Water Environment Management and Performance Evaluation in Central China: A Research Based on Comprehensive Evaluation System

: As a developing country with insu ﬃ cient water resources, China’s water environment management and performance evaluation have important research value. The three provinces (Henan, Hubei


Introduction
Water is the source of life, the key to production, and the foundation of ecology [1][2][3]. Water is one of the most valuable and irreplaceable resources in the world, on which all the life on Earth depends for survival and development [4][5][6]. With the rapid development of the economy, population (1) A comprehensive evaluation model based on PCA and the PSR model was constructed to analyze the sustainable development of water environment in central China. The main advantage of PCA is that it can effectively reorganize discrete variables by mathematical statistical methods and reflect the data characteristics by a few variables. The main advantage of the PSR model is that it highlights the causal relationship between the environment and the stress facing the environment, as well as the mutual restriction and interaction between the three layers of stress, state, and response. Hence, the comprehensive evaluation model in this paper can determine a few composite variables from various variables to replace the existing variables by mathematical dimension reduction methods to explore the causal relationship between human activities and environmental changes based on the evaluation of the sustainability of environmental systems. (2) A performance evaluation system for water environment management, which could comprehensively evaluate the performance of water environment treatment and effectively reveal the correlation between various indicators, was established. The principal factors in water environment management can be obtained by this evaluation system. Therefore, the evaluation indicator system and the weights of different indicators in this system can be determined for quantitative calculation by substituting the standardized values into the indicator system. This performance evaluation system can be used to evaluate the performance of water environment management and sustainable development. After careful selection of specific indicators and use of official statistics from the three provinces in central China, the objectivity of calculation results was ensured in this paper to contribute to evaluate the performance of water environment management and sustainable development in China.
The structure of this paper is as follows: Section 2 is the literature review, Section 3 introduces the research methods used in this paper, Section 4 lists the calculation results, Section 5 analyzes the water environment management and performance evaluation of the three provinces in central China from 2011 to 2017, and Section 6 summarizes the findings in this paper and provides corresponding policy recommendations.

Literature Review
It is generally agreed by the academia that performance evaluation should consider various factors including efficiency, effectiveness, and satisfaction [26][27][28].
Among the existing studies, Lu et al. established a credibility-based optimization model for water resources management in South central China to show the confidence level of the optimal management strategies. Their results indicated that an aggressive strategy should be considered if system benefit is not the major concern of the government. They also suggested that part of system benefit could be sacrificed to protect local groundwater resources [29].
Cai et al. used the composite index method to conduct a spatiotemporal analysis of water resources vulnerability in China. They found that water resources in north and central China are more vulnerable than in the western area. Moreover, water pollution was worsening remarkably in central China, and water resource shortage has been one of the most serious challenge for sustainable development there [30].
Yao et al. investigated the 14 antibiotics in groundwater and surface water at the Jianghan Plain in central China. They demonstrated that the total concentrations of antibiotics in the spring samples were higher than those in summer and winter. By the risk quotient and mixture risk quotient methods, they evaluated the environmental risks for surface water and groundwater in central China [31].
Hu et al. analyzed 13 antibiotics in the Hanjiang River, one of the main rivers in central China. Their results showed that the hazard quotients of antibiotics were higher in the sediment than those in the water body of the Hanjiang River. Moreover, antibiotic mixtures posed higher ecological risks to water resource in central China than aquatic organisms [32].
Jia et al. constructed an index system to quantify the water environmental carrying capacity. They showed that the potential of water environmental carrying capacity is decreasing from the east China to the west. Moreover, the water resource vulnerability in the west is higher than that of central China [33].
Zhou et al. established a non-radial directional distance function to measure the performance of water use and wastewater emission. Their results indicated that eastern China performs better than central China, with the average technology gap of 51%. Since the technological heterogeneity directly affected the environmental efficiency of industrial water in China, they also assessed the technological efficiency of each province and provided corresponding improvement targets for them [34].

Principal Component Analysis
The PCA method was first introduced by the American statistician Pearson in the study of biological theory [35]. The main idea is to reorganize discrete variables by mathematical statistical methods and attempt to reflect the data characteristics by a few variables [36][37][38]. This method determines a few composite variables from various variables to replace the existing variables by mathematical dimension reduction methods, such that these composite variables contain as much amount of information as the original variables and are independent from each other. This method could remove overlapping information in quantitative analysis in order to reflect the same amount of information with a minimum number of mathematical variables [39,40].
The PCA method uses variance as a measure of information amount. It attempts to reorganize the various existing variables with certain correlation with each other into a new set of mutually independent composite variables to replace the existing variables. If the first linear combination selected, i.e., the first composite variable, is denoted as F1, and the information amount carried by each variable is measured by the variance, then the larger the Var(F1) value, the larger the amount of information is contained. Therefore, among all the linear combinations, the F1 with the largest variance should be selected. Such F1 is also called the first principal component. If the first principal component could not sufficiently represent all the information contained in the original p variables, a second linear combination should be considered. In order to effectively reflect the information in the original variables, the information contained by F1 does not need to be covered by F2 again. By applying the same mathematical method, F2, i.e., the second principal component, could be obtained given that Cov(F1, F2) = 0. By the same methods, the third, the fourth, the fifth, . . . and the pth principal component could be determined.
Based on this method, this paper constructed a matrix of water environment sample data of central China: where X ij stands for the jth indicator of the ith data.
(1) Standardize the raw data X: The x ij in the above equation is the observed sample data and x * ij is the standardized data, where x j is the average of the jth indicator: var(x j ) is the standard deviation of the jth indicator: (2) Construct a correlation coefficient matrix R for the standardized data x * ij : R is a p × p matrix in which the element r ij can be defined as: (3) Calculate Eigenvalues and Eigenvectors In the above formula, λ i (i = 1, 2, 3 · · · , p) is the eigenvalue and E is an identity matrix of the same order as R. By solving the above formula, the eigenvalues can be obtained. The eigenvalues were further sorted by value. The eigenvalue λ i represents the variance of the ith principal component, reflecting the degree of influence of each principal component. The contribution rate of Principal Component A i to the variance can be written as: The cumulative contribution rate of the first n principal components to the variance can be written as:

(5) Determine the Principal Components
Based on the standardized raw data, the contribution rates of different principal components can be obtained by substituting the principal components into the expressions above.

The Comprehensive Evaluation Method Based on the PSR Model
The PSR model was developed by Rapport and Friend in Canada to assess the impact of human activities on the ecological environment [41]. This model highlights the causal relationship between the environment and the stress facing the environment, as well as the mutual restriction and interaction between the three layers of stress, state, and response [42][43][44]. The main purpose of the PSR model is to explore the causal relationship between human activities and environmental changes based on the evaluation of the sustainability of environmental systems [45,46]. Therefore, the PSR model can be used to study the sustainable development of the water environment in central China.
The water environment is a dynamic environment. This paper adopted the PSR model to study the changes in water environment in central China during the study period and to further analyze the sustainability of the water environment. To evaluate water environment sustainability under the PSR framework based on the construction of distance function and discrete coefficients, the formula following formula was used: In the above formula, CI is the coordination degree function, and X 1 , X 2 , and X 3 represent the scores corresponding to the pressure, state, and response layers, respectively. The closer the scores under the pressure, state, and response layers to each other, the closer the coordination coefficient is to √ 3, indicating a better sustainability level.

Comprehensive Evaluation of the Performance of Water Environment Management and Sustainable Development
Through calculation based on the above method, this paper constructed an evaluation indicator system and determined the weights of different indicators in this system. Next, this paper performed quantitative calculation by substituting the standardized values into the indicator system. The specific method is: In the formula above, ICP is the water environment management index, P i is the indicator value, and W i is the weight of the indicator. This index can be used to evaluate the performance of water environment management and sustainable development. As can be seen from the above formula, the value of the index should range from [0, 1].

Indicator Selection and Data Source
In the selection of specific indicators, this paper emphasized the principle of comprehensiveness and objectivity to ensure that the indicator system could comprehensively evaluate the performance of water environment management and sustainable development. The data of the indicators were all from official statistics to ensure the objectivity of calculation results, and the study period was from 2011 to 2017 [47][48][49][50][51]. The finalized indicator system is shown in Table 1. In the above table, the indicators of the pressure layer were measured by the discharge of major pollutants. The lower the indicator value, the lower the pressure on the water resources caused by pollutant emission during economic development. The indicators of the state layer were divided into two categories: The gross domestic product and the change in population. The higher the indicator value, the bigger achievement in water quality improvement. The indicators of the response layer represent the expenditure or investment of the government in order to take actions against water pollution. The higher the indicator value, the more emphasis the local government has put on water pollution control and the stronger the enforcement.

Results
Based on the model and methodology introduced in Section 3, as well as the indicators selected, this paper obtained the below calculation results from the PSR model (as shown in Table 2): As can be seen from the eigenvalues and variance contribution rates in Table 2, there were five indicators whose eigenvalues are greater than 1, which thus became the candidates of the principal component variables. These variables are: NH 3 -N Emissions, Natural Population Growth Rate, Mercury Emissions, Regional Secondary Industry Output, and Investment in Ecosystem Construction and Protection, whose cumulative variance contribution rate reached 82.44%, indicating that the five principal component variables could explain 82.44% of the information contained in the 21 indicators. These principal component variables were then sorted by their variance contribution rates and expressed as Z1, Z2, Z3, Z4, and Z5 respectively. The factor variance contribution rates are shown in Table 3 below. The rotated factor load matrix indicates the correlation between the 21 indicators and the five principal components, as shown in Table 4.  It can be seen from Table 4 that: ( Therefore, Z1 and Z3 could be defined as the principal components of the stress layer, which comprehensively reflect the overall conditions of the pressure indicators; Z2 and Z4 could be defined as the principal components of the state layer, which comprehensively reflect the overall improvement of the state indicators; and Z5 could be defined as the principal component of the response layer, which comprehensively reflects the overall conditions of the response indicators. Based on the calculation method introduced in Section 3, this paper further obtained the expressions of Z1, Z2, Z3, Z4, and Z5 (see Equations (12) The component score matrix of Equations (12)- (16) is also shown in Table A1. The evaluation scores of the above five principal components can be integrated into one Comprehensive Evaluation Index Z, as shown in (17) below:

Discussion
Based on the above expressions of the Pressure Index Z1, the State Improvement Index Z2, the Response Index Z3, and the Comprehensive Evaluation Index Z, this paper obtained the scores of each index in each of the central China provinces within the study period and made further comparison on the index scores by year and by province, respectively (see Figure 1 below and Table A2).
This paper further discretized the Comprehensive Evaluation Index (Sustainability Index) Z in order to define the corresponding intervals for each sustainability level. The results are shown in Figure 2 below and Table A3: Per the common standards followed by academic researches, a discretized evaluation value between 0 and 0.3 indicates a poor level of sustainable development, a discretized evaluation value between 0.3 and 0.6 indicates a medium level of sustainability, a discretized evaluation value between 0.6 and 0.9 indicates a good level of sustainability, and a discretized evaluation value above 0.9 indicates an excellent level of sustainable development [52,53]. It can be seen from the data in Table A3 that the level of sustainable development of the three provinces in central China during 2011 and 2012 was generally poor, but it was on an improvement trend. The overall level of sustainable development was in the medium range during 2013 and 2014, which improved compared with the previous two years, but there was still room for improvement. By 2015-2016, due to the government's strong implementation of environmental protection policies, strengthened environmental protection supervision, and the introduction of a series of laws and regulations such as the Environmental Protection Law, the sense of responsibility for environmental protection became deeply rooted in the hearts of the people [54][55][56]. Therefore, during this period and beyond 2017, the sustainability level of the water environment in these provinces has seen huge improvements.
It can be noted by sorting the discretized comprehensive evaluation scores in 2017 that the Hubei Province had the best sustainability level in water environment, the Henan Province achieved the biggest improvement in terms of water environment sustainability, and the Hunan Province's sustainability level in water environment was medium. Overall, the sustainability level of water environment in central China has improved. Based on Section 3.3, this paper further calculated the comprehensive evaluation scores of the water environment management and sustainable development performance in the three provinces of central China (see Figure 3 below and Table A4). This paper further discretized the Comprehensive Evaluation Index (Sustainability Index) Z in order to define the corresponding intervals for each sustainability level. The results are shown in Figure 2 below and Table A3: Per the common standards followed by academic researches, a discretized evaluation value between 0 and 0.3 indicates a poor level of sustainable development, a discretized evaluation value between 0.3 and 0.6 indicates a medium level of sustainability, a discretized evaluation value between 0.6 and 0.9 indicates a good level of sustainability, and a discretized evaluation value above 0.9 indicates an excellent level of sustainable development [52,53]. It can be seen from the data in Table   - This paper further discretized the Comprehensive Evaluation Index (Sustainability Index) Z in order to define the corresponding intervals for each sustainability level. The results are shown in Figure 2 below and Table A3: Per the common standards followed by academic researches, a discretized evaluation value between 0 and 0.3 indicates a poor level of sustainable development, a discretized evaluation value between 0.3 and 0.6 indicates a medium level of sustainability, a discretized evaluation value between 0.6 and 0.9 indicates a good level of sustainability, and a discretized evaluation value above 0.9 indicates an excellent level of sustainable development [52,53]. It can be seen from the data in Table   -  biggest improvement in terms of water environment sustainability, and the Hunan Province's sustainability level in water environment was medium. Overall, the sustainability level of water environment in central China has improved. Based on Part 3.3, this paper further calculated the comprehensive evaluation scores of the water environment management and sustainable development performance in the three provinces of central China (see Figure 3 below and Table A4).  (1) Ammonia nitrogen refers to the nitrogen in water in the form of free ammonia and ammonium ions. Human activities have caused nitrogenous substance to enter the water environment mainly through untreated urban household wastewater and industrial wastewater, as well as various kinds of leachates. The main reason why ammonia nitrogen exceeds the acceptable standard is that the designed size of the sewage treatment facility is too small and the treatment equipment is underloaded, so the free ammonia in the sewage cannot fully complete the nitrification reaction. In addition, excessive sewage discharge has also resulted in a sharp increase in ammonia nitrogen, which has seriously hammered the sustainable development of the water environment. During the study period, the Henan Province strictly regulated sewage discharge, achieved an overall balance of water resources by reducing ammonia nitrogen emissions, strengthened the promotion of water resource protection, and made great efforts to enhance the sense of responsibility of all sectors of society for water resource protection [57]. At the same time, the Henan Province actively introduced highly efficient energy-saving technologies to timely process the sewage, regularly investigated and monitored the sources of water pollution, and conducted statistical analysis on sewage treatment results to derive the performance of water pollution control during defined periods of time, which helped the Henan Province achieve satisfactory water environment management results [58]. (2) As one of the key indicators defining the sustainability of water environment, the Natural Population Growth Rate reflects the relationship between human and the nature, as well as the social aspect of environmental protection, industrialization, and urbanization. As a populous province, the Henan Province strictly implemented the family planning policy in order to control the natural population growth rate and actively utilized market-based approaches to adjust the natural population growth rate in the context of the Chinese government gradually liberalizing the birth control policies in China [59,60], thus contributing to the sustainable development of the water environment.

(3) As a response layer indicator for sustainable development, the Investment Amount in Ecosystem
Construction and Protection reflects the sense of responsibility and commitment of the local enterprises and government regarding ecological environment construction. During the study period, the average annual investment in ecosystem construction and protection in the Henan Province was around 7 billion RMB [61], which exceeded the investment amount by other provinces. This also explains the significant improvement in water environment protection and sustainable development achieved by the Henan Province in the past five years.
The Hubei Province, which showed the best overall sustainability level during the study period, did not only take a series of measures in the above key areas that contribute to the sustainable development of water environment as the Henan Province [62], it also paid more attention to scientific and technological innovation, such as adopting the new clean wastewater treatment technology in the treatment and control of pollutants including mercury [63,64]. Therefore, the Hubei Province achieved outstanding pollution control results in terms of the pressure layer indicators such as ZP6.
Basing on the actual conditions of water environment management in the three provinces of central China and the availability of data, we mainly selected Afforestation Area, Constructed Wetland Area, Comprehensive Utilization of General Industrial Solid Waste, Investment in Industrial Wastewater Treatment, Investment in Industrial Waste Treatment, and Investment in Ecosystem. Construction and Protection were the indicators of results. In future research, we will further supplement the indicators as references to the real effects of the pressures and the corrections of the externalities caused by human activities. These indicators include, but are not limited to, the conditions and impact of the discharges of treated wastewater on the natural environment, the increase in corporate profits brought about by the recycling of wastewater, the costs saved by the recycling of wastewater (such as management fees and sewage charges), fines for compensation for water environmental treatment, etc.

Conclusions
This paper selected the performance of water environment management and sustainable development in the three provinces of central China as the research object and constructed a comprehensive evaluation system for water environment management and sustainable development by integrating the PCA method and the PSR model in order to comprehensively analyze the performance of water environment management and sustainability of development. The constructed evaluation system could comprehensively analyze the result of water environment treatment in a certain region and is able to effectively reveal the correlation between different indicators, thus determining the principal factors in water environment management. With the help of this system, this paper evaluated the performance of water environment management in the three provinces of central China from 2011 to 2017.
The results show that the sustainability level of the water environment in central China has shown an improvement trend during the study period, with the largest improvement seen during 2011-2014. The evaluation results vary among different provinces. The Henan Province has experienced the most significant improvement during the study period. Its comprehensive evaluation score of water environment management and sustainable development reached 1.671 in 2017, ranking second in central China. Overall, Hubei Province maintained the best water environment management and sustainable development level during the study period, with a comprehensive evaluation score of 1.692 in 2017.
The contributions of this paper are: (1) A comprehensive evaluation model based on PCA and the PSR model was constructed to analyze the sustainable development of water environment in central China. The main advantage of PCA is that it can effectively reorganize discrete variables by mathematical statistical methods and reflect the data characteristics by a few variables. The main advantage of the PSR model is that it highlights the causal relationship between the environment and the stress facing the environment, as well as the mutual restriction and interaction between the three layers of stress, state, and response. Hence, the comprehensive evaluation model in this paper can determine a few composite variables from various variables to replace the existing variables by mathematical dimension reduction methods, to explore the causal relationship between human activities and environmental changes based on the evaluation of the sustainability of environmental systems. (2) A performance evaluation system for water environment management, which could comprehensively evaluate the performance of water environment treatment and effectively reveal the correlation between various indicators, was established. The principal factors in water environment management can be obtained by this evaluation system. Therefore, the evaluation indicator system and the weights of different indicators in this system can be determined for quantitative calculation by substituting the standardized values into the indicator system. This performance evaluation system can be used to evaluate the performance of water environment management and sustainable development. After careful selection of specific indicators and use of official statistics from the three provinces in central China, the objectivity of calculation results was ensured in this paper to contribute to evaluate the performance of water environment management and sustainable development in China.
Based on the evaluation results, the authors proposed the following policy recommendations for the improvement of water environment management and sustainable development in central China: (1) Strengthen the promotion and education about the importance of sustainable development of the water environment, accelerate the accumulation of human capital in the provinces of central China, and raise people's awareness of water conservation. The provinces of central China should further increase the investment in the education of water resource protection knowledge and technologies to the citizens, cultivate their awareness of ecological protection related to the water environment, help the citizens form a habit of reducing water resource input in production as well as reducing water pollution emissions in daily life, and enhance the public's understanding of the ecological and social benefits of sustainable development of the water environment. At the same time, governments at all levels below the provincial level should place great emphasis on the sustainable development of the water environment, include it in the government's key agenda, and effectively strengthen the protection of the water environment based on the actual local conditions. (2) Establish a long-term incentive mechanism for the sustainable development of the water environment. The distribution of precipitation, the layout of industrial and agricultural production, and the level of economic development vary greatly among the provinces of central China. It is important to comprehensively consider the regional differences and the economic feasibility for the local residents when establishing a long-term mechanism to motivate the sustainable development of the water environment. For example, special funds could be appropriated to support the technology upgrade of water pipelines and surface water delivery [65], as well as water recycling technologies that have higher costs such as the micro-irrigation technology [66]. At the same time, the local governments should reduce the administrative interventions during the promotion of water environment improvement technologies in order not to burden the residents and enterprises while promoting the sustainable development of water environment. (3) Further increase investment in fixed assets for water pollution control. Compared with general fixed asset investment, the investment in environmental pollution control has its own positive environmental externalities and environmental spillover effects apart from the benefits of increasing household consumption and stimulating demand for related industries [67]. Therefore, investment in pollution control has more social and environmental implications than general fixed asset investment. It should be noticed that although the growth rate of fixed assets investment in water pollution control in these three provinces of central China has been higher than that of the overall environmental investment in recent years, there is still a gap in the proportion of pollution control investment in national income when compared with the average level of developed countries [47]. Thus, the investment in water pollution control should be further increased in the future.
Therefore, the public participation and long-term incentive mechanism for the sustainable development of the water environment will be included in future research. Meanwhile, the indicators, which reflect the real effects of the pressures and the corrections of the externalities caused by human activities to make our research more perfect, will also be supplemented.

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