R.S.; Keefer, T.N. Simple method for predicting dispersion in streams. J. Environ. Eng. Div.
**1974**, 100, 997–1011. [Google Scholar] [CrossRef] - Tufford, D.L.; McKellar, H.N. Spatial and temporal hydrodynamic and water quality Modelling analysis of a large reservoir on the South Carolina (USA) coastal plain. Ecol. Model.
**1999**, 114, 137–173. [Google Scholar] [CrossRef] - Bae, S.; Seo, D. Analysis and modelling of algal blooms in the Nakdong River. Korea Ecol. Model
**2018**, 372, 53–63. [Google Scholar] [CrossRef] - Lenhart, T.; Eckhardt, K.; Fohrer, N. Comparison of two different approaches of sensitivity analysis. Phys. Chem. Earth
**2002**, 27, 645–654. [Google Scholar] [CrossRef] - Francos, A.; Elorza, F.J.; Bouraoui, F. Sensitivity analysis of distributed environmental simulation models: Understanding the model behaviour in hydrological studies at the catchment scale. Reliab. Eng. Syst. Saf.
**2003**, 79, 205–218. [Google Scholar] [CrossRef] - Peng, S. Research on Uncertainty Water Quality Model Based on WASP Model; Tianjin University: Tianjin, China, 2010. (In Chinese) [Google Scholar]
- Allen, J.I.; Holt, J.T.; Blackford, J.; Proctor, R. Error quantification of a high-resolution coupled hydrodynamic-ecosystem coastal-ocean model: Part 2. Chlorophyll-a, nutrients and SPM. Mar. Syst.
**2007**, 68, 381–404. [Google Scholar] - Guo, X.; Yan, S. Sensitivity analysis of rigid pile composite foundation model parameters based on orthogonal design method. J. Self -Sci. Xiangtan Univ.
**2015**, 37, 15–20. [Google Scholar] - Cerny, V. Thermodynamical approach to the traveling salesman problem: An efficient simulation algorithm. Optimiz. Theory App.
**1985**, 45, 41–51. [Google Scholar] [CrossRef] - Putnam, H. Trial and error predicates and the solution to a problem of Mostowski. Symb. Logic.
**1965**, 30, 49–57. [Google Scholar] [CrossRef] - Refsgaard, J.C.; Knudsen, J. Operational validation and intercomparison of different types of hydrological models. Water Resour. Res.
**1996**, 32, 2189–2202. [Google Scholar] [CrossRef] - Liu, X. The Heavy Metal Water Quality Target Management Research on Yangtze River Based on Human Health Risk; Central South University: Changsha, China, 2017. (In Chinese) [Google Scholar]
- Sin, G.; de Pauw, D.J.W.; Weijers, S.; Vanrolleghem, P.A. An efficient approach to automate the manual trial and error calibration of activated sludge models. Biotechnol. Bioeng.
**2008**, 100, 516–528. [Google Scholar] [CrossRef] - Nash, J.E.; Sutcliffe, J.V. River flow forecasting through conceptual models part I—A discussion of principles. Hydrology
**1970**, 10, 282–290. [Google Scholar] [CrossRef]

**Figure 3.**(

**a**), (

**b**) and (

**c**) Comparison of the simulation results and measured data of COD, AN and TP in Segment 5 (2020).

Segment | Volume | VelocityMult | DepthMult | Length | Width | Slope | BottomRough |

1 | 9.43 × 10^{5} | 1.05 | 1 | 2.21 × 10^{4} | 42.74 | 9.97 × 10^{4} | 0.03 |

2 | 9.94 × 10^{5} | 1.26 | 1.5 | 1.20 × 10^{4} | 55.42 | 8.36 × 10^{4} | 0.03 |

3 | 4.72 × 10^{6} | 0.89 | 2 | 3.18 × 1^{4} | 74.17 | 2.82 × 10^{4} | 0.03 |

4 | 3.18 × 10^{6} | 1.55 | 3 | 1.20 × 10^{4} | 88.47 | 5.01 × 10^{4} | 0.03 |

5 | 1.51 × 10^{6} | 0.98 | 1 | 5.46 × 10^{4} | 27.61 | 8.60 × 10^{4} | 0.03 |

6 | 1.82 × 10^{6} | 1.04 | 1.3 | 3.79 × 10^{4} | 36.99 | 6.85 × 10^{4} | 0.03 |

^{3}; the unit of length, width and depth are m, and the unit of velocity is m/s.

Segment | COD | AN | TP |
---|---|---|---|

Segment 1 | 26769.6 | 1031.616 | 277.416 |

Segment 2 | 29318.4 | 1226.751 | 305.424 |

Segment 3 | 22910.4 | 359.452 | 187.704 |

Segment 4 | 210024 | 6440.76 | 2496.268 |

Segment 5 | 28687.68 | 681.883 | 140.112 |

Segment 6 | 26618.4 | 506.800 | 124.56 |

Segments | COD | AN | TP |
---|---|---|---|

Segment 1 | 46656 | 747.468 | 226.8 |

Segment 2 | 42026.4 | 706.176 | 153.216 |

Segment 3 | 25934.4 | 301.176 | 261.432 |

Segment 4 | 286387.2 | 4972.032 | 2865.384 |

Segment 5 | 28296 | 427.104 | 110.664 |

Segment 6 | 24429.6 | 361.44 | 78.192 |

Segments | E_{X} | Segments | E_{X} |
---|---|---|---|

Segment 1 | 43.047 | Segment 4 | 405.725 |

Segment 2 | 48.939 | Segment 5 | 62.246 |

Segment 3 | 79.564 | Segment 6 | 53.134 |

Boundary | COD (mg/L) | AN (mg/L) | TP (mg/L) |
---|---|---|---|

Segment 1 | 13.5 | 0.284 | 0.09 |

Segment 4 | 10 | 0.12 | 0.1 |

Segment 5 | 13 | 0.31 | 0.06 |

**Table 6.**Sensitivity classes [7].

Sensitivity Index | Grading Standard | Level |
---|---|---|

0.00 ≤ S_{S} ≤ 0.05 | Small to negligible | 1 |

0.05 ≤ S_{S} ≤ 0.20 | Medium | 2 |

0.20 ≤ Ss ≤ 1.00 | High | 3 |

Ss ≥ 1.00 | Very high | 4 |

K_{12} | Θ_{12} | K_{83} | Θ_{83} | K_{d} | Θ_{d} |
---|---|---|---|---|---|

0.0972 | 0.832 | 0.144 | 0.856 | 0.2772 | 0.8376 |

Measure | Coefficient of Determination | Performance Rating |
---|---|---|

0.00 ≤ R^{2} ≤0.40 | Unsatisfactory | |

R^{2} | 0.40 ≤ R^{2} ≤ 0.60 | Satisfactory |

0.60 ≤ R^{2} ≤ 0.75 | Good | |

0.75 ≤ R^{2} ≤ 1 | Very Good |

Indicators | I | II | III | IV |

COD | 15 | 15 | 20 | 30 |

AN | 0.15 | 0.5 | 1.0 | 1.5 |

TP | 0.02 | 0.1 | 0.2 | 0.3 |

Symbol | COD | AN | TP | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|

(−50%) | (+50%) | S_{S} | Class | (−50%) | (+50%) | S_{S} | Class | (−50%) | (+50%) | S_{S} | Class | |

K_{12} * | 0 | 0.000 | 0.000 | 1 | 0.025 * | 0.075 * | 0.05 * | 2 * | 0 | 0 | 0 | 1 |

Θ_{12} * | 0 | 0.000 | 0.000 | 1 | 0.369 * | 0.387 * | 0.378 * | 3 * | 0 | 2.062 * | 1.031 * | 4 * |

K_{NO3} | 0 | 0.000 | 0.000 | 1 | 0.001 | 0.003 | 0.002 | 1 | 0 | 0 | 0 | 1 |

T_{min} | 0 | 0.000 | 0.000 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 1 |

K_{2D} | 0 | 0.000 | 0.000 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 1 |

Θ_{2D} | 0 | 0.000 | 0.000 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 1 |

K_{NO3} | 0.000 | 0.000 | 0.000 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 1 |

K_{71} | 0.000 | 0.000 | 0.000 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 1 |

K_{83} * | 0.000 | 0.000 | 0.000 | 1 | 0 | 0 | 0 | 1 | 0 | 0.111 * | 0.055 * | 2 * |

Θ_{71} | 0.000 | 0.000 | 0.000 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 1 |

Θ_{83} * | 0.000 | 0.000 | 0.000 | 1 | 0 | 0 | 0 | 1 | 0.44 * | 0.413 * | 0.427 * | 3 * |

K_{d} * | 0.065 * | 0.196 * | 0.131 * | 2 * | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 1 |

Θ_{d} * | 0.319 * | 0.612 * | 0.465 * | 3 * | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 1 |

K_{COD} | 0.003 | 0.009 | 0.006 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 1 |

f_{D5} | 0.000 | 0.000 | 0.000 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 1 |

R | 0.000 | 0.000 | 0.000 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 1 |

Parameter Name | Symbol | Value | Unit |
---|---|---|---|

Nitrification Rate Constant at 20 °C | K_{12} | 0.081 | d^{−1} |

Nitrification Temperature Coefficient | Θ_{12} | 1.04 | |

Half Saturation Constant for Nitrification Oxygen Limit | K_{NO3} | 0.5 | mgO_{2}/L |

Minimum Temperature for Nitrification Reaction | T_{min} | 3 | °C |

Denitrification Rate Constant at 20 °C | K_{2D} | 0.045 | d^{−1} |

Denitrification Temperature Coefficient | Θ_{2D} | 1.04 | |

Half Saturation Constant for Denitrification Oxygen Limit | K_{NO3} | 0.1 | mgO_{2}/L |

Dissolved Organic Nitrogen Mineralization Rate Constant at 20 °C | K_{71} | 0.075 | d^{−1} |

Dissolved Organic Phosphorus Mineralization Rate Constant at 20 °C | K_{83} | 0.12 | d^{−1} |

Dissolved Organic Nitrogen Mineralization Temperature Coefficient | Θ_{71} | 1.08 | |

Dissolved Organic Phosphorus Mineralization Temperature Coefficient | Θ_{83} | 1.07 | |

COD Decay Rate Constant at 20 °C | K_{d} | 0.231 | d^{−1} |

COD Decay Rate Temperature Correction Coefficient | Θ_{d} | 1.047 | |

COD Half Saturation Oxygen Limit | K_{COD} | 0.5 | mgO_{2}/L |

Fraction of Detritus Dissolution to COD | f _{D5} | 0.2 | |

Fraction of COD Carbon Source for Denitrification | R | 0.2 |

Segments | COD | AN | TP | ||||||
---|---|---|---|---|---|---|---|---|---|

Normal | Wet | Dry | Normal | Wet | Dry | Normal | Wet | Dry | |

Segment 1 | 8.837 | 5.223 | 8.654 | 0.392 | 0.314 | 0.418 | 0.130 | 0.097 | 0.105 |

Segment 2 | 19.140 | 9.468 | 13.320 | 0.731 | 0.469 | 0.795 | 0.191 | 0.160 | 0.195 |

Segment 3 | 12.376 | 7.732 | 9.623 | 0.325 | 0.243 | 0.325 | 0.095 | 0.092 | 0.110 |

Segment 4 | 38.393 | 44.830 | 38.922 | 1.355 | 1.258 | 0.844 | 0.425 | 0.645 | 0.290 |

Segment 5 | 15.609 | 10.697 | 11.671 | 0.269 | 0.374 | 0.218 | 0.061 | 0.067 | 0.067 |

Segment 6 | 16.141 | 10.712 | 10.229 | 0.263 | 0.373 | 0.157 | 0.071 | 0.068 | 0.063 |

Segments | COD | AN | TP | ||||||
---|---|---|---|---|---|---|---|---|---|

Normal | Wet | Dry | Normal | Wet | Dry | Normal | Wet | Dry | |

Segment 1 | 14,072.94 | 17,147.7 | 10,998.18 | 469.098 | 571.59 | 366.606 | 93.8196 | 114.318 | 73.3212 |

Segment 2 | 21,286.8 | 19,512.9 | 17,975.52 | 709.56 | 650.43 | 599.184 | 141.912 | 130.086 | 119.8368 |

Segment 3 | 13,481.64 | 14,427.72 | 11,944.26 | 449.388 | 480.924 | 398.142 | 89.8776 | 96.1848 | 79.6284 |

Segment 4 | 135,525.96 | 185,195.16 | 85,738.5 | 4517.532 | 6173.17 | 2857.95 | 903.5064 | 1234.634 | 571.59 |

Segment 5 | 15,846.84 | 12,417.3 | 10,288.62 | 528.228 | 413.91 | 342.954 | 105.6456 | 82.782 | 68.5908 |

Segment 6 | 12,949.47 | 12,831.21 | 7686.9 | 431.649 | 427.707 | 256.23 | 86.3298 | 85.5414 | 51.246 |

Total | 14,072.94 | 17,147.7 | 10,998.18 | 469.098 | 571.59 | 366.606 | 93.8196 | 114.318 | 73.3212 |

Date (MM/DD/YY) | COD | AN | TP | ||||||
---|---|---|---|---|---|---|---|---|---|

Simulated | Observed | f | Simulated | Observed | f | Simulated | Observed | f | |

1/15/2020 | 13.000 | 12 | 8.33% | 0.115 | 0.107 | 7.04% | 0.055 | 0.05 | 9.79% |

2/15/2020 | 14.083 | 13 | 8.33% | 0.118 | 0.115 | 2.98% | 0.065 | 0.06 | 8.25% |

3/15/2020 | 18.842 | 16 | 17.76% | 0.105 | 0.091 | 15.09% | 0.012 | 0.01 | 24.86% |

4/15/2020 | 15.831 | 14 | 13.08% | 0.387 | 0.372 | 4.02% | 0.104 | 0.09 | 15.23% |

5/15/2020 | 13.682 | 13 | 5.25% | 0.466 | 0.486 | 4.12% | 0.065 | 0.06 | 7.54% |

6/15/2020 | 18.629 | 18 | 3.50% | 0.252 | 0.267 | 5.46% | 0.022 | 0.02 | 11.65% |

7/15/2020 | 6.313 | 6 | 5.21% | 0.749 | 0.812 | 7.82% | 0.151 | 0.14 | 7.73% |

8/15/2020 | 8.019 | 8 | 0.24% | 0.296 | 0.334 | 11.40% | 0.042 | 0.04 | 5.63% |

9/15/2020 | 9.828 | 10 | 1.72% | 0.199 | 0.227 | 12.38% | 0.051 | 0.05 | 2.11% |

10/15/2020 | 8.979 | 9 | 0.24% | 0.235 | 0.257 | 8.50% | 0.061 | 0.06 | 1.67% |

11/15/2020 | 12.007 | 11.8 | 1.76% | 0.155 | 0.165 | 6.31% | 0.072 | 0.07 | 2.88% |

12/15/2020 | 12.700 | 11.8 | 7.62% | 0.171 | 0.165 | 3.90% | 0.074 | 0.07 | 6.31% |

**Table 15.**Accuracy evaluation statistics of the simulation results in Segment 5: January 1, 2020– December 31, 2021.

COD | AN | TP | ||||
---|---|---|---|---|---|---|

Statistics | Measured | Simulated | Measured | Simulated | Measured | Simulated |

Count | 12 | 12 | 12 | 12 | 12 | 12 |

Min | 6.000 | 6.313 | 0.091 | 0.104 | 0.010 | 0.012 |

Max | 18.000 | 18.842 | 0.812 | 0.749 | 0.140 | 0.151 |

Coef of Det (R^{2}) | 0.97 | 0.99 | 0.99 | |||

Mean Abs Error | 0.764 | 0.021 | 0.004 | |||

RMS Error | 1.129 | 0.027 | 0.006 | |||

Norm RMS Error | 0.088 | 0.078 | 0.079 | |||

Index of Agreement (r) | 0.98 | 0.99 | 0.99 |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Obin, N.; Tao, H.; Ge, F.; Liu, X. Research on Water Quality Simulation and Water Environmental Capacity in Lushui River Based on WASP Model. *Water* **2021**, *13*, 2819.
https://doi.org/10.3390/w13202819

**AMA Style**

Obin N, Tao H, Ge F, Liu X. Research on Water Quality Simulation and Water Environmental Capacity in Lushui River Based on WASP Model. *Water*. 2021; 13(20):2819.
https://doi.org/10.3390/w13202819

**Chicago/Turabian Style**

Obin, Nicolas, Hongni Tao, Fei Ge, and Xingwang Liu. 2021. "Research on Water Quality Simulation and Water Environmental Capacity in Lushui River Based on WASP Model" *Water* 13, no. 20: 2819.
https://doi.org/10.3390/w13202819