Hydro-environmental response to the inter-basin water resource development in the middle and lower Han River, China

The construction of water conservancy projects in river basins is an important way for human beings to utilize water resources comprehensively. However, the increasing intensity of human activities imposed on the natural river system seriously threatens the river water environment (Bond et al. 2008; Pruden et al. 2012; Koster et al. 2021). In the past decades, the hydro-environmental problems caused by strong inter-basin water resource development have attracted much attention (Palmer 2010; Vörösmarty et al. 2010; Bakker 2012; O'Connor et al. 2015). Construction and operation of large-scale water conservancy projects generally have continuously physical and ecological influences on the downstream river channels. Liechti et al. (2015) analyzed the impact of water conservancy construction on the hydrological situation of the Zambezi River in Africa and southern China based on historical data and found that the reservoirs have a significant impact on river flow fluctuations and seasonal hydrological changes. Bonacci & Oskorus (2010) conducted a study on the hydrological situation of the Drava River in Italy for nearly 30 years and clarified the effects of the construction and operation of the reservoirs on the water level, flow, and sediment transport in the lower reaches of the river. More than half of the large rivers in the world are under the influence of strong human activities, these rivers are facing higher pressures of hydro-environmental changes than the ones under natural conditions (Gurría 2009; Yang et al. 2017).

Sustainable development and integrated utilization of water resources while considering river health have become an increasingly important issue in the current research areas of water conservancy and hydro-environment (Scionti et al. 2018; Hao et al. 2021). At present, the construction of water conservancy projects in China has gradually shifted from traditional water conservancy to ecological water conservancy, more and more attention was paid to the coordinated development of engineering and environment, and more importance was attached to maintaining the ecological environment and safeguarding humanity (Cheng et al. 2019). To improve the nationwide utilization of water resources in China, the Three Gorges Reservoir (TGR) and the South-to-North Water Transfer Projects (SNWTP) were constructed and put into service successively. The TGR and SNWTP, up to now, are the largest hydraulic structure and the inter-basin water transfer project all over the world. Both projects have greatly improved the utilization of water resources and alleviated the water shortage in northern China, which contributed a great deal to the optimal water resources allocation and economic development (Gao et al. 2016; Rogers et al. 2020).

However, such large-scale water conservancy projects also brought a series of hydro-environmental impacts on the downstream river channels (Zhou 2020; Sayed et al. 2021). As the largest tributary of the Yangtze River, the middle and lower Han River suffered from the hydrological impact resulting from the TGR and SNWTP. The middle and lower Han River has been seriously polluted for six consecutive years after the middle route of SNWTP operation, which resulted in the frequent occurrence of river eutrophication (Jing et al. 2019). Diatom blooms frequently broken out after the SNWTP operation despite adopting some mitigation measures, e.g., the Yangtze-Han Water Diversion Project (YHWDP) (Zhu et al. 2008). The current river regulation problem urgently needs to be solved is how to alleviate the adverse influences resulting from the TGR and the middle route of SNWTP in the lower Han River and maintain hydro-environmental health. In recent years, although the effect of the middle route of SNWTP on the middle and lower Han River was studied from different perspectives, little quantitative research was conducted to examine the TGR impact on the hydro-environmental processes of downstream tributaries (Wilson et al. 2017; Zhou et al. 2017; Zhang et al. 2018; Du et al. 2019; Zhuang et al. 2019). Besides, the former research on the middle route of SNWTP impact on the hydro-environment of the middle and lower Han River is mainly based on assuming or planning conditions because the middle route of SNWTP was just put into operation since 2014 (Zhu et al. 2008; Jing et al. 2019). However, the actual hydrological processes of the Han River are bound to be different from the planning situations. The numerous environmental factors' variation characteristics and response mechanisms in the middle and lower Han River before and after the scientific identification of water diversion need to be further studied.

This study focused on the hydro-environmental response of the middle and lower Han River to the TGR and the middle route of the SNWTP operation. The research aimed to clarify two key questions: (1) What changes have occurred in hydrology, water quality in the middle and lower Han River before and after the TGR, and the middle route of SNWTP operation? (2) How to identify and quantify the different influences that resulted from both projects, which provides a basis for revealing the causes of the frequent blooms in the lower Han River. Based on a hydro-environmental model and the statistical methods, hydrological changes and the water quality variations under the flow regulation of the TGR and the middle route of SNWTP operation were examined, and quantitative evaluation was conducted in this study. The scientific identification of the formation and evolution mechanism of the hydro-environmental changes in the middle and lower Han River was carried out, which provided an important scientific basis for maintaining the sustainable management of water resources and water environmental health in the Han River basin.

The stretch of the Han River falls between 106°∼114°E, 30°∼34°N, the drainage area is 15.1 × 10⁴ km², and the total length of the mainstream is 1,577 km. As shown in Figure 1, Danjiangkou and Zhongxiang divide the mainstream into the upper, middle, and lower reaches. The middle and lower Han River is about 580 km long and flows through Laohekou, Xiangyang, Zhongxiang, Qianjiang, Xiantao, Hanchuan, and Wuhan cities. As the largest tributary of the Yangtze River, the Han River connects its channels with the Yangtze River in Wuhan city. The climate in the Han River basin is mild and humid, and the rainfall is abundant. The multi-year average temperature is about 15∼17 °C.

There are four dams (Danjiangkou, Wangfuzhou, Cuijiaying, and Xinglong) along the middle and lower Han River, constituting the cascade reservoirs system. The dams are typical low-head run-of-river hydroelectric plants except for the Danjiangkou dam. The middle route of SNWTP is an enormous inter-basin water transfer project, which significantly contributed to the optimization of China's water resource allocation. The middle route of SNWTP draws water from the Danjiangkou Reservoir (DJKR) in the upper reaches of the Han River. Since its gate of the middle route of SNWTP officially opened on December 12, 2014, it has transferred water up to 2.55 × 10¹⁰ m³ and greatly alleviated the water shortage in northern China.

The TGR is located 603 km upstream of the Han River estuary (Figure 1), which was put into operation in 2003. Significant hydrological changes occurred after the TGR operation during its flow regulation periods, which has an appreciable impact on the downstream flow processes and stage variation along the channels. The variations in the water level at the conference area of the two rivers also affect the outflow from the Han River. Therefore, this study focused on the coupled effects of TGR and the middle route of SNWTP on the hydro-environment along the lower Han River.

The research data include topographical, hydrological, meteorological, and water quality datasets. The digital elevation model data was provided by the Bureau of Hydrology, Changjiang Water Resources Commission. The hydrological data were obtained from the hydrological yearbook of the Institute of Geographical Sciences and Natural Resources Research of the Chinese Academy of Sciences. The daily monitoring data were collected, including the inflow and outflow of the Danjiangkou Reservoir, series of gauged flow and stage at the Huangjiagang, Xiangyang, Huangzhuang, Xiantao, and Hanchuan hydrological stations in the middle and lower Han River and the Hankou station in the Yangtze River. The hydrological data cover the periods from 1996 to 2019.

The meteorological data is downloaded from the China Meteorological Data Network in the middle and lower Han River, including the gauged data at the Laohekou, Zaoyang, Zhongxiang, Tianmen, and Wuhan stations (Figure 1). The time series of meteorological data are consistent with the hydrological data. Besides, the monthly averaged water quality data from 1996 to 2019 at the Shilou, Yujiahu, and Zongguan stations were provided by the Water Environment Monitoring Center of the Yangtze River Basin. The monitoring indicators at these stations include water temperature, ammonia nitrogen (NH₃-N), total phosphorus (TP), biochemical oxygen demand (BOD₅), and pH. Besides, the locations and the waste flow data of the sewage outfalls and water intakes along the middle and lower Han River were collected from the Hubei Environmental Protection Bureau. The data of socioeconomic indicators used in this study was collected from the Hubei Statistical Yearbooks. All the collected data are consistent, and there is no missing data in the time series.

In this study, a hydro-environmental model was set up for the middle and lower Han River based on the integrated simulation modules developed by the Danish Hydraulic Institute. The framework and modules of the model are shown in Figure 2. The coupled model includes hydrodynamic module and environmental module, and the governing equations are shown as follows.

In this study, flow and mass transport boundary conditions were specified in this hydro-environmental model. The inflow of the Danjiangkou reservoir was the upstream flow boundary condition for the coupled model, and the stage-discharge relation curve at the Hankou station was specified as the downstream model condition. The lateral inflow and outflow of branches and the agricultural and industrial water demand were calculated as the source or sink terms of the continuity equation. Moreover, the cascade reservoirs downstream of the Danjiangkou Reservoir are low-head run-of-river hydropower plants, and their scheduling rules are not complex. The reservoir stage will be kept at the normal pool level when the upstream inflow is less than the designed flow.

On the contrary, the reservoir will discharge according to its discharge capacity curve when the inflow is larger than the designed flow. In this study, the dispatching rules of cascade reservoirs are adopted during the hydro-environmental model setup. Besides, the load boundary conditions were specified at the sewage outfall and water intakes based on the collected data. The non-point loads from the subbasins were estimated using the tool of Mike Load Calculator, which was also specified in the coupled model (Zhu et al. 2008). The detailed information on the boundary conditions for flow and pollution load along the middle and lower Han River were presented in the Supplementary Material (Text S1).

The middle and lower channel of the Han River was divided into four reaches in the coupled model by three reservoirs (Wangfuzhou, Cuijiaying, and Xinglong) with 316 sections. The largest section distance is 3.51 km, and the average distance is 1.83 km. The hydrological calibration of the model was conducted using the hydrological processes from April 1, 2013 to December 31, 2015, and its validation was carried out by the gauged hydrological data from January 1, 2016 to December 31, 2017. As shown in Supplementary Material, Table S1, the calibrated Manning's roughness coefficient (n) ranged from 0.018 to 0.053. The calibration and validation results are shown in Supplementary Material, Figure S1. The simulation results were evaluated using the Nash-Sutcliffe efficiency coefficients (NSE), and the normalized root mean squared error (NRMSE) (Zhang et al. 2018). The NSE at the gauging stations is larger than 0.92, and RMSE is smaller than 0.16, implying the coupled model can be applied to predict and evaluate the hydrological changes accurately under the influences of inter-basin water resource development.

The parameters for the water quality simulation of the coupled model were also calibrated and validated based on the observed data in 2006 and 2007. The results indicate that the longitudinal dispersion coefficient k ranges from 0.035 to 0.050 m²/s. The detailed information for the model calibration results is shown in Table S1. Supplementary Material, Figure S2 presents the comparison results of water quality indicators between the observed and the simulated data. The statistical comparison results indicate that the average relative errors of NH₃-N, TP, and BOD₅ are 23.5%, 20.1%, and 17.9%, respectively. Therefore, the coupled model can be used to simulate the actual hydro-environmental processes within the acceptable error range.

Based on the 1D hydro-environmental model and the collected data, as shown in Table 1, three simulation scenarios for the hydro-environmental evaluation of the middle and lower Han River were conducted in this study. Scenario 1 was applied to simulate the hydro-environmental processes without the water transfer projects (SNWTP, YHWTP) and TGR. Scenario 2 was used to reveal the backwater effect in the lower Han River. Moreover, Scenario 3 was applied to examine the coupled influences that resulted from the water transfer projects and TGR.

Figure 3 presents the flow duration curve at the Huangjiagang station. If the flow of less than 800 m³/s was considered the low flow, the low flow duration significantly increased after water diversion, which extended from 3.2–5.1 months to 7.0–10.1 months. As shown in Table 2, the average annual flow at the Xiantao station increased from 843.67 m³/s to 874.62 m³/s after the TGR operation, while it decreased to 775.31 m³/s after water diversion. In particular, there was much more distinct flow alteration during the flood seasons than the dry seasons. It should be noted that the multi-year average flow of scenario 3 decreased both in flood seasons and dry seasons comparing to Scenarios 1 & 2. Spearman correlation analysis was carried out based on the monthly flow of the three scenarios. The test results showed that the was less than the critical value in Scenarios 1 & 2, which indicated no obvious changing trend in flow series during the two periods. In particular, although there were evident flow variations during the dry seasons and flood seasons after the TGR and SNWTP operation, the Spearman test results did not show a distinct changing trend within the statistical periods of the three scenarios.

The calculation results of flow distribution before and after the water diversion are shown in Table 3. There was a significant increase in the coefficient of variation and flow variation range of Scenario 3, which indicated the much more uneven flow distribution occurred after the water diversion. Compared with Scenario 1, there was no significant increase in the hydrologic alteration indicators of Scenario 2, which indicated that the TGR operation exerted slight influences on the flow distribution in the lower Han River.

Moreover, the water level at the estuary of the Han River is the vital outflow boundary condition, which greatly contributed to the outflow processes variation of the lower Han River. The stage variation at the Hankou station can reflect the estuary outflow boundary changes, which are presented in Supplementary Material, Figure S3. There was a significant variation in the water level at the Hankou station after the TGR operation. The monthly average water level decreased obviously from April to December, especially stage decrease during October and November because of TGR impoundment. The maximum stage decrease was up to 2.34 m. On the contrary, the water level increased slightly during the water compensation from the TGR (from January to March), and the average stage increase was 0.78 m.

Based on the simulation results of the hydro-environmental model, the changes in water quality indicators, NH₃-N, TP, and BOD₅, were presented in Figure 4. The environmental processes at the Yujiahu, Shilou, and Zongguan stations reflected the spatiotemporal variations in water quality along the Han River channels. As shown in Figure 4, there was a distinctly rising trend of Scenario 3 in NH₃-N concentration at the Yujiahu, Shilou, and Zongguan stations compared with Scenarios 1 and 2. The largest monthly average concentration increase at the Yujiahu station was up to 0.109 mg/l. Compared with Scenario 1, the concentration variations in the NH₃-N of Scenario 2 presented a different trend during the dry and flood seasons. In Scenario 2, there was a slight increase in the NH₃-N concentration during the dry seasons (December to the following February), while it decreased to a different extent at the gauging stations during the flood seasons. The maximum NH₃-N concentration decrease occurred at the Zongguan station by 0.06 mg/l during August. In terms of spatial variation along the channels, the closer to the estuary of the Han River, the more significant concentration variations occurred at the gauging stations. The monthly average NH₃-N concentration increase of Scenario 2 was 0.048 mg/l at the Zongguan station than that of Scenario 1, which increased by 12.6% in contrast to that at the Yujiahu station.

There were similar changing trends of TP, BOD₅ at the gauging stations. In contrast to Scenario 1, the TP and BOD₅ concentrations variated to a different extent in Scenarios 2 & 3. In particular, there was a significant concentration increase in TP and BOD₅ of Scenario 3. Compared with Scenario 1, the largest concentration increases in TP and BOD₅ of Scenario 3 at the gauging stations were 0.04 mg/l and 0.15 mg/l, respectively. In addition, the monthly average concentration of TP and BOD₅ of Scenario 2 decreased by 0.005 mg/l and 0.089 mg/l than the ones of Scenario 1 during the flood seasons. On the contrary, the monthly average TP and BOD₅ concentration of Scenario 2 increased by 0.002 mg/l and 0.037 mg/l than the ones of Scenario 1 during the dry seasons.

Supplementary Material, Figure S4 shows the changing trend of the integrated evaluation index along the middle and lower Han River. There was an obvious increase in the integrated evaluation index of Scenario 3 at the gauging stations. After the TGR and SNWTP were implemented, the integrated evaluation index variated to different degrees at the Yujiahu, Shilou, and Zongguan stations. In particular, there was not a good consistency in the variations of integrated evaluation index P in Scenario 2. The largest P-decrease of Scenario 2 was 0.021, 0.035, and 0.041 than Scenario 1 during the periods from March to November, while the largest P-increase were 0.018, 0.020, and 0.022 during the dry seasons (December to the following February). Compared to the increase of integrated evaluation index P between the Shilou and Zongguan stations, the more obvious P-increase occurred at the estuarine Zongguan station of the Han River. It should be noted that the integrated evaluation index was still less than 1.0, which indicated that it was not larger than the highest allowable value along the middle and lower Han River.

The integrated evaluation index was examined by the Spearman test, and the results are shown in Table 4. The critical values of the Spearman test in scenarios 1–3 at the significance level of 0.05 are 0.786, 0.587, and 0.900, respectively. The integrated evaluation index at the gauging stations is less than the critical values. Therefore, there was no obvious increase or decrease trend in the integrated evaluation index at the gauging stations. Moreover, the SNWTP significantly influenced pollutants duration along the middle and lower channels when it was put into operation (Figure 5). In particular, the NH₃-N concentration of Scenario 3 at the stations was larger than 0.30 mg/l during more than 50% of the statistical time. Also, the changing processes of TP and BOD₅ had a similar trend with that of NH₃-N. The duration of higher pollutant concentration was drastically extended in Scenario 3.

The hydrological changes resulted in the variations in water quality indicators and influenced the nutrient assimilative capacity along the middle and lower Han River. The spatiotemporal variations in nutrient assimilative capacity are shown in Figure 6. Although the water quality indicators fluctuated, the decreasing trend was evident after the SNWTP operation. The more runoff flowed through the channels, the larger the nutrient assimilative capacity there was along the Han River, e.g., the nutrient assimilative capacity increased significantly during the rainy periods from 2011 to 2013.

The average allowable capacity of NH₃-N for the middle and lower Han River was 853.85 g/s, 870.69 g/s, and 586.27 g/s in Scenarios 1, 2, and 3, respectively. As shown in Figure 6, the average allowable capacity of NH₃-N was much larger than that of TP. The TP was evident out of the limit downstream of the Xiantao city. In particular, the closer to the estuary of the Han River, the larger decrease occurred in nutrient assimilative capacity. Also, there was a similar spatiotemporal variation in the BOD₅ along the lower Han River. With the increase of pollutants, it needed more oxygen to decompose the organic matter. Therefore, the available capacity for the BOD₅ along the middle and lower Han River decreased significantly.

As shown in the results, the water environment of the Han River is deteriorating under the influences of inter-basin water resource development. The hydrological changes along the middle and lower Han River significantly contributed to the water environmental variations. The runoff decreased substantially after the SNWTP operation, directly increasing pollutant concentration and rapidly decreasing nutrient assimilative capacity (Figures 4–6).

The flow decrease in the Han River and the backwater effect of the water surface of the Yangtze River increased the estuarine water level of the Han River, which led to the hydraulic gradient decrease and velocity reduction along the lower Han River to a different extent. The monthly averaged flow velocity at the Yujiahu, Shilou, and Zongguan stations was presented in Supplementary Material, Figure S5. The flow velocity during the dry seasons decreased slightly after the TGR and SNWTP operation, and the flow velocity larger than 1.5 m/s rarely occurred along the lower channels. In particular, the monthly average flow velocity decreased by 0.1 m/s during the dry seasons under the coupled influences of the TGR and SNWTP (Supplementary Material, Figure S5). Due to the evident runoff reduction and flow velocity decrease, the pollutant dispersion becomes much slower. Therefore, water self-purification capacity decreased and the pollutants accumulated in the middle and lower Han River.

However, there was no consistency in the flow variations all year because of the different floods encountered between the two rivers. Compared with Scenario 1, the TGR indeed contributed to a certain degree to the estuarine level decrease and flow velocity increase along the lower Han River during the TGR impounding periods (September and October). There was a flow velocity increase by 0.13–0.52 m³/s in Scenario 2 during the TGR impounding periods. However, the flow velocity increase during the TGR impoundment was still unable to improve the water environment of the middle and lower Han River. Because the runoff decrease of the middle and lower Han River drastically reduced the water level after the SNWTP operation, there was no extremely significant flow velocity increase at all. In particular, the TGR impoundment impact lasts no more than 2.5 months annually. Although the flow in the Han River increased after the YHWDP was put into operation, the low flow duration had no remarkable changes (Zhang et al. 2018). Therefore, the YHWDP can improve the water quality after its operation but will not completely offset the adverse influences resulting from the SNWTP. The runoff decrease is the primary cause of water quality deterioration along the middle and lower Han River.

The concentration increase in NH₃-N, TP, and BOD₅ deteriorated the water quality in the lower Han River, which significantly reduced the nutrient assimilative capacity. Consequently, several algal bloom events occurred in recent years, which posed a great threat to water safety in the Han River basin (Xin et al. 2020). To deal with the water quality degradation and the series of negative influences resulted from the project operation on the hydro-environment, the outflow into the middle and lower channels of Han River should be increased to offset the flow decrease after the SNWTP operation. Moreover, pollutant emissions should be strictly limited, and wastewater treatment should also be taken seriously in the Han River basin. Last but not least, optimizing water resource scheduling of the inter-basin water resource may also be an effective measure to deal with the deterioration of water quality under the limited water quantity.

The SNWTP and TGR, as the important hydraulic engineering systems to optimize water distribution across watersheds, also had significant adverse effects on downstream river health. In this study, the hydro-environmental modeling and the middle and lower Han River data analysis were applied to evaluate the hydrological changes and water-quality variations resulting from the inter-basin water resource regulation. The lower Han River is under the combined effects of the hydrological variations in the Han River and the Yangtze River. After the SNWTP operation, the low flow duration was extended from 3.2–5.1 months to 7.0–10.1 months.

On the other hand, the water level of the Han River estuary, as one of the most important outflow boundary conditions, had remarkable variations after the TGR operation. There was an obvious stage decrease during the TGR impoundment, while the estuarine stage of the Han River increased slightly when the TGR began to supplement water to the downstream channels. The water quality of the middle and lower Han River deteriorated under the influences of the inter-basin water resource regulation of the SNWTP. The concentration increase in NH₃-N, TP, and the BOD₅ resulted in an obvious increase in the integrated evaluation index at the gauging stations. Consequently, the allowable nutrient assimilative capacity decreased significantly. Effective measures should be taken to improve the hydro-environment along the middle and lower Han River to reduce pollutant emission and optimize the reservoir operation.

This work was supported by the National Key Research and Development Plan (2016YFC0400901) and the National Natural Science Foundation of China (51509273).

The authors have declared no conflict of interest.

All relevant data are included in the paper or its Supplementary Information.