Analysis of the characteristics of major pollutants discharged from wastewater in China’s provinces

In recent years, the discharge of major pollutants in China's wastewater has been decreasing but remains at a high level. Controlling the discharge of pollutants in sewage is of great importance for protecting water quality and maintaining ecological balance. Based on data collected from 31 provinces in China from 2011 to 2020 (except 2018), this study analyzes the spatiotemporal variation emissions of the wastewater pollutants: chemical oxygen demand (COD), ammonia nitrogen (NH3-N), total nitrogen (TN), and total phosphorus (TP). The entropy method was used to evaluate the effectiveness of water pollution control in different provinces. Our results revealed that the total emission per gross domestic product (GDP) for COD, NH3-N, TN and TP in China decreased by 50.7%, 81.9%, 65.4% and 70.8%, respectively. In terms of regional annual emission differences, the Northwest region was the lowest compared with other regions, accounting for 4.87%-6.59% of the national pollutant emissions, and the Central China region was the highest, accounting for 22.4%-26.05% of the national pollutant emissions. The average value of pollutant emissions per unit of GDP decreased year-to-year overall, but Guangxi and Tibet showed a trend of first decreasing and then increasing. The correlation results indicated a significant correlation (0.977) between TN and TP emissions in wastewater in China during 2011–2020. Through clustering and Multidimensional Scaling model (MDS) analysis, Beijing and Shanghai have been performing well in controlling water pollution discharge, while the provinces of Tibet and Guangxi must still increase their efforts in water pollution control. Furthermore, these results demonstrate the experience and achievements of the Chinese government in the treatment of wastewater pollution and provide a useful reference for treatment of wastewater pollution in the world.


Introduction
Water plays an important role in the entire ecosystem (water purification, water retention, and climate regulation) (Grizzetti et al., 2016).After the reform and opening, the rapid development of China's economy led to many streams coming from domestic sewage, agricultural sewage, industrial sewage and mine drainage that directly discharged into rivers and lakes.This excessive discharge exceeds the environmental carrying capacity, resulting in a series of ecological problems.For example, water pollution, shortage of water resources, and eutrophication of water bodies can lead to the gradual deterioration of the environment (Chen et al., 2016;Han et al., 2019;Lu et al., 2019;Marinelli, 2018;Tian et al., 2021;Wang et al., 2022).
China ranks among the world's largest contributors to wastewater discharge and eco-environmental damage (Liu et al., 2021;Zhang et al., 2023).Investigations showed that the eco-environmental damage caused by China's wastewaters discharges accounted for more than 1/4 of gross domestic product (GDP) in 2011-2015.Wastewater eco-environmental damage has decreased by about 50% in 2016 and 2017 compared with that in 2015, and the effect of government policies was remarkable (Liu et al., 2021).Given China's large geographical territory and marked regional differences in the main components and characteristics of sewage discharge, it becomes necessary to further understand the pollution levels associated with sewage discharge in various provinces throughout recent years.Such analysis will help identify disparities in pollutant discharge characteristics across regions and provide valuable data to inform future government decision-making.
Sewage discharge brings high concentrations of pollutants to most rivers in China, of which more than 40% are heavily polluted and 80% of lakes are eutrophic, and the threat to health from polluted waters has raised widespread public concern (Uddin & Jeong, 2021).Point source pollution such as industrial wastewater and domestic sewage, as well as non-point source pollution related to agriculture, all contribute to serious pollution loading on the nation's waters (Wang et al., 2015a, b).Industrial wastewater contains high concentrations of heavy metals, organic compounds and pathogens of a wide variety and complex composition; these are difficult to purify by natural degradation and are also difficult to treat (Bu et al., 2021).Domestic wastewater is a type of wastewater originating from daily human activities and consists mainly of detergents, soaps, vegetable oils, food residues, human waste, and organic substances and harmful compounds, which can lead to water pollution if discharged into rivers without proper treatment (Mustafa & Hayder, 2020, 2021).Non-point source pollution caused by large amounts of fertilizer inputs in agricultural production activities also causes significant pollution of surface water as well as groundwater (Bryan & Kandulu, 2011;Zhang et al., 2019).
Therefore, after recognizing the current situation of polluted waters, the importance of water quality monitoring has been emphasized, and many researchers have proposed various water quality monitoring models such as diffusion and receptor models to evaluate water quality (Bhanuprasad et al., 2008;Cao et al., 2015;Huang et al., 2010;Li et al., 2003;Liang et al., 2015;Liu et al., 2020;Wang et al., 2015a, b).In recent years, multivariate statistical methods (MSTs) such as cluster analysis (CA), discriminant analysis (DA), principal component analysis/factor analysis (PCA/FA), and multiple linear regression (MLR) have been effectively applied for the assessment of water quality (Shrestha & Kazama, 2007;Simeonov et al., 2003;Singh et al., 2005;Su et al., 2011).
The Chinese government regards these indicators, such as chemical oxygen demand (COD), ammonia nitrogen (NH 3 -N), total nitrogen (TN) and total phosphorus (TP), as part of the binding indicators for water environmental protection, and these indicators are also widely used in water quality control worldwide (Cardoso et al., 2007;Carvalho et al., 2008;Heiskary & Wilson, 1988;Huo et al., 2013;Paul & Gerritsen, 2002;Poikane et al., 2010;Walker et al., 2007).China has also gradually standardized various environmental protection regulations since the promulgation of the first Environmental Protection Law (for trial implementation) in 1979, with special emphasis on sewage treatment (Shao, 2010).In recent years, the Chinese government has made TP, NH 3 -N and COD indicators as binding targets for environmental protection planning.From the "10th" to the "13th Five-Year Plan", COD emissions have been reduced by 10%, 10%, 8%, and 10%, respectively; NH 3 -N emissions have been reduced by 10% except for the "12th Five-Year Plan".The control measures for TP have also changed from key basin control to total control in coastal areas, and TP emissions have dropped by 10% during the "13th Five-Year Plan" period.Therefore, our goal is to analyze the characteristics of the main pollutants discharged from wastewater in Chinese provinces, identify the sources of pollution in the most polluted provinces, and analyze their spatial and temporal changes.
For our knowledge, there are few published studies on the empirical and subjective evaluation of the Vol.: (0123456789) characteristics of sewage discharge in various provinces.Hence, this study provides important guidance for government departments to develop sewage control and management plans tailored to local conditions.It aims to examine the emission characteristics of COD, NH 3 -N, TN, and TP in wastewater.Additionally, it compares the emissions between different regions in China from 2011 to 2020 (except 2018).Therefore, the effectiveness of water pollution control in various provinces is evaluated using the entropy value method and cluster analysis and MDS analysis are conducted to identify the provinces that demonstrated the most effective control of pollutant discharge.

Data sources and analysis
In this study, COD, NH 3 -N, TN, and TP contents in wastewater discharged from 22 provinces, 5 autonomous regions, and 4 municipalities directly under the central government of China from 2011 to 2020 were collected (data from 2 special administrative regions of Hong Kong and Macau and Taiwan Province were not counted).COD, NH 3 -N, TN, and TP data were collected from the China Statistical Yearbook (National Bureau of Statistics of China (NBSC), 2011(NBSC), -2021)) The correlation analysis of the four pollutants discharged in the water bodies and the variability analysis of different geographical regions were carried out using SPSS 21.0 (International Business Machines Corporation, USA).The GDP-based pollutant emission intensity data for each province were calculated by Eq. ( 1) (COD, NH 3 -N, TN, and TP in wastewater were divided by the GDP of the corresponding province), (Yang et al., 2020).
where P ei is the pollutant emission intensity, T/billion RMB-year, P e is the pollutant emission, T, and GDP is the gross domestic product, billion RMB-year.
Based on the regional pollutant emission intensity values, the entropy value method was used to analyze the pollutant emission intensity status of each province in 2011 to 2020, and to analyze and evaluate the ranking of each province.

Standardization of data
The pollutants (COD, NH 3 -N, TN, and TP) per unit of GDP wastewater discharge were used as evaluation indicators, and all data are standardized or dimensionless and all indicators are negative indicators which can be calculated by the following Eq.( 2) (Wang et al., 2022;Yang et al., 2020): where X is the normalized assignment of data metrics and X i is the raw data collected for each metric.

Determination of index weights-entropy method
The entropy method is employed to determine the weights of the above four indicators based on the degree of variation of their actual measured values, which is an objective assignment method.The specific calculation process is divided into the following three steps, as the following Eqs.(3), ( 4), ( 5), and (6) showed: (1) where U j is entropy values of each indicator, R j is entropy inversion, W j is weights, C i is composite index, m is the number of provincial-level regions, n is the number of evaluation indicators; and S ij is the normalized data of the j-th indicator for the i-th provincial-level region (Wang et al., 2022). (5)

Analysis of the emission characteristics of COD, NH 3 -N, TN, and TP in wastewater
The annual levels of COD in wastewater discharged from each province in China between 2011 and 2020 are presented in Fig. 1.From 2011 to 2015, there were no significant overall changes in COD levels.However, in 2016 and 2017, the levels of COD during these two years exhibited a significant decrease compared to other years.This result is consistent with previous research findings (Liu et al., 2021).Compared with the past, the COD of some regions in 2019 and 2020 changed significantly, such as Hubei, Guizhou and Tibet.During this period, the maximum annual emission of COD is 1.982 million tons in Shandong Province in 2011.
Using the COD of each province at from the aforementioned period, the annual average value of COD emissions in each province can be calculated (million tons/year).Based on this comparison, 10 provinces exceeded 1 million tons/year, namely; Shandong, Henan, Hunan, Heilongjiang, Sichuan, Jiangsu, Hebei, Hubei, Liaoning.Among these provinces, Guangdong had the highest value of COD emissions, reaching 1.549 million tons/year.The annual average value of COD emissions in the five northwestern provinces of Xinjiang, Shaanxi, Gansu, Ningxia, Qinghai and Tibet were below 0.4 million tons/year, with the two areas of Qinghai and Tibet having the lowest emission values (0.089 million tons/year) for each of the above provinces.
Between 2011 and 2015, the overall NH 3 -N change profile was similar, and in 2016 and 2017, the overall level of NH 3 -N in these two years shows a significant decrease compared to previous years.The overall level of NH 3 -N emissions in 2019 and 2020 was similar, with significant decreases observed in the peak points of NH 3 -N emissions such as Hebei, Liaoning, Heilongjiang, Jiangsu, Shandong, Guangdong, and Sichuan compared to 2011.During this period, the maximum annual emission of NH 3 -N was 0.231 million tons (Guangdong, 2011).With the exception of Tibet, all provinces showed a decreasing trend in annual ammonia nitrogen emissions year-to-year.
Based on the annual NH 3 -N emissions in each province at the above time, it can be concluded that the average annual emissions (million tons/year) totaled more than 0.1 million tons in Guangdong, Hunan, Shandong, Jiangsu, Sichuan, and Henan.Guangdong had the highest value of NH 3 -N emissions in those six provinces (0.173 million tons/year).NH 3 -N emissions in the annual average value of 0.02 million tons/year below the city were Tianjin, Hainan, Beijing, Ningxia, Qinghai and Tibet, where the Tibet region has the lowest emissions value of the above provinces (0.003 million tons/year).
The overall profile change of TN was relatively similar across the country between 2011 and 2015.After 2016, there was a significant change in the overall change profile of TN, where the peak points in the overall curve of TN emissions in Shandong, Hebei and Heilongjiang had a significant decrease in TN emissions compared to 2011.The overall profile curve of TN emissions in 2019 and 2020 was closer, with TN emissions in Guangdong being the peak point of the curve.During this period, the maximum emission of TN was 0.677 million tons (Shandong, 2015).
According to the annual value of TN emissions in each province, the annual average value of TN emissions in each province (million tons/year) can be obtained, and in order of the size of this value, the top ten provinces were Shandong, Henan, Hebei, Guangdong, Sichuan, Hunan, Heilongjiang, Jiangsu, Hubei, and Anhui.Shandong had the highest value of NH 3 -N emissions of those 10 provinces (0.394 million tons/year).The provinces with annual average values of TN emissions below 0.03 million tons/year were Tianjin, Beijing, Ningxia, Qinghai and Tibet, with the Tibetan region having the lowest emission value (0.007 million tons) of the above provinces.
The overall contours of TP changes were relatively similar across the country between 2011 and 2015.After 2016, the peak points in the overall curve of TP emissions in Shandong, and Hebei had a significant decrease in TP emissions.The overall profile curve of TP emissions in 2016 and 2017 was relatively similar.The overall profile curve of TP emissions in 2019 and 2020 was also relatively similar.In the overall curve of TP in 2020, the TP emissions in Guangdong became the peak point of the curve.
During the above-mentioned years, the maximum value of annual TP emissions in Shandong Province in 2015 was 0.082 million tons.Based on the annual values of TP emissions in each province, the average value of TP emissions (million tons/year) can be calculated for each province, and the top ten provinces in terms of annual average value of TP emissions were Shandong, Guangdong, Henan, Hebei, Hunan, Sichuan, Hubei, Liaoning, Heilongjiang, and Anhui.Shandong had the highest value of TP emissions of those 10 provinces (0.042 million tons/year).The provinces with annual average values of TP emissions below 0.003 million tons/year were Beijing, Ningxia, Shanghai, Tibet, and Qinghai.Qinghai had the lowest value of TP emissions of those five provinces (0.0009 million tons/year).
Vol:. ( 1234567890) Inter-regional wastewater discharge variation analysis In order to understand the regional differences of the pollutants discharged from the above wastewater, the regional differences of COD, NH 3 -N, TN and TP were analyzed and compared according to the geographical division of China, and the results were shown in Fig. 2.
Based on the analysis of COD emission differences among different regions, it is evident that between 2011 and 2020, the average value of COD emissions in Northwest China was the lowest compared with other regions, while Central China had the highest average emissions among all regions.There was a significant difference between the emissions in Northwest China and Northeast and Central China before 2016.COD emissions in Northeast China decrease and there was no significant difference with the emissions in Northwest China significant difference; while the emissions in central China were still significantly different from those in northwest China after 2016.
From the analysis of the differences in NH 3 -N emissions in different regions, it can be seen that between 2011 and 2020, the average value of NH 3 -N emissions in Northwest was the lowest compared to other regions, while the average value of emissions in Central China was basically the top of all regions.There were significant differences between Northwest China and Central China, East China, and South China; there were no significant differences between Central China, South China, East China, and Northeast China in 2011.The regional differences in the average ammonia emissions were mainly between Northwest China and Central China after 2012.There was no regional difference in TN emissions between regions in 2011, this may be related to the provinces and cities of administrative divisions, and the range of changes in these data is relatively large.There was a significant difference between TN emissions in Central China and Northwest China between 2012 and 2014.There was no regional difference in TN emissions between regions in 2015.There remained a significant difference between TN emissions in Central China and Northwest China after 2016.
There was no difference in TP emissions between regions in 2011, and no difference in TP emissions between regions in 2015, 2016 and 2020.There were significant differences between Central China and both regions, Northwest and Southwest in 2012.There were significant differences between Central China and Northwest China during 2013, 2014, 2017 and 2019.

Analysis of the characteristics of pollutants discharged per unit of GDP in wastewater
The COD emissions per unit GDP by province were shown in Fig. 3. Between 2011 and 2017, the overall COD emissions per unit GDP showed a decreasing trend year-to-year.In 2019 and 2020 COD emissions per unit GDP increased significantly in some provinces, such as Tibet and Guizhou.Between 2011 and 2020, the maximum value of annual COD emissions per unit GDP was 27.881 T/billion RMB-year (Tibet) and the minimum value was 0.148 T/billion RMB-year (Beijing).Based on the COD per unit GDP of each province at the above time, the annual average value of COD emissions per unit GDP (T/billion RMB-year) can be calculated, and the top three were Heilongjiang, Ningxia, Xinjiang.Heilongjiang has the highest value of COD emissions per unit GDP (7.961 T/billion RMB-year) of those three provinces.The least three were Tianjin, Shanghai, and Beijing.Beijing had the lowest value of COD emissions per unit GDP (0.63 T/billion RMB-year).
From 2011 to 2020, the majority of provinces showed an overall decreasing trend in NH 3 -N emissions per unit GDP year by year; only individual provinces such as Guangxi and Tibet showed a decreasing and then increasing trend in NH 3 -N emissions per unit GDP.Between 2011 and 2020, the maximum value of NH 3 -N emissions per unit GDP was 0.902 T/ billion RMB-year (Hainan) and the minimum value was 0.008 T/billion RMB-year (Beijing).Based on the NH 3 -N emissions per unit GDP in each province at the above time, the annual average value of NH 3 emissions per unit GDP can be calculated, and the top three cities, ranked from most to least, were Hainan, Jiangxi and Hunan.Hainan had the highest value of NH 3 -N emissions per unit GDP (0.501 T/billion RMB-year) of those three provinces.The cities with the least were Shanghai, Tianjin, and Beijing.Beijing had the lowest value of NH 3 -N emissions per unit GDP (0.063 T/billion RMB-year).
Between 2011 and 2020, the majority of provinces showed an overall decreasing trend in TN emissions per unit GDP year-to-year; only individual provinces such as Guangxi and Tibet show a U-shaped trend in TN emissions per unit GDP.In the period 2011-2020, the maximum value of TN emissions per unit GDP was 2.581 T/billion RMByear (Xinjiang) and the minimum value was 0.03 T/ billion RMB-year (Beijing).Based on the TN emissions per unit GDP in each province at the above time, the annual average value of TN emissions per unit GDP was calculated, and the top three cities, from most to least, were Xinjiang, Heilongjiang, and Hainan.Xinjiang had the highest value of TN emissions per unit GDP (1.31 T/billion RMB-year) of those three provinces.The bottom three were Tianjin, Beijing, and Shanghai.Shanghai had the lowest value of TN emissions per unit GDP (0.111 T/billion RMB-year).
From 2011 to 2020, Hebei and Heilongjiang showed an overall significant decreasing trend of TP emissions per unit GDP year by year; however, Guangxi, Guizhou and Tibet showed a decreasing and then increasing trend of TP emissions per unit GDP.The remaining provinces showed a decreasing trend in general.From 2011 to 2020, the maximum value of TP emissions per unit GDP was 0.278 T/billion RMB-year (Hebei) and the minimum value was 0.001 T/billion RMB-year (Beijing).Based on the TP emissions per unit GDP for each province at the above time, the annual average value of TP emissions per unit GDP was calculated, and the top three cities were Heilongjiang, Hainan, and Hebei.Heilongjiang had the highest value of TP emissions per unit GDP (0.127 T/billion RMB-year) of those 3 provinces.The bottom three cities were Tianjin, Beijing, and Shanghai.Shanghai had the lowest value of TP emissions per unit GDP (0.008 T/billion RMB-year).

Comprehensive evaluation of water pollution control effects in different provinces
The effectiveness of water pollution control in different provinces from 2011 to 2020 was evaluated using the entropy method (Fig. 4).In order to increase the comparability across years, after taking the entropy value inverse to determine the index weights for COD, NH 3 , TN and TP, the composite index of each index was derived, and the intensity was classified into four levels using natural folding points.From a national perspective, from 2011 to 2016, the intensity of pollutant discharge in wastewater gradually decreases, then the intensity of pollutant discharge in wastewater gradually increased.The results revealed that the total emission of per GDP for COD, NH 3 -N, TN and TP in China decreased by 50.7%, 81.9%, 65.4% and 70.8%, respectively.Beijing and Shanghai were the regions with the highest comprehensive ranking from 2011 to 2020 due to the lowest emission intensity, while Tibet and Guangxi were at the lowest comprehensive ranking in 2020, suggesting these two provinces have high wastewater emission.The province with the greatest decline in the comprehensive ranking was Tibet, which fell from 14th (2011) to 31st (2020), followed by Guizhou, which fell from 15th (2011) to 28th (2020).The province with the most significant rise in the overall ranking was Henan Province, which rose from 21st (2011) to 11th (2020).The rankings of the remaining provinces did not change.
Regarding the indicator weights, it is observed that TP had the highest weight value, followed by COD.These  two indicators also have the greatest impact on the ranking of the provinces.Therefore, in the context of wastewater emission reduction, Tibet and Guangxi provinces still need to put in considerable efforts, particularly in reducing the intensity of TP and COD emissions.It is crucial for them to further intensify their measures in these areas.

Correlation analysis of each pollutant index in wastewater
Correlation analysis was conducted on the indicators of each pollutant in the wastewater for each of the mentioned years, and the results are shown in Table 1.
The table reveals that from period 2011 to 2020, the mean correlation values between COD, NH 3 , TN and TP emissions at the 0.01 level were 0.906, 0.886, and 0.897, respectively, and the mean values of correlation coefficients between NH 3 and TN and TP were 0.824 and 0.833, respectively.The correlation coefficients of NH 3 and TN and TP were 0.824 and 0.833, respectively, and the correlation coefficients of TN and TP were 0.977.In order to explore the interrelationship of the main pollutants in wastewater discharge, a linear regression Fig. 5 Cluster analysis (A) and multidimensional scale analysis (B) of pollutants in wastewater from each province in China analysis was performed for each indicator.Assuming that COD is y, the linear relationship can be obtained by regression analysis with three indicators NH 3 -N (x 1 ), TN (x 2 ), and TP (x 3 ), respectively.As showed in the following Eqs.( 7), ( 8) and ( 9).
According to the correlation analysis of pollutants discharged from water, it was found that the correlation between TN (x 2 ) and TP (x 3 ) content is extremely high (R 2 = 0.977).Regression analysis was conducted on the content of the two, and the following Eq.( 10) was obtained: Cluster analysis and multidimensional scale analysis of pollutants in wastewater Cluster analysis and multidimensional scale analysis of wastewater discharges in each province were performed using Primer 6 (Fig. 5).The results showed that three clusters were formed.The results of the multidimensional scale analysis were consistent with the results of the cluster analysis.

Conclusions
This study analyzed the levels of COD, NH 3 -N, TN, and TP discharged from wastewater in 31 provincial-level administrative regions in China.The findings indicated that the average COD and TN discharges in these regions exhibit a fluctuating trend from 2011 to 2020, with a decrease followed by an increase.In contrast, the average NH 3 -N and TP discharges demonstrate a consistent decreasing trend over the years.In terms of regional variations, the Northwest region had the lowest emissions compared to other regions, while the Central China region had the highest emissions.From 2011 to 2017, there was a decline in the average COD emissions per unit of GDP.However, from 2019 to 2020, an increase in COD emissions per unit of GDP was observed.From 2011 to 2020, the emissions of NH 3 -N per unit GDP, TN per unit GDP and TP per unit GDP in most of the provinces showed a decreasing trend year-to-year, but the emissions in Guangxi and Tibet showed a decrease and then and increase.The correlation results indicated a significant correlation (0.977) between TN and TP emissions in wastewater in China during 2011-2020.Clustering and MDS analyses showed that Beijing and Shanghai had the lowest emissions in terms of all indicators.Therefore, Beijing and Shanghai have been performing quite well in controlling the intensity of water pollution discharge from 2011 to 2020.Conversely, Tibet and Guangxi require concerted efforts to strengthen the control and emission reduction of pollutants.The results of this work explored the differences in the emission intensity of major pollutants in various provinces of China, as well as the potential for emission reduction.Furthermore, it is considered an important practical applications for proposing policy recommendations to reduce the emission intensity of pollutants in various provinces and regions of China.

Fig. 1
Fig. 1 Emissions of different pollutants in wastewater; A COD emissions, B NH 3 -N emissions, C TN emissions, and D TP emissions

Table 1
Beijing and Shanghai clustered into one category at Euclidean distance 83.15; Tianjin, Jiangsu, Zhejiang, Inner Mongolia, Shandong, Guangdong, Shaanxi, Fujian, and Chongqing clustered into one category at Euclidean distance 77.8; and other provinces clustered into one group at Euclidean distance 81.94.From 2011 to 2015, Beijing demonstrated the lowest indicators of COD per unit GDP and NH 3 -N per unit GDP among all regions in China, while Shanghai exhibited the lowest indicators of TN per unit GDP and TP per unit GDP; in 2016 and 2017, Beijing's indicators were the lowest, while Shanghai's emissions were second only to Beijing; in 2019 and 2020, both Beijing and Shanghai's indicators were the lowest in China.Therefore, Beijing and Shanghai were clustered into a separate category in the cluster analysis.