Characteristics and Adaptability Assessment of Commonly Used Ecological Flow Methods in Water Storage and Hydropower Projects, the Case of Chinese River Basins

: The construction and operation of water storage and hydropower projects affects the structure of water ecosystems of downstream rivers, and the establishment of ecological flow in rivers below the water storage and hydropower projects has significant impacts on maintaining the stability of river ecosystems. A database was established based on 2000–2017 environmental impact assessment (EIA) reports on water storage and hydropower projects in China and ecological flow (e-flow) methods, and the three widely used e-flow methods for water storage and hydropower projects in China were identified on the database. Furthermore, an ecological flow satisfaction degree model was used to evaluate the methods using long series of historical hydrological data from the hydrological stations in the Yellow River basin, the Yangtze River basin, and the Liao River basin. The results showed that hydrological methods are the type most often used for water storage and hydropower projects in China, including the Tennant method, the minimum monthly average flow with 90% guarantee rate method (Mm9M method), and the measured historical minimum daily average flow rate method (MDM method). However, the ecological flow methods selected are not significantly different among different basins, indicating that the selection of ecological flow methods is rather arbitrary, and adaptability analysis is not available. The results of the ecological flow satisfaction model showed that the Tennant method is not suitable for large river basins. The results of this study can provide technical support for establishment and management decisions surrounding ecological flow.


Introduction
A river ecosystem is the unity of material circulation and energy circulation between organisms and their environments and is a continuous system that relies on the hydrological regime [1,2]. The socio-economic development of human beings is closely related to rivers. To make more efficient use River basin [25,26]. In recent years, government management departments have grown increasingly concerned about ecological water demand. The national scientific and technological research projects in the Tenth Five Year Plan focused on the ecological restoration of the Liaohe River basin; based on the present conditions of the Liaohe River basin, the second-level water ecological subregion of the Liaohe River basin was constructed. In 2009, major national science and technology projects focused on the ecological environment restoration of the Weihe River basin. The national scientific and technological research projects in the Eleventh Five Year Plan completed and improved a number of different regions, different types and different periods, and completed an ecological water demand study at the river basin scale in China.
In general, recent studies have focused on constructing calculation methods for specific river segments. However, the concept and connotations of ecological flow have not been unified, the focuses of these calculation methods are not the same, and the scope of their application also varies. It is difficult to form a set of highly adaptive calculation systems.
The development of ecological flow for water storage and hydropower projects in China has been relatively slow. In the beginning, the definition of ecological flow was not unified, and there were many related concepts, such as the minimum flow and the in-stream flow. In addition, there were also no official recommended methods. Before 2000, environmental impact assessment (EIA) of water storage and hydropower projects did not mention the demand for ecological flow. Until 2006, the 'Technical Guide for Environmental Impact Assessment of River Ecological Flow, Cold Water, and Fish Passage Facilities for Water Conservation Construction Projects (Trial) EIA Letter (2006) No.4' was issued by the State Environmental Protection Administration, which defined the concepts and methods of ecological flow in EIA for the first time, recommending 10% of the average annual natural runoff as the minimum flow. However, this "Guideline" does not explicitly propose the appropriate calculation methods and the usage criteria for each method, and results from the adaptability assessment of the relevant ecological flow calculation methods are few [27].
This paper attempts to address the aforementioned issues by examination of existing EIA reports from 2000 to 2017 and determining the common methods used for ecological flow in China's water storage and hydropower projects. Moreover, we constructed a model to assess the degree of ecological flow satisfaction to emphasize the applicability of commonly used ecological flow methods in a water storage and hydropower projects in China. The research results can provide technical support for the establishment of ecological flow and for management decision-making, which is of great significance for protecting river ecosystem after damming.

EIA 2000-2017 Database
We constructed a database to restore the 206 EIA reports of water storage and hydropower projects during the period 2000-2017. The analyzed projects had passed the EIAs, which were downloaded from the EIA Center website.
The EIA is an administrative support procedure used by authorities to identify, describe, and assess the environmental impacts of a project. The results reflect the general opinions of the government, experts, and scholars. From the EIA, we can obtain the current status of ecological flow methods used in China's water storage and hydropower projects.
Ecological flow is not only a scientific issue but also a management issue. China's regulations did not mandate the rule of ecological flow from water storage and hydropower projects until 2006. Therefore, the analysis in this paper emphasized that 2006 was an important time node.

Hydrological Data
The Chinese Hydrological yearbook (The Red Book) is compiled by the hydrology department based on actual monitoring data. The data have passed through reliability, consistency, and representativeness tests, which were analyzed by hydrographic officers.
Considering the impact of water storage and hydropower projects and human activities on the current environment, the Red Book of each hydrological yearbook form 1950 to 1986 were extracted to analyze the adaptability of the ecological flow calculation method.

Assessment Model
There are more than 200 types of methods for ecological flow worldwide, and it is evidenced that results from different e-flow methods are different. Thus, it is important to choose an appropriate ecological flow method for determining the ecological flow of rivers. This paper uses the ecological flow satisfaction index to construct an evaluation model to initially evaluate the adaptation of the widely used ecological flow methods in China.
The ecological flow satisfaction degree is a commonly used method to assess the ecological water demand, which is usually the ratio of the series length of the actual runoff of the river course that is greater than the ecological water demand threshold to the total series length [28]. The assessment of the e-flow satisfaction degree usually uses the monthly average flow ecological water conservation rate to reflect the current river ecosystem water shortages; the assessment results can result in ecosystem water shortages [29,30].
where P i is the satisfaction degree of the ecological water requirement in the ith year, D i is the number of months that the ecological water requirement was met in the ith year, Q ij is the average flow at the jth month in the ith year, and E ij is the e-flow of the jth month in the ith year. From Equation (1), it can reflect the degree of the actual runoff meets the e-flow demand. However, Equation (1) cannot reflect a certain satisfaction degree of e-flow, especially in EIAs of water storage and hydropower projects in China. For example, if the e-flow is smaller than the measured minimum flow in the river, the result of Equation (1) is 100%, and this result is unreasonable. Moreover, if the e-flow is smaller than the minimum e-flow requirement from the "Guideline", which means E ij is less than 10% of the multiple average annual flow (MAF), the results of Equation (1) still do not reflect this phenomenon. Therefore, we add the above restrictions on the basis of Equation (1), and an e-flow satisfactoriness model is proposed for the preliminary evaluation of the adaptability of e-flow methods for water storage and hydropower projects in China.
where P i is the satisfaction degree of ecological water requirement in the ith year, D i is the number of months that the ecological water requirement met in the ith year, Q ij is the average flow at the jth month in the ith year, E ij is the e-flow of the jth month in the ith year, Q min is the smallest monthly mean flow measured in the ith year, and Q avg is the annual average flow for 36 years from 1950 to 1986.

Study Area
China has a vast territory and a relatively large geographical span. There is typically greater precipitation in southern China than in northern China, and the ecosystem structure of different regions is quite different, showing prominent seasonal and regional characteristics [30]. The choice of an appropriate calculation method for determining the different annual and seasonal e-flows in different regions of China is important for determining the spatial distribution of e-flow in China.
In this paper, we chose the 6 representative hydrological stations in the first level or second level basins in China to assess the widely used e-flow methods (Figure 1  regions is quite different, showing prominent seasonal and regional characteristics [30]. The choice of an appropriate calculation method for determining the different annual and seasonal e-flows in different regions of China is important for determining the spatial distribution of e-flow in China. In this paper, we chose the 6 representative hydrological stations in the first level or second level basins in China to assess the widely used e-flow methods (Figure 1), which are the Han River basin (Ankang Station, AK), Yangtze River basin (Pingshan Station, PS; Xiaodeshi Station, XDS; Kangding Station, KD), Yellow River basin (Anningdu Station, AND), Songhua River basin (Songhuajiang Station, SHJ).
In addition, the differences in the calculation methods among different river sections in the same basin were also analyzed. We selected four hydrological stations in different parts of the Han River basin to analyze the adaptability of the calculation method of e-flow in different parts of the Han River basin. Such as the Wuhou town (WHZ) station upstream of the mainstream, the Huangjiagang station (HJG) and Hongjiadu (HJD) station in the middle stream, and the Xincheng (XC) station downstream of the mainstream, as well as 5 stations in the tributaries, including the Zijingguan (ZJG) station at the Danjiang River, the Nanguangping (NKP) Station in the Jiahe River, the Tanghe (TH) Station in the Tang River, the Kaifengyu (KFY) Station in the Nanhe River, and the Chaiping (CP) Station in Xunhe River ( Figure 2).  In addition, the differences in the calculation methods among different river sections in the same basin were also analyzed. We selected four hydrological stations in different parts of the Han River basin to analyze the adaptability of the calculation method of e-flow in different parts of the Han River basin. Such as the Wuhou town (WHZ) station upstream of the mainstream, the Huangjiagang station (HJG) and Hongjiadu (HJD) station in the middle stream, and the Xincheng (XC) station downstream of the mainstream, as well as 5 stations in the tributaries, including the Zijingguan (ZJG) station at the Danjiang River, the Nanguangping (NKP) Station in the Jiahe River, the Tanghe (TH) Station in the Tang River, the Kaifengyu (KFY) Station in the Nanhe River, and the Chaiping (CP) Station in Xunhe River ( Figure 2).

Construction of Water Storage and Hydropower Projects
Among the 206 water storage and hydropower projects, 75 were constructed before 2006 (37 water storage projects and 38 hydropower projects), and 131 were built after 2006, of which 75 were water storage projects, whereas 56 were hydropower projects.
A total of 133 projects calculated the e-flow (59 were water storage projects and 74 were hydropower projects), and 73 projects did not calculate the e-flow (53 water storage projects and 20 hydropower projects) ( Table 1).
Among the 133 projects, 30 were constructed before 2006 (3 water storage projects and 27 hydropower projects), and 103 were built after 2006, of which 56 were water storage projects and 47 hydropower projects.

Construction of Water Storage and Hydropower Projects
Among the 206 water storage and hydropower projects, 75 were constructed before 2006 (37 water storage projects and 38 hydropower projects), and 131 were built after 2006, of which 75 were water storage projects, whereas 56 were hydropower projects.
A total of 133 projects calculated the e-flow (59 were water storage projects and 74 were hydropower projects), and 73 projects did not calculate the e-flow (53 water storage projects and 20 hydropower projects) ( Table 1). Among the 133 projects, 30 were constructed before 2006 (3 water storage projects and 27 hydropower projects), and 103 were built after 2006, of which 56 were water storage projects and 47 hydropower projects.

Use of Ecological Flow Methodology
Among the 133 water storage and hydropower projects discussed above, the hydrological method is the most widely used method, accounting for 171 uses. The hydraulic method and habitat simulation methods were both used approximately 20 times. The comprehensive analysis method was rarely used in China, and only one case was found in this study. Several water storage and hydropower projects consider the landscape water requirements, which occurred a total of 25 times (Figure 3a). Among the 133 water storage and hydropower projects discussed above, the hydrological method is the most widely used method, accounting for 171 uses. The hydraulic method and habitat simulation methods were both used approximately 20 times. The comprehensive analysis method was rarely used in China, and only one case was found in this study. Several water storage and hydropower projects consider the landscape water requirements, which occurred a total of 25 times (Figure 3a).
We also analyzed the conditions of e-flow methods in different watersheds. Hydrological methods were not used in the Hai River basin but were used in other basins. The hydraulic method was rarely used in China. It was used primarily in the south-eastern rivers (three times), the Yangtze River basin (nine times), the Songhua River, and the Pearl River basin (two times) and was not used in other basins. The use of the habitat simulation method is mainly in the Yangtze River basin (10 times), and the comprehensive analysis method was only used once in a river basin in north-western China and not in any other basins (Figure 3b). Among the ecological methods, we summarized the three most widely used e-flow methods, which are all hydrological methods, including the Tennant method, the minimum monthly average flow with a 90% guarantee rate method (Mm9M), and the minimum daily average flow method (MDM) ( Table 2).
The Tennant method is the most commonly used in case studies (84 times). The Tennant method was first reported by Tennant in 1976, who established the relationship among aquatic organisms, the river landscape, recreation and river discharge by analysing cross-sectional data of 11 rivers in the United States, suggesting that 10% of the MAF is the minimum e-flow. At present, the Tennant method is the most widespread hydrological method used worldwide and is the second most widely applied e-flow method in North America. It is also one of the recommended methods of the "Guideline".
The 7Q10 method is a hydrological method defined as low flow based on a return period of 10 years spanning seven consecutive days. The 7Q10 method uses hydrological parameters and water quality factors (river self-purification capacity) to calculate ecological water demand. We also analyzed the conditions of e-flow methods in different watersheds. Hydrological methods were not used in the Hai River basin but were used in other basins. The hydraulic method was rarely used in China. It was used primarily in the south-eastern rivers (three times), the Yangtze River basin (nine times), the Songhua River, and the Pearl River basin (two times) and was not used in other basins. The use of the habitat simulation method is mainly in the Yangtze River basin (10 times), and the comprehensive analysis method was only used once in a river basin in north-western China and not in any other basins (Figure 3b).
Among the ecological methods, we summarized the three most widely used e-flow methods, which are all hydrological methods, including the Tennant method, the minimum monthly average flow with a 90% guarantee rate method (Mm9M), and the minimum daily average flow method (MDM) ( Table 2).
The Tennant method is the most commonly used in case studies (84 times). The Tennant method was first reported by Tennant in 1976, who established the relationship among aquatic organisms, the river landscape, recreation and river discharge by analysing cross-sectional data of 11 rivers in the United States, suggesting that 10% of the MAF is the minimum e-flow. At present, the Tennant method is the most widespread hydrological method used worldwide and is the second most widely applied e-flow method in North America. It is also one of the recommended methods of the "Guideline".
The 7Q10 method is a hydrological method defined as low flow based on a return period of 10 years spanning seven consecutive days. The 7Q10 method uses hydrological parameters and water quality factors (river self-purification capacity) to calculate ecological water demand.
The Mm9M method is the second most frequently used method, which is improved from the 7Q10 method and has been used 40 times in different projects and selected basins. Due to the high requirements of the 7Q10 method for pollutant discharge, the "Technical Principles and Methods for Establishing Local Water Pollution Discharge Standards" (GB3839-1983) ruled that the minimum monthly average flow with a 90% guarantee rate should be used as the minimum e-flow [31].
The MDM method is the third most widely used method in EIAs of water storage and hydropower projects and was used 19 times in our case study. The MDM method is a hydrological method applied to watersheds in China, which is based on the Tennant method. It usually takes the minimum daily average flow measured over many years as the minimum e-flow. Note: Considering the difference between e-flow methods, multiple e-flow methods were occasionally used in a single project to determine the e-flow during the process of environmental impact assessment (EIA), which is the main reason that the number of e-flow methods exceeds the number of projects.

Ecological Flow Satisfaction Degree
We calculated the e-flow satisfaction degree of each method by the assessment model, which are the Tennant method, the MDM method, and the Mm9M method.
We first calculated the e-flow satisfaction degree of different methods at the control stations of each major basins. The results indicated the Tennant method was widely used to calculate the e-flow of water storage and hydropower projects in various basins, its ecological satisfaction degree at all of the stations were zero. In contrast, the MDM method performed well in all control stations and all of its e-flow satisfaction degrees were 100%. In contrast, the results of the Mm9M method varied, and the results at the six stations were only approximately 50% (Figure 4).
The adaptability of the ecological methods in different parts of one basin also varies. Based on the results from the nine control stations in Han River basin (Figure 5), the results showed that the e-flow satisfaction degrees calculated from the Mm9M method in Han River basin were generally not high, resulting in e-flow satisfaction degrees of 20-100% ( Figure 6).
The Tennant method performed differently between the mainstream river and the tributaries. The e-flow satisfaction degrees at the HJG, XC, and AK stations were zero, which are on the main channel of Han River basin. The results at the upstream stream control station WHZ was approximately 80%, and the tributary stations (ZJG, NKP, and TH) exhibited a value of 100%, although the ecological satisfaction degrees at the other tributary stations (KFY and CP) were zero.
Although the MDM method performed well, the results were also zero at ZJG, WHZ, and TH and 100% at the other stations.  Although the MDM method performed well, the results were also zero at ZJG, WHZ, and TH and 100% at the other stations.

Discussion
Determining the e-flow for water storage and hydropower projects is significant for river ecosystem. Therefore, the Ministry of Environmental Protection issued a "Guideline" in 2006, which clear stated the minimum standards for e-flows from water conservancy and hydropower projects. After more than ten years of development, we found that the "Guideline" played an essential role in promoting the discharge of e-flow, and the percentage of projects that disseminated e-flow increased from 30/75 to 103/131. However, there are still several problems in management, such as the lack of theoretical basis for choosing the calculation methods of e-flow, and lack of a rationality analysis used for e-flow methods.

Practice of Ecological Flow Methods
Ecosystems have well-defined regional characteristics [32], and several scholars have proposed dividing China into five major regions based on the present regional conditions [33,34].
Different ecological flow methodologies should be and are used in different areas and for different purposes, which range from global water resources planning to detailed planning for managed dam releases. The results of this paper show that the selection of methods within the study area does not differ much within China; e-flow in major river basins has been mainly calculated using hydrological methods, including the Tennant method, the Mm9M method, and the MDM method. Other methods, such as habitat simulation, hydraulics, and comprehensive analyses have been infrequently used; these results are similar to those of other papers [35,36]. This shows that the choice of e-flow methods in China is somewhat random. By comprehensively analyzing the aforementioned results, we believe that the main reasons for the current situation are as follows: 1.
China did not explicitly clarify the demand for e-flow in the EIA of water storage and hydropower projects before 2006, and the regulations did not expressly clarify the calculation method guidelines; therefore, the selection of the e-flow calculation methods in the EIA of water storage and hydropower projects relied on the subjective judgment by designers [37];

2.
The hydrological method is the most widely used in the world because of its simple use and low threshold. Furthermore, China's water storage and hydropower projects are mostly located in remote mountainous areas in the southwestern region [35,36,38] and the data in this area are mostly concentrated in river flow monitoring, while other basic data monitoring systems are lacking, which limits the application and popularization of different methods.
The hydraulic and habitat simulation methods have been commonly used in the Yangtze River basin and the Pearl River basin, where many water storage and hydropower projects have occurred. However, economic capacity is positively correlated with the abundance of primary hydrological data, and those basins represented one of the most developed regions in China, which can support the data demand of the various methods [39].
Relative to the hydraulic method, the habitat simulation method, which considers the biological and ecological water requirements, is relatively mature and has a solid theoretical basis. It is more suitable for the analysis of e-flow on the river scale, and it is currently thought to be able to quantify e-flow and consider ecological factors, making it a relatively credible assessment method [40,41]. However, this method has particular data requirements. The status quo of the current hydraulic fundamental data in China limits the extensive use of this method; in the future, China should vigorously strengthen hydraulic data to meet the data demand of various methods [42].
The comprehensive analysis method rarely appeared in our statistical results. This method studies the relationship among flow, sediment transport, riverbed shape, and riparian community based on expert advice from the entire ecosystem so that the recommended river flow can simultaneously satisfy multiple functions such as biological protection, habitat maintenance, sediment deposition, pollution control, and landscape maintenance. This method has a relatively high demand for data and emphasizes the river's natural properties. However, it is difficult to restore waterways to pristine conditions under the impact of a dam; thus, the adaptability of this method in the ecological restoration of post-dammed rivers is low [43].

3.
More studies on the resilience of e-flow methods for local areas are needed. There are few studies on the advantages and disadvantages of hydrological methods for small watersheds, which results in a lack of theoretical basis for the application of calculation methods; 4.
There is a lack of new methods or techniques for determining e-flow for different conditions. We believe that there should be specific methods for assessing e-flow problems, which distinguish spatial and temporal ecological problems or the planning purposes; 5.
E-flow conflicts with power generation efficiency; therefore, water storage and hydropower project managers usually chose the minimum standard specified in the "Guideline" to pass the environmental impact assessment. That is, the minimum value in the Tennant method and the rationality of the established results have not been deeply analyzed.
In the statistical process, we found that the French Rural method (defined as 5% of the AMF) was used eight times; this result surprised us because this contradicts the lower limit of the e-flow specified in the guidelines, which has no reference significance in the calculation of the e-flow of water storage and hydropower projects in China. However, this result indicates several problems in the computation and supervision of current e-flow calculations.

Assessment of Ecological Flow Methods
Hydrological methods are widely applied in China, yet the spatial and temporal variability of the e-flow criteria have been frequently questioned because these method usually have an absolute value for a certain year [44]. Using the normal fish spawning period (April to September) as an example, the hydrological regime of rivers below a dam calculated by these methods does not change, even during the period of fish spawning, resulting in a single hydrological regime and a single biological environment in the channel. This also explains the gradual shrinking of the population of Chinese sturgeon in the Yangtze River basin [45].
The Tennant method's e-flow satisfaction degree results are usually 0 in most main streams of rivers, which demonstrates that the calculated results are lower than the historical measured minimum flow and indicates that the Tennant method may not be suitable for water storage and hydropower projects in large river basins. This conclusion is inconsistent with the general belief that the Tennant method applies to rivers and perennial rivers [46,47]. The reason may be that the Tennant method is a well-known method in the United States and the world, and it may be more suitable for the actual situation of rivers in the United States and is not as adaptable to research in China. In addition, the management of e-flow is different in rivers and damming rivers according to the policy [48]. Therefore, an adaptation study of this method in downstream areas of China's water storage and hydropower projects should be carried out.
The MDM method is mainly based on the minimums of historical daily flow records, which are improved from the Tennant method for water storage and hydropower projects in China; this method performs well for large rivers but is not suitable for rivers with low flow, such as the tributary of the Han River basin. Rivers exhibiting the "cut-off" phenomenon, with minimum hydrological flow of 0, are most susceptible to errors when using this method. The Mm9M method's results are generally large, which indicate the ecological satisfaction rate is usually low in the mainstreams and tributary channels of rivers and that e-flow will cause waste of water resources. Thus, adaptive analytical research related to this method is sorely needed.
In this paper, we did not analyze the adaptability of various e-flow methods in different water use periods (time), nor did we consider the e-flow requirements of different organisms in the river. Therefore, the results in this study are only the initial screening of the adaptability of ecological flow calculations in China. In the future, a calculation framework based on time and space needs to be established to refine the calculation rules of e-flow and provide technical support for the ecological protection of river segments.

Conclusions
The calculation and establishment of e-flow in rivers affected by water storage and hydropower projects have significant impacts on maintaining the stability of river ecosystems. This paper collected 206 EIA reports for water storage and hydropower projects in China after the year 2000 and summarized the practice characteristics and application of e-flow methods in China and in first-level basin. Furthermore, the applicability of commonly used calculation methods in China was also discussed. The results of this study can provide technical support for establishment and management decisions surrounding ecological effects.
In summary, after implementing e-flow management, a variety of e-flow calculation methods are synthetically used, which greatly improve the application of e-flow measures in China. The most commonly used methods of e-flow are hydrological methods, including the Tennant method, the Mm9M method, and the MDM method.
Although e-flow management has played a positive role in the protection of river ecosystems, a series of problems remain, such as the blind calculation of e-flow, un-regionalization selection of e-flow methods, the unreasonable calculation of e-flow rates, and the lack of reasonable analysis of e-flow method. Therefore, in the future, the framework of e-flow management should determine the relationship regional ecological method model to optimize the e-flow method, and the ecological adaptability of the regulatory system; these changes will provide basic research support for the development of e-flow management policies.