*2.3. Methods*

### 2.3.1. Relative Difference

This study introduces the concept of "relative gap" to compare and analyze the difference between the upstream and downstream pollutant concentrations [29]. It is defined as:

$$DR = \frac{\mathbb{C}u - \mathbb{C}d}{\mathbb{C}u} \times 100\% \tag{1}$$

where *DR* is the relative difference between the pollutant concentrations upstream and those downstream, *Cu* is the pollutant concentration upstream, and *Cd* is the pollutant concentration downstream.

If the calculated result is negative, it means that the pollutant concentration downstream of the river is higher than that upstream, while if it is positive, it means that the pollutant concentration downstream is lower than that upstream.

#### 2.3.2. Gaussian Fitting

In order to more deeply quantify the difference between pollutant concentrations upstream and downstream when the sluices are closed, the frequency distribution of the relative difference of pollutant concentrations was fitted using the Gauss fitting method in the normal distribution model in OriginLab software. The formula is as follows:

$$y = y\_0 + \frac{A}{w\sqrt{\pi/2}}e^{\frac{-2(x-x\_0)^2}{w^2}}\tag{2}$$

where *y*<sup>0</sup> is the baseline, *x*<sup>0</sup> is the mean, *w* is the discrete degree parameter, and *A* is the shape parameter.

#### **3. Results and Discussion**

#### *3.1. Variation of Water Quality Indicators in Different Seasons*

Urban rivers often take into account the role of discharging domestic sewage, most of which are nutrient pollutants. Nutrient pollutants mainly include nutrients represented by nitrogen and phosphorus, which are not considered pollutants in themselves. However, when the level of nutrients contained in sewage is relatively high, it will contribute to the proliferation of algae in the water and eutrophication of the water body, which leads to a series of hazards [30]. Furthermore, some studies have indicated that the main pollutants in inlet and outlet rivers around the whole of Taihu Lake are dominated by nitrogen pollutants, followed by organic pollution such as phosphorus [31]. Combined with the cyanobacterial water pollution events that have occurred in Taihu Lake [32], TN and TP are selected as water quality indicators in this study. The comparison results of TN and TP are shown in Figures 2 and 3.

The concentrations of TP and TN were relatively low during the wet and dry seasons compared to those in the flat-water season due to the fact that the wet and dry seasons are the summer and autumn seasons in Wuxi, with more rainfall. On the one hand, the high precipitation leads to the high storage capacity of rivers and the constant turnover of the water body. These are conducive to pollutant concentration reduction. On the other hand, for flood control purposes, the sluices are opened to release flood water when there is excessive rainfall. Furthermore, the river flows faster, and some pollutants from the river will flow into larger water bodies such as the Beijing–Hangzhou Grand Canal, which contributes to reducing pollutant concentrations.

In addition, there are some cases of excessive pollutant concentrations during the wet and dry seasons; for example, the TN concentrations upstream of the rivers in the wet seasons (Figure 2c,e,h) and the TP concentrations upstream of the rivers during the wet and dry seasons (Figure 3f,h), respectively. This is due to the fact that summer and autumn are not only the peak season of precipitation in Wuxi but also the peak period for industrial production and domestic sewage discharge. The combination of effluent discharge, sluice closure, and high temperatures during periods brings about high concentrations of these pollutants, especially during the time when it does not rain. This result is similar to previous studies [33].

#### *3.2. Variation of Water Quality Indicators in the Upstream and Downstream*

To further analyze the variation in the pollutant concentrations upstream and downstream of the rivers, the relative differences in the pollutant concentrations were calculated by Equation (1). The relative differences upstream and downstream of the rivers for TN and TP pollutants are shown in Tables 1 and 2.

**Figure 2.** Comparison of TN upstream and downstream of the rivers in the closed state of the sluices. (**a**) TN concentrations at point 1; (**b**) TN concentrations at point 2; (**c**) TN concentrations at point 3; (**d**) TN concentrations at point 4; (**e**) TN concentrations at point 5; (**f**) TN concentrations at point 6; (**g**) TN concentrations at point 7; (**h**) TN concentrations at point 8.

**Figure 3.** Comparison of TP upstream and downstream of the rivers in the closed state of the sluices. (**a**) TP concentrations at point 1; (**b**) TP concentrations at point 2; (**c**) TP concentrations at point 3; (**d**) TP concentrations at point 4; (**e**) TP concentrations at point 5; (**f**) TP concentrations at point 6; (**g**) TP concentrations at point 7; (**h**) TP concentrations at point 8.


**Table 1.** Relative difference in TN concentration.

**Table 2.** Relative difference in TP concentration.


The 24 groups of TN concentrations are compared in Table 1. Two groups of TN concentrations in the upper sites were smaller than those in the lower sites, and the relative difference between the upstream and downstream was not obvious. While another 22 groups of TN concentrations upstream were higher than those downstream, among which 13 groups, relative differences ranged from 0 to 30% between upstream and downstream, the relative differences of the five groups ranged from 30 to 100%, and four groups relative differences exceeded 100% between the upstream and downstream.

The results indicated that the water quality in the upper reaches of the river was worse than that in the lower reaches when the sluice status was closed, with the maximum relative difference between the upper and lower reaches of the TN concentration being greater than 100%. Meanwhile, the TN concentrations were relatively high in all rivers. There were a few rivers where the differences in TN concentrations between the upper and lower reaches were not significant, but the comparison of the differences in TN concentrations between the upper and lower reaches in most rivers was very obvious.

The TP concentration in the rivers of Wuxi was relatively lower compared to TN (Figure 2 and Table 2). Among the 24 sets of data, there were eight sets of relative differences within 10%, which is not a significant comparison. Furthermore, among the remaining 16 sets of data, the TN concentrations in three sets were in the upper reaches, less than those in the lower reaches, and the other 13 sets in the upper reaches were greater than that in the lower reaches. Meanwhile, of these 16 sets, there were seven sets with relative differences between upstream and downstream from 10% to 30%, six sets were between 30% and 100%, and three sets where the difference exceeded 100%. The concentration of TP was not high in general, and some of the data were not obvious enough for a clear comparison. However, for the 16 groups of data, it can still be concluded that the water quality in the upper reaches of the rivers is worse than that in the lower reaches.

Generally, the concentrations of TP were relatively low, and the degree of variation in TP was not as great as that of TN. Furthermore, the levels of TP and TN also differed significantly at different times of the year at the same site in the same river, reflecting seasonal variability. In addition to this, an important preliminary conclusion was drawn: the pollutant concentrations in the upper reaches of the Wuxi rivers were higher than those

in the lower reaches when the sluices were closed. The result is highly consistent with the previous investigations [34,35].

In order to further quantitatively analyze the pollutant concentrations in the upper and lower sites of rivers, the interval length of 20% was firstly selected for the frequency distribution map for frequency distribution statistics, and then the frequency distribution was fitted by applying Equation (2). The specific frequency distribution plots, as well as the fitted curves, are shown in Figure 4.

**Figure 4.** Relative difference frequency distribution and Gauss fitting curve (**a**) TN (**b**) TP.

The value of the Gauss fitting curve parameter xc for the relative disparity frequency of TN was 14.42 (Figure 4). That is, under the normal distribution model, the average value of the relative differences between the TN concentrations in the upper and lower reaches when the sluices were closed was 14.42%. Similarly, the mean value of the relative differences in TP concentration was 13.80%. This indicates that the TN and TP concentrations in the upper reaches of urban rivers in Wuxi were 14.42% and 13.80% higher than those in the lower reaches under the closed state of the sluices, respectively. This is consistent with our preliminary conclusions.

#### *3.3. Variation of Water Quality Indicators in Urban Rivers*

When the sluices are closed, the water quality upstream of urban rivers may be worse than that downstream. The reason for this phenomenon is inextricably linked to the water quality conditions of the urban rivers themselves.

First of all, the pollution sources of urban rivers have special characteristics compared to the general large rivers such as the Yangtze River and the Yellow River, has a great distinction. Some studies have concluded that the main pollutants in the Yangtze River come from urban domestic sewage and agricultural pollution [36]. Furthermore, the crosssections below the Three Gorges Dam are mainly located in the main urban living area, which causes some pollutant indicators in the lower reaches of the sluices and dam to be significantly higher than those in the upper reaches [37]. For urban rivers, in Wuxi, the city is located in the middle and upper reaches of rivers, where domestic sewage and industrial sources are the most important sources of nutrients and pollution [38]. These effluents flow into the rivers from the middle and upper sites, and these pollutants accumulate in the upper reaches of the sluices when they are closed. The presence of numerous sluices leads to the accumulation of large amounts of industrial wastewater, domestic sewage, and solid waste sediment in the upper reaches of the sluices. Although domestic, agricultural, and industrial wastewater is also discharged into the middle and lower reaches of urban rivers, the lower reaches usually connect to larger water bodies, such as Taihu Lake and the Beijing–Hangzhou Grand Canal. Consequently, the water quality upstream of urban rivers is often worse than that downstream.

Secondly, the lower reaches of urban rivers connect to larger water bodies, which own large volumes and lightly polluted water, as well as relatively clear water quality compared to that of urban rivers, and the concentrations of TN and TP are also relatively low. The concentrations of TN and TP in Taihu Lake and the Beijing–Hangzhou Grand Canal are shown in Figure 5.

It can be seen from Figure 5 that although the concentrations of the TN and TP in Taihu Lake and the Beijing–Hangzhou Grand Canal have partially exceeded the standards of Class III or Class IV WEQSC (GB3838-2002), they were still much lower than the average levels of TN and TP concentrations in urban rivers, as shown in Figures 2 and 3. In addition, the water storage volume of Taihu Lake and the Beijing–Hangzhou Grand Canal are more than other rivers in the city, and their dilution effect on pollutants is more obvious. Therefore, when the downstream of the river is connected with these water bodies with relatively clear water quality and huge water volume, the pollutants can be effectively diffused and decomposed, and the water quality downstream of the sluices is better than upstream.

#### **4. Conclusions**

The effect of the construction of sluices on the water quality of urban rivers in Wuxi was investigated, and the difference between TN and TP in the upper and lower reaches of the urban rivers after the construction of the sluices was detected. In this paper, the water quality in the urban rivers showed obvious seasonality and was usually better in seasons

that have more rainfall, such as summer and autumn. However, irregular sluice regulation often causes some water quality pollutant concentrations to rise abnormally. Additionally, in the state of sluice closure, the water quality of the urban rivers upstream was worse than that downstream; the concentrations of TN and TP downstream were, on average, 14.42% and 13.80% lower than that upstream, respectively. The concentrations of pollutants showed different degrees of variation with time and space, and there were discrepancies between the data and conclusions at individual monitoring sites. For example, the pollutant concentrations in the Ancient Canal and Huancheng River were higher downstream than upstream as well as the relative difference between the upstream and downstream pollutant concentrations in the Liangtang rivers was extremely obvious. In the future, additional sample sites and numerous data will be required for in-depth exploration.

**Author Contributions:** F.L.: Conceptualization, Methodology, Software. W.H.: Data collection and curation, Writing—original draft, preparation. Y.Y.: Software, Validation. Writing—review & editing. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the Oversea Study Fellowship from the China Scholarship Council, National Natural Science Foundation of China (No.42007151), China's National Key Research and Development Program (No.2017YFC0505803), the Natural Science Foundation of Colleges and Universities of Jiangsu Province (No.19KJB610015), the Philosophy and Social Science Foundation of Colleges and Universities of Jiangsu Province (No.2019SJA0108), Jiangsu Funding Program for Excellent Postdoctoral Talent (JB0206015). The authors wish to thank Nanjing University for providing MIKE software to be used in this study.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.
