Water Compensation and Its Implication of the Three Gorges Reservoir for the River-Lake System in the Middle Yangtze River, China

water

Article Water Compensation and Its Implication of the Three Gorges Reservoir for the River-Lake System in the Middle Yangtze River, China

Junhong Zhang 1,2, Luojie Feng 1, Lu Chen 3,*, Dangwei Wang 4, Minglong Dai 5, Wensheng Xu 6 and Tao Yan 7

1 College of Resources and Environmental Science, South-Central University for Nationalities, Wuhan 430074, China; [email protected] (J.Z.); [email protected] (L.F.) 2 Department of Civil & Environmental Engineering, University of Alberta, Edmonton, AB T6G 2W2, Canada 3 College of Hydropower & Information Engineering, Huazhong University of Science & Technology, Wuhan 430074, China 4 China Institute of Water Resources and Hydropower Research, Beijing 100038, China; [email protected] 5 Bureau of Hydrology, Changjiang Water Resources Commission, Wuhan 430010, China; [email protected] 6 Changjiang River Scientiﬁc Research Institute, Wuhan 430010, China; [email protected] 7 Tianjin Research Institute for Water Transport Engineering, Tianjin 300456, China; [email protected] * Correspondence: [email protected]; Tel.: +86-189-7167-6115

Received: 29 May 2018; Accepted: 25 July 2018; Published: 31 July 2018

Abstract: Dam construction is an important means to improve water use efficiency and the aquatic environment. However, the flow regulation of the Three Gorges Reservoir (TGR) in the middle Yangtze River has attracted much attention because the severe drought occurred in the river-lake system downstream of the TGR. In this paper, the Dongting Lake was selected as a case study in order to detect the possible relationship between the flow regulation of the TGR and the extreme drought in the river-lake system based on a coupled hydrodynamic model. The results not only confirmed the significant role of the TGR to relieve drought in the river-lake system, but also indicated that the outflow of the TGR and the hydraulic gradient between the Zhicheng to Chenglingji stations were the crucial factors to affect the water exchange between the rivers and the Dongting Lake. The adjustment of hydraulic gradient within a proper range during the water compensation of the TGR will be an effective measure to improve the water exchange and water environment in the river-lake system. These findings present the quantitative influence of these important factors on the water exchange between rivers and lakes and provide a scientific reference for environmental and ecological management of other river-lake systems.

Keywords: water compensation; the Three Gorges Reservoir; river-lake system; water exchange

1. Introduction With the rapid development of social economy, the exploitation and utilization of water resources of rivers and lakes have been highly concerned around the world [1–3]. In particular, dams have been playing an increasingly important role in the integrated utilization of water resources. The Itaipu dam in the Parana River provides 17.4% and 74.1% electrical energy for Brazil and Paraguay, respectively [4]. The Aswan dam in the Nile River not only provides ﬂood protection and electrical energy, but it also supplies adequate water for irrigation [5]. The vast Mead Lake was formed after the construction of Hoover dam in the Colorado River, which has been an important habitat for animals [6]. However, the widespread dams in the world also caused some trouble in the environmental management. Subsequent changes in the timing and duration of streamﬂow were involved when dams were put

Water 2018, 10, 1011; doi:10.3390/w10081011 www.mdpi.com/journal/water Water 2018, 10, 1011 2 of 17 into operation [7–9]. Consequently, the original ecological balance was partially or fully disrupted in the river basins [10–13]. The Yangtze River, which is located in the central China, is not excluded from the influence of water resources development. The impoundment impact of the Three Gorges reservoir on hydrology, climate, environment, and ecology has attracted much attention [14,15]. There are obvious hydrological regime changes in the Yangtze River after the operation of the Three Gorges Reservoir (TGR), which have an extensive influence on the health of the river-lake system [16,17]. Water level drop occurs downstream the TGR due to the severe riverbed erosion [18]. Therefore, the water exchange between the rivers and the lakes in the middle Yangtze River is disturbed, which greatly affects the wetland environment that is the main habitat of aquatic plants and animals [19]. The Yangtze River is divided into channel segments by a cascade reservoir system because of the development of water resources, which generally leads to a decrease in biodiversity [20]. In addition, the longer duration of beach exposure in the river-lake system after the operation of the TGR greatly deteriorated the ecological environment [21]. Drought occurred more frequently and its duration increased by about 30% after the operation of the TGR [22]. Therefore, it is undeniable that the environmental impact is remarkable during the impoundment periods. Although the TGR impoundment contributes a lot to the drought in the river-lake system, its contribution to the drought as the main factor remains contentious [23–25]. The flow discharge generally increases during the extreme dry seasons from December to the following April, because the water compensation from the TGR is one of the important rules of flow regulation [26]. However, whether the water compensation during the dry seasons can ease the extreme drought of the river-lake system in the middle Yangtze River has not drawn definitive conclusions. Most of the current research is focused on the TGR impoundment and its implication, little work has been performed on the influences of water compensation after the TGR operation on the river-lake system, let alone the research on its mechanism of the water exchange between the mainstream and the lakes [27,28]. Therefore, this paper aims to analyze the real effect of the water compensation from the TGR on the extreme drought in the middle Yangtze River. The Dongting river-lake system was selected as a case study to reveal the impact of water compensation from the TGR based on a one-dimensional/two-dimensional (1D/2D) coupled hydraulic model. Whether the flow regulation of the TGR exacerbated or relieved the extreme drought during the dry seasons was given a definite answer in this study. The variation in water level, flow discharge, and the water exchange between the Dongting Lake and the mainstream was calculated and analyzed. In addition, the influence factors and dynamic mechanism were further discussed. The hydraulic gradient and the outflow from the TGR, as two important factors, were proposed as a first attempt to regulate and control the extreme drought during the dry seasons. The research results can provide some technical support and reference for the administrative departments of water conservancy.

2. Materials and Methods

2.1. Study Area and Data Collection Dongting Lake, which is located in the middle Yangtze River (111◦190–113◦340 E, 27◦390–29◦510 N), is one of the famous freshwater lakes in the world (Figure1). The lake area is approximately 2579 km 2, accounting for 0.143% of the entire Yangtze basin. The river network around the lake area is the important hydraulic connection between the Dongting Lake and the Yangtze River. Flow diversion occurs from the mainstream of the Yangtze River to the Dongting Lake through the tributaries distributed between the both. There are three ﬂow inlets along the mainstream, as shown in Figure1, Songzi, Taiping and Ouchi. In addition, four rivers (Li River, Yuan River, Zi River, and Xiang River) pour their water into the Dongting Lake. The ﬂow outlet of the Dongting Lake, Chenglingji, is located at the lowest site of the lake area. All of the rivers and the lake constitute a complex river-lake system in the middle Yangtze River. There are some gauging stations in the study area, which monitored the Water 2018, 10, x FOR PEER REVIEW 3 of 17 Water 2018, 10, 1011 3 of 17 river‐lake system in the middle Yangtze River. There are some gauging stations in the study area, whichwater levelmonitored and discharge the water processes level and at discharge the key channel processes sites at (Figure the key1). channel The detailed sites information(Figure 1). The for detailedthese stations information is listed for in these Table 1stations. is listed in Table 1.

FigureFigure 1. 1. LocationLocation of of the the river river-lake‐lake system system in in the the middle middle Yangtze Yangtze River. River.

TableTable 1. 1. TheThe detailed detailed information information of of gauging gauging stations stations in in the the river river-lake‐lake system. Location Series No. Station Name Location Catchment Area (km2) Parameters Gauged Series No. Station Name Longitude Latitude Catchment Area (km2) Parameters Gauged 1 Yichang 111°17Longitude′ E 30°42 Latitude′ N 1,005,501.00 2 1Zhicheng Yichang 111°30 111◦′ 17E 0 E30°30 30◦42′ 0NN 1,005,501.001,024,131.00 ◦ 0 ◦ 0 3 2Shashi Zhicheng 112°13 111 ′ 30E E30°18 30 30′ NN 1,024,131.001,028,415.00 ◦ 0 ◦ 0 4 3Jianli Shashi 112°53 112 ′ 13E E29°49 30 18′ NN 1,028,415.001,033,274.00 4 Jianli 112◦530 E 29◦490 N 1,033,274.00 5 Chenglingji 113°08′ E 29°25′ N 262,344.00 water level, discharge 5 Chenglingji 113◦080 E 29◦250 N 262,344.00 water level, discharge 6 6Luoshan Luoshan 113°22 113◦′ 22E 0 E29°40 29◦40′ 0NN 1,294,911.001,294,911.00 7 7Taojiang Taojiang 112°06 112◦′ 06E 0 E28°55 28◦55′ 0NN 25,788.0025,788.00 8 8Taoyuan Taoyuan 111°29 111◦′ 29E 0 E28°54 28◦54′ 0NN 90,530.0090,530.00 ◦ 0 ◦ 0 9 9Shimen Shimen 111°23 111 ′ 23E E29°37 29 37′ NN 17,942.0017,942.00 10 10Nanzui Nanzui 112°18 112◦′ 18E 0 E29°03 29◦03′ 0NN- ‐ 11 11Heyehu Heyehu 112°54 112◦′ 54E 0 E28°52 28◦52′N0N- ‐ 12 Shawan 112◦180 E 28◦550 N- water level 12 Shawan 112°18◦′ E 0 28°55◦ ′ 0N ‐ water level 13 13Lujiao Lujiao 113°07 113 ′ 07E E29°08 29 08′ NN- ‐ 14 Yingtian 112◦550 E 28◦510 N- 14 Yingtian 112°55′ E 28°51′ N ‐ ◦ 0 ◦ 0 15 15Xinjiangkou Xinjiangkou 111°47 111 ′ 47E E30°18 30 18′ NN- ‐ 16 Shadaoguan 111◦550 E 30◦180 N- 16 Shadaoguan 111°55′ E 30°18′ N ‐ 17 Mituosi 112◦070 E 30◦220 N- water level, discharge 17 18Mituosi Kangjiagang 112°07 112◦′ 18E 0 E30°22 29◦73′ 0NN- ‐ water level, discharge 18 19Kangjiagang Guanjiapu 112°18 112◦′ 19E 0 E29°73 29◦73′ 0NN- ‐ 19 Guanjiapu 112°19′ E 29°73′ N ‐

WaterWater 20182018, 10, 10, x, 1011FOR PEER REVIEW 4 4of of 17 17

The TGR is located at approximately 45 km upstream of the Yichang hydrological station. Its lengthThe and TGR height is are located 2309 atm and approximately 185 m, respectively. 45 km upstream According of to the the Yichang operation hydrological rules of the station.TGR, thereIts length are four and heightcontinuous are 2309 periods m and of 185flow m, regulation respectively. within According a complete to the run operation cycle: rules(1) flood of the preTGR,‐discharged there are control, four continuous the water periodsin the reservoir of flow regulation is emptied within to the adownstream complete run before cycle: the (1) main flood floodpre-discharged seasons [28,29]; control, (2) theflood water regulation, in the reservoir the upstream is emptied floods to are the regulated downstream and before then discharged the main flood to theseasons downstream [28,29]; during (2) flood the regulation, periods from the upstreamJuly to August floods [30]; are (3) regulated reservoir and impoundment, then discharged the toTGR the beginsdownstream to store during water the during periods the from approximate July to August periods [30 ];from (3) reservoir September impoundment, to October; theand TGR (4) water begins compensation,to store water duringthe outflow the approximate from the periodsTGR is fromincreased September in order to October; to meet and the (4) downstream water compensation, water demandthe outflow during from the the dry TGR seasons. is increased The water in order compensation to meet the generally downstream occurred water from demand December during theto the dry followingseasons. TheApril water to meet compensation the industrial generally and agricultural occurred from needs December and maintain to the the following water depth April toin meetthe downstreamthe industrial waterways. and agricultural The TGR needs was and put maintaininto operation the water in 2003 depth and in its the impoundment downstream test waterways. lasted fiveThe years. TGR wasImpoundment put into operation test of the in 2003 normal and pool its impoundment level was conducted test lasted in five 2008, years. and Impoundment the normal flow test regulationof the normal of the pool TGR level began was conductedsince 2009. in The 2008, inflow and theand normal outflow flow processes regulation of the of the TGR TGR from began 2008 since to 20152009. are The shown inflow in andFigure outflow 2. There processes were the of thesimilar TGR regulated from 2008 flow to 2015 progresses are shown because in Figure of the2. Theresame strictwere scheduling the similar rules regulated of the flow TGR. progresses because of the same strict scheduling rules of the TGR.

Figure 2. Inﬂow and outﬂow process of the Three Gorges Reservoir (TGR) from 2008 to 2015. Figure 2. Inflow and outflow process of the Three Gorges Reservoir (TGR) from 2008 to 2015.

InIn addition, addition, the the topographic andand hydrologicalhydrological data data for for the the modeling modeling research research were were collected collected from fromthe Bureauthe Bureau of Hydrology, of Hydrology, Changjiang Changjiang Water Water Resources Resources Commission, Commission, China. China. The The hydrological hydrological data dataseries series cover cover the water the water level andlevel discharge and discharge gauged gauged at the stations at the instations the study in area.the study The topographic area. The topographicmaps of the maps river-lake of the system river‐ werelake measuredsystem were from measured 2008 to 2012. from Moreover, 2008 to 2012. the inﬂow Moreover, and outﬂow the inflow data andafter outflow the operation data after of the the TGR operation from 2008 of tothe 2015 TGR were from collected 2008 to from 2015 the were Three collected Gorges from Corporation the Three [31 ]. Gorges Corporation [31]. 2.2. Methodologies 2.2. Methodologies 2.2.1. Coupled Hydrodynamic Model

2.2.1.1D Hydraulic Coupled Hydrodynamic Model for River Model Network

1D HydraulicThe Yangtze Model River for River and the Network tributaries around the Dongting Lake were modeled based on the Mike11 hydrodynamic module [32]. The unsteady ﬂow in rivers was described by the Saint-Venant The Yangtze River and the tributaries around the Dongting Lake were modeled based on the equations in the hydrodynamic module, as follows: Mike11 hydrodynamic module [32]. The unsteady flow in rivers was described by the Saint‐Venant equations in the hydrodynamic module, as follows:∂Q ∂z + B = q (1) ∂x ∂t L

∂Q ∂Q ∂z ∂A n2|u|Q + 2u +(gA − Bu2) − u2 + g = 0 (2) ∂x ∂x ∂x ∂x R4/3

Water 2018, 10, 1011 5 of 17

where, t is time; x is the Cartesian coordinate; A is cross-sectional area; Q is flow discharge; qL is the lateral inflow or outflow; z is the water level at cross sections; B is the width of water surface; R is the hydraulic radius; u is mean velocity at a cross section; g is the gravitational acceleration; and, n is bed roughness.

2D Shallow Water Model for the Lake The 2D shallow-water equations were used to describe the water movement in the lake area based on the Mike 21 hydrodynamic module, as follows [33]:

∂h ∂hU + i = 0 (3) ∂t ∂xi

∂U ∂U ∂h τ ∂2U i + i = − + + bi + i Ui g f i Ah 2 (4) ∂t ∂xi ∂xi hρ ∂xj in which, i = 1, 2 denotes the x and y directions of the Cartesian coordinate; η is river bed elevation; d is static water depth; h is the total water depth, h = η + d; ρ is the water density; fi is the Coriolis force; Ah is diffusion coefﬁcient of horizontal turbulence; and, τbi is the bottom friction.

Model Coupling Strategies and Moving Boundary Processing The 1D and 2D hydrodynamic models were coupled in the frontal zones using the MIKE FLOOD, which is an integrated simulation system [34]. Water changes between the models occurred through the overlapping cross sections, which should be kept conservative and continuous. To complete the water exchange, the inner boundary conditions of the model junctions were calculated, as follows. In the ﬁrst place, the stage at the inner boundary sections of the 1D model was assigned the averaged water level at the cell faces of the overlapping cross section:

n+1 M z S n+1 2D,k k Z1D = ∑ (5) k=1 S

n+1 n+1 where, Z1D is the boundary stage of the 1D model at the next time step; z2D,k is the updated stage at the cell faces; sk is the width of cell faces; and, S is the total width of the boundary. Secondly, the discharge and stage of all the cross sections were calculated by the 1D model, n+1 n+1 (Z1D , Q ). Finally, based on the water level of the inner boundary and cell topography, the mean water depth n+1 at the cell faces was obtained. Then, the flow velocity (Uk ) at each cell face was computed by the Chezy and Manning formula. Then, the stage and velocity at all the 2D mesh points were calculated n+1 by the 2D model, (z2D , Ui). In addition, discontinuous flow in the rivers and wetland exposure in the lake may occur during the whole hydrological processes. Therefore, to deal with the moving water-land boundary, three controlling water depths were applied in the computation: drying depth (Ddry), flooding depth (D f lood), and wetting depth (Dwet)[35,36]. When the water depth in any cell face was smaller than Ddry, the equations would not be solved in the cell face; when the water depth was larger than Dwet, the equations of momentum and mass conservation in the 2D model should be solved during the computation; and, when the water depth was between the Ddry and Dwet, only the mass conservation equation was solved. It should be pointed out that the flooding depth must be larger than the drying depth and smaller than the wetting depth, Ddry < D f lood < Dwet). Unrealistically high flow velocities and stability problems may occur if a very small value is assigned to the wetting depth. Therefore, the default value was used in this study, as follows: Ddry = 0.005 m, D f lood = 0.05 m, Dwet = 0.1 m, which has been validated using a series of laboratory measurements by DHI. Similarly, in order to judge whether there was water flowing in the river channels, a critical water depth (Dc = 0.005 m) in Water 2018, 10, 1011 6 of 17 the 1D model was applied during the computing process. When the water depth at the cross sections is smaller than the critical water depth, the channels should be considered as dried-up riverways.

2.2.2. Impoundment Evaluation of the TGR In order to reveal and evaluate the water compensation to the Dongting river-lake system after impoundment of the TGR, the inflow and outflow of the TGR from 2008 to 2011 were used as a case study to simulate the hydrological processes downstream of the TGR. The flow processes cover several complete scheduling cycles; therefore, the hydrological changes can be found by comparing the hydrological processes with and without the TGR. Flow inlets were applied at the Yichang, Shimen, Taoyuan, Taojiang, and Xiangtan gauging stations with the observed data. The stage-discharge relation at the Luoshan station was used as the outflow boundary condition. In addition, on account of the distance between the TGR and the model inlet, the flow boundary of the coupled model without the TGR at the Yichang station should be calculated from the inflow of TGR while using the Muskingum flood-routing method [37]. Therefore, the time effect of flood routing from the TGR to the Yichang station has been taken into account in the simulation scenario that without the TGR. The other boundary conditions were kept the same both in the two scenarios that with and without the TGR. In order to quantitatively evaluate the simulation results, the Nash-Sutcliffe efficiency coefficients (NSE) and the normalized root mean squared error (RMSE) were applied to compare the predicted and observed data, as follows [38,39]:

= − N ( − ˆ )2 N ( − )2 NSE 1 ∑i=1 Ti Ti / ∑i=1 Ti Ti (6) r 1 N 2 RMSE = (T − Tˆ ) /T (7) N ∑i=1 i i i where, N denotes the number of data sample; Ti is the observed data; Tˆi is the predicted value of the coupled model; Ti is the mean observed data.

3. Results

3.1. Model Calibration and Validation The model was calibrated based on the hydrological process in 2008 and validated using the observed data in 2011. Flooding and drying events were both covered in the simulating periods. The channels in the river network were divided by 862 cross sections in the 1D model. The Dongting Lake area was divided into 23,436 grid cells and 12,923 nodes in the 2D hydrodynamic model. The grid layout was not equally spaced in this study. In particular, the grid of the deep channels in the lake area was greatly reduced the space, about 30~120 m, while the grid space of lake shoals was increased. It is well known that the Dongting Lake shape during the dry seasons is much more like a wide river due to the low stage. There is no water in the vast beach during the dry seasons. Therefore, the grid space was adjusted based on the topographical changes in the lake area, which was proven to be an effective mean to shorten the computing time on the premise of guaranteeing computation accuracy. The initial value of the Manning roughness coefficient n was assigned according to the water depth β in the river reaches and cell faces of the coupled model based on an empirical formula, n = n0h [40], in which, n denotes the Manning roughness coefficient of the sections or cells; n0 is the roughness coefficient when water depth is equal to 1 m; h is the total water depth; β is an empirical coefficient, generally set as −1/6. In the calibration process, the value of the roughness coefficient was calibrated and adjusted based on the initial value and the observed water level at the different gauging stations of the river-lake system. Table2 and Figure3 show the calibration and validation results of the coupled model. There was an obvious variation in the Manning roughness coefficient from one site to another due to the changes of surface friction characteristics. Crops, weeds, shrubs, and trees in the river point bars and lake shoals Water 2018, 10, x FOR PEER REVIEW 7 of 17

Table 2 and Figure 3 show the calibration and validation results of the coupled model. There was an obvious variation in the Manning roughness coefficient from one site to another due to the Water 2018, 10, 1011 7 of 17 changes of surface friction characteristics. Crops, weeds, shrubs, and trees in the river point bars and lake shoals generally increase the bed roughness. As shown in the Figure 3, there is a good generallyagreement increase between the bedthe roughness.simulated Asand shown observed in the data Figure at3, therethe main is a good gauging agreement stations between that theare simulateddistributed and in the observed lake and data rivers. at the In main addition, gauging water stations volume that aredifference distributed between in the the lake accumulated and rivers. Ininflow addition, volume water and volume the outflow difference volume between should the accumulated be equal to inflowthe difference volume andbetween the outflow the initial volume and shouldfinal water be equal volume to the inside difference the lake between area and the river initial channels. and final The water computing volume insideresults the of lakecalibration area and in river2008 and channels. validation The computing in 2011 presented results ofa good calibration balance in of 2008 water and volume, validation the in relative 2011 presented error were a good2.8% balanceand 3.2%, of waterrespectively, volume, which the relative indicated error that were the 2.8% water and conservation 3.2%, respectively, was whichsatisfied indicated in the coupled that the watermodel. conservation was satisfied in the coupled model.

Table 2.2. Manning roughness coefﬁcientscoefficients inin thethe coupledcoupled model.model.

Subarea SubareaDeep Deep Channels Channels ShoalsShoals in in the the Lake Lake Point BarsPoint in Rivers Bars in Rivers Lake area 0.021~0.034 0.032~0.059 ‐ Lake area 0.021~0.034 0.032~0.059 - MainstreamMainstream 0.018~0.035 0.018~0.035 ‐ - 0.025~0.0460.025~0.046 TributariesTributaries 0.019~0.033 0.019~0.033 ‐ - 0.023~0.0420.023~0.042

30 30 3 (a)Zhicheng (b)Jianli (c)Xinjiangkou 25 25

/s) simulated 3 2 20 m observed

3 20 15 15 1 10 10 5 0

Discharge(×10 5 0 35 37 39 41 43 20 22 24 26 28 30 32 32 34 36 38 40 Water level(m) Water level(m) Water level(m) 800 1200 (d)Shadaoguan (e)Mituosi 1200 (f)Guanjiapu

600 900 /s)

3 900

400 600 600

200 300 300 Discharge(m

0 0 0

32 33 34 35 36 37 38 39 40 30 32 34 36 38 27 28 29 30 31 32 33 34 Water level(m) Water level(m) Water level(m) 31 30 28 (g)Nanzui (h)Heyehu (i)Chenglingji 30 28 26

29 26 24

28 24 22 27 22 20 26 20 18

Simulated water level(m) 25 18 25 26 27 28 29 30 31 18 20 22 24 26 28 30 18 20 22 24 26 28 Observed water level(m) Observed water level(m) Observed water level(m)

Figure 3. Model calibration results at the gauging stations in thethe river-lakeriver‐lake system.system.

The NSENSE andand RMSERMSE werewere computedcomputed andand areare shownshown inin TableTable3 .3. The The NSENSE atat thethe mostmost gauginggauging stationsstations waswas largerlarger thanthan 0.90.9 andand RMSERMSE waswas smallsmall enough,enough, whichwhich indicatedindicated thethe coupledcoupled modelmodel couldcould predictpredict andand evaluateevaluate the hydrological changes accurately after the TGR operation in the study area. TheThe calibrationcalibration andand validationvalidation resultsresults ofof thethe coupledcoupled model model based based on on the the observed observed datadata inin 20082008 andand 2011 showshow that therethere isis aa goodgood agreementagreement betweenbetween thethe simulatedsimulated andand observedobserved datadata atat thethe DongtingDongting river-lakeriver‐lake system.system. The two important evaluation indicators, the NSE and RMSE, indicateindicate thatthat thethe coupledcoupled modelmodel can can predict predict and and evaluate evaluate the the hydrological hydrological changes changes accurately accurately after after the the operation operation of the of TGRthe TGR in the in studythe study area. area.

Water 2018, 10, 1011 8 of 17

Table 3. Computed results of Nash-Sutcliffe efﬁciency coefﬁcients (NSE) and the normalized root mean squared error (RMSE) at the main gauging stations.

2008 2011 No. Stations Discharge (m3/s) Elevation (m) Discharge (m3/s) Elevation (m) NSE RMSE NSE RMSE NSE RMSE NSE RMSE 1 Zhicheng 0.992 0.132 0.993 0.004 0.991 0.134 0.988 0.004 2 Shashi 0.986 0.121 0.988 0.011 0.984 0.124 0.979 0.007 3 Jianli 0.989 0.136 0.969 0.016 0.99 0.129 0.984 0.012 4 Nanzui - - 0.975 0.005 - - 0.972 0.01 5 Heyehu - - 0.984 0.014 - - 0.982 0.013 6 Chenglingji 0.896 0.152 0.991 0.014 0.891 0.161 0.99 0.011 7 Shadaoguan 0.931 0.156 0.961 0.018 0.932 0.071 0.984 0.019 8 Xinjiangkou 0.947 0.142 0.987 0.021 0.939 0.069 0.991 0.023 9 Mituosi 0.968 0.233 0.982 0.016 0.972 0.105 0.976 0.014 10 Kangjiagang 0.927 0.152 0.914 0.023 0.935 0.138 0.913 0.025 11 Guanjiapu 0.935 0.131 0.905 0.017 0.929 0.124 0.901 0.019

3.2. Water Exchanges between the Mainstream and the Dongting Lake The flow regulation of the TGR greatly changed the downstream hydrological processes (Figure2). The peak discharge was significantly reduced, while the low-flow discharge was compensated to some extent in the mainstream after the TGR operation. Therefore, there was an unavoidable impact on the water exchange in the river-lake system during the dry seasons (from the December to the following April). The processes of total flow diversion through the three inlets (Songzi, Taiping, and Ouchi) are shown in Figure4 based on the computed results. However, there was no drastic flow increase at these inlets during the most time of water compensation. The largest flow increase generally occurred in April because of the beginning of flood pre-discharge control in the late April. The mean flow diversion during the four dry seasons from 2008 to 2011 was about 84.7 m3/s and 56.7 m3/s with and without the TGR, and the mean flow increase was approximately 28 m3/s after the TGR operation. The water compensation from the TGR during the dry seasons increased the inflow to the river-lake system, which contributed to some extent to the changes of the diversion ratio. The diversion ratio referred to the value that the total discharge transferred from the mainstream to the tributaries and the Dongting Lake through the three inlets (Songzi, Taiping, and Ouchi) was normalized by the one at the Zhicheng station. The monthly mean diversion ratio with and without the TGR during the dry seasons is presented in Table4. The diversion ratio under the TGR regulation was larger than the one without the TGR during the dry seasons. The maximum difference of diversion ratio occurred in April, about 0.56%. During the whole dry seasons from 2008 to 2011, the monthly mean regulated flow diversion ratio was about 0.894%, which was higher than the one without the TGR by 0.304%. The increased water diversion volume was about 4 × 107 m3 during the dry seasons, which was about the 4.6% of the annually diverted volume without the TGR. Water 2018, 10, 1011 9 of 17 Water 2018, 10, x FOR PEER REVIEW 9 of 17

1600 1600 (a)2008 (b)2009 1200 1200 1200 Natural process 1200 Flow regulation /s) /s) 3 /s) 3 /s)

900 3 900 800 Flow difference 3 800

400 600 600 400

0 300 0 300 Flow diversion ( m ( diversion Flow Flow difference (m Flow difference Flow diversion ( m ( diversion Flow

-400 ( m Flow difference 0 -400 0 -800 -300 -800 -300 01/01 01/21 02/10 03/01 03/21 04/10 12/01 12/21 01/10 01/30 02/19 03/11 03/31 04/20

900 3 900 /s)

800 800 3

600 600 400 400

0 300 0 300 Flow diversion ( m ( Flow diversion Flow diversion ( m ( diversion Flow Flow difference ( m ( difference Flow

-400 0 -400 0 m ( difference Flow

-800 -300 -800 -300 12/01 12/21 01/10 01/30 02/19 03/11 03/31 04/20 12/01 12/21 01/10 01/30 02/19 03/11 03/31 04/20 Time (MM/DD) Time (MM/DD)

FigureFigure 4. Flow 4. Flow diversions diversions during during the the dry seasonsseasons with with and and without without the the TGR. TGR.

Table 4. Monthly mean discharge and flow diversion ratio with and without the TGR during the dry Table 4. Monthly mean discharge and ﬂow diversion ratio with and without the TGR during the dry seasonsseasons from from 2008 2008 to 2011, to 2011, %. %. Diversion Ratio Monthly Mean Discharge (m3/s) Runoff Processes Diversion Ratio Monthly Mean Discharge (m3/s) Runoff ProcessesDec. Jan. Feb. Mar. Apr. Three Outlets Zhicheng Dec. Jan. Feb. Mar. Apr. Three Outlets Zhicheng TGR, R1 0.44 0.48 0.38 0.45 2.72 84.7 6210 TGR, R1 0.44 0.48 0.38 0.45 2.72 84.7 6210 No TGR, R2 0.36 0.2 0.04 0.19 2.16 56.7 5590 No TGR, R2 0.36 0.2 0.04 0.19 2.16 56.7 5590 R1 – R2 R1 – R2 0.08 0.080.28 0.280.34 0.34 0.26 0.26 0.560.56 2828 620 620

Note: theNote: R1 and the R12 anddenoteR2 denote the flow the ﬂow diversion diversion ratio ratio with and and without without the TGRthe TGR during during the dry seasons.the dry seasons.

The Thecoefficient coefficient of water of water exchange exchange was was introduced introduced toto show show the the variation variation in thein the water water exchange exchange betweenbetween the therivers rivers and and the the Dongting Dongting Lake Lake during during thethe drydry seasons seasons in in Figure Figure5. The 5. The coefficient coefficient of of waterwater exchange exchange was was the the ratio ratio of of total flow flow through through the the three three inlets inlets (Songzi, (Songzi, Taiping, Taiping, Ouchi) toOuchi) the flow to the flow dischargedischarge at at the the Chenglingji Chenglingji station. station. As shown As shown in the in Figure the Figure5, the coefficient 5, the coefficient of water exchangeof water exchange with the TGR was generally larger than the one without the TGR. The increase in flow ratio implied that with the TGR was generally larger than the one without the TGR. The increase in flow ratio implied there was more water being stored up in the Dongting Lake. In particular, the increase in water volume that there was more water being stored up in the Dongting Lake. In particular, the increase in water of the Dongting Lake was about 0.38 × 109 m3, 0.23 × 109 m3, 0.39 × 109 m3, and 0.47 × 109 m3, 9 3 9 3 9 3 9 3 volumerespectively, of the Dongting during the Lake dry was seasons about from 0.38 2008 × 10 to m 2011,, 0.23 which × 10 avoidedm , 0.39 the× 10 extensive m , and drying-up 0.47 × 10 m , respectively,in the lake during area from the dry January seasons to March. from 2008 The increasedto 2011, which flow and avoided higher the stage extensive during thedrying water‐up in the lakecompensation area from in the January mainstream to March. increased The the increased flow diversion flow and and restrained higher the stage outflow during from the the lake water compensationarea. The water in the compensation mainstream to increased the downstream the flow of the diversion TGR improved and restrained the water exchange the outflow processes from the lake inarea. the river-lakeThe water system. compensation to the downstream of the TGR improved the water exchange processes in the river‐lake system.

Water 2018, 10, x FOR PEER REVIEW 10 of 17 Water 2018, 10, 1011 10 of 17 Water 2018, 10, x FOR PEER REVIEW 10 of 17 0.16 0.16 (a) 2008 (b) 2009

0.16 Nature process 0.16 0.12 (a) 2008 Flow regulation 0.12 (b) 2009

Nature process

out 0.12 Flow regulation 0.12

/Q 0.08 0.08 in Q out

/Q 0.08 0.08

in 0.04 0.04 Q

Q 0.12 0.12 out 0.08 0.08 /Q in Q 0.08 0.08 0.04 0.04

0.04 0.04 0.00 0.00 121234 121234 Month Month 0.00 0.00 121234 121234 Month Month Figure 5. Variation in coefficient of water exchange in the river‐lake system with and without the TGR. Qin is the total flow through the three inlets (Songzi, Taiping, Ouchi); Qout is the outflow Figure 5. VariationVariation in in coefficient coefficient of of water water exchange exchange in thein the river-lake river‐lake system system with with and withoutand without the TGR. the through the Chenglingji outlet; The natural processes are the results without the TGR, while the flow TGR.Qin is Q thein totalis the flow total through flow through the three the inlets three (Songzi, inlets Taiping, (Songzi, Ouchi); Taiping, Qout Ouchi);is the outflowQout is the through outflow the regulation is the results under the operation of the TGR. throughChenglingji the outlet;Chenglingji The natural outlet; processesThe natural are processes the results are without the results the TGR,without while the the TGR, flow while regulation the flow is regulationthe results is under the results the operation under the of theoperation TGR. of the TGR. 3.3. Changes of Water Level in the River‐Lake System 3.3. Changes The water of Water compensation Level in the River River-Lakeafter‐ Lakethe System TGR operation not only increased the inflow of the mainstreamThe waterwater and compensation compensationflow diversion, after after but the it TGRthe also TGR operation changed operation the not process only not increased onlyof water increased the level inflow in the the of the Dongtinginflow mainstream of Lakethe mainstreamandarea. flow The diversion,stage and‐ durationflow butdiversion, curves it also butwith changed it alsoand thechangedwithout process thethe TGR ofprocess water at theof level water Nanzui, in level the Lujiao, in Dongting the and Dongting Chenglingji Lake Lake area. area.Thestations stage-duration The are stage shown‐duration curvesin Figure curves with 6. and Therewith without andwas withoutno the obvious TGR the at differenceTGR the Nanzui, at the in Nanzui, the Lujiao, extremely andLujiao, Chenglingji low and water Chenglingji stations level at stationsarethe shown Nanzui are in shownstation Figure in6with. ThereFigure and was 6.without There no obvious thewas TGR. no difference obvious However, indifference the it was extremely slightlyin the low extremely higher water than level low the atwater one the Nanzuiwithoutlevel at thestationthe Nanzui TGR with at stationthe and Lujiao without with and and the Chenglingji without TGR. However, the stations TGR. it However,when was slightly the percentageit was higher slightly than of time higher the stage one than without exceeded the one the 62%.without TGR The at thelow TGR Lujiao water at and thelevel Lujiao Chenglingji generally and Chenglingji occurred stations when from stations theDecember percentage when theto the percentage of following time stage of early exceededtime stageApril, 62%. exceeded about The 120 low 62%. days, water The in lowlevelthe Yangtzewater generally level River occurred generally basin, from whichoccurred December just from was to Decemberthe the period following to of the water early following compensation. April, early about April, 120 Therefore, days, about in 120 thethe days, Yangtzechanges in theRiverin theYangtze basin, extremely whichRiver lowbasin, just water was which the level period just were was of water thethe responseperiod compensation. of towater the water Therefore,compensation. compensation. the changes Therefore, As in theshown the extremely changes in the inlowFigure the water extremely 6, the level flow were low regulation thewater response level impact were to on the thethe water Dongtingresponse compensation. Laketo the was water not As compensation.the shown same in in the the Figure Asdifferent shown6, the subarea. in flow the FigureregulationThe influence 6, the impact flow of regulationwater on the compensation Dongting impact Lake on theon was Dongtingthe not eastern the Lake same Dongting was in the not differentLake the same was subarea. inmore the differentobvious The influence subarea.than the of Thewaterwestern influence compensation parts ofof waterthe on lake thecompensation area. eastern The Dongting closer on the to Lake easternthe lake was Dongting outlet, more obvious the Lake greater thanwas stage themore western influenceobvious parts thanof ofwater the westernlakecompensation area. parts The closer ofhas. the tolake the area. lake outlet,The closer the greater to the stagelake influenceoutlet, the of greater water compensation stage influence has. of water compensation has. 32 32 32 (a)Nanzui (b)Lujiao (c)Chenglingji 3229 3229 3229 (a)Nanzui (b)Lujiao (c)Chenglingji 2926 2926 2926

2623 2623 2623 Water level(m) Water 20 20 20 23 Natural process 23 23 Flow regulation Water level(m) Water 2017 2017 2017 0 20406080100 Natural process 0 204060801000 20406080100 Flow regulation 17 Percentage of time stage exceeded(%) 17 Percentage of time stage exceeded(%) 17 Percentage of time stage exceeded(%) 0 204060801000 204060801000 20406080100 Percentage of time stage exceeded(%) Percentage of time stage exceeded(%) Percentage of time stage exceeded(%) FigureFigure 6.6. Stage-durationStage‐duration curvescurves withwith andand withoutwithout thethe TGRTGR atat thethe mainmain lake lake stations. stations. Figure 6. Stage‐duration curves with and without the TGR at the main lake stations. TheThe ChenglingjiChenglingji stationstation waswas locatedlocated inin thethe lowestlowest sitesite ofof thethe DongtingDongting Lake,Lake, whichwhich gaugedgauged thethe ﬂowflowThe processesprocesses Chenglingji fromfrom thethestation lakelake was areaarea located to to the the mainstream. mainstream.in the lowest Therefore,Therefore, site of the the theDongting variationvariation Lake, inin the thewhich lowlow gauged waterwater levellevel the flow processes from the lake area to the mainstream. Therefore, the variation in the low water level

Water 2018, 10, 1011 11 of 17 Water 2018, 10, x FOR PEER REVIEW 11 of 17 at the Chenglingji station can reveal, to some extent, the discharging and impounding results of the Dongting Lake. Lake. In In order order to to describe describe the the changes changes in inthe the low low water water level level at the at theChenglingji Chenglingji station, station, the thewater water level level L95, Lwith95, with the theexceedance exceedance frequency frequency of 95%, of 95%, was was selected selected as a as characteristic a characteristic parameter parameter to toshow show the the stage stage changes changes during during the the dry dry seasons. seasons. The The variation variation in the in theL95 at L95 theat Chenglingji the Chenglingji with with and andwithout without the TGR the TGR from from 2008 2008 to to2011 2011 is isshown shown in in Figure Figure 7.7. The The annual characteristiccharacteristic parameterparameter L L9595 under the operation of the TGR was obviously larger than the one without the TGR, TGR, except except for for 2008. 2008. The characteristic parameterparameter LL9595 under the TGR regulation from 2008 to 2011 was 18.42 m, which was 0.26 m higher than the one without the TGR.

20.0

L95 Annual natural process

L95 Annual flow regulation

19.5 L95 Flow regulation during 2008-2011

L95 Natural process during 2008-2011

19.0

18.5 18.42 18.16 18.0 Water level at Chenglingji(m) at level Water

17.5

17.0 2008 2009 2010 2011 Year

Figure 7. Changes in the water level with a possibility of 95% at the Chenglingji station.

3.4. Variation in Lake Volume The variationvariation in in the the water water exchange exchange after after the TGRthe TGR operation operation was certainwas certain to lead to to lead the lake to the volume lake change.volume Figurechange.8 showsFigure the8 shows volume the variation volume ofvariation the Dongting of the LakeDongting during Lake the during dry seasons the dry from seasons 2008 tofrom 2011. 2008 When to compared2011. When with compared the natural with processes, the natural there wasprocesses, remarkable there volume was remarkable increase during volume the dryincrease seasons during due tothe the dry water seasons compensation due to the ofwater the TGR.compensation Because the of normalthe TGR. ﬂow Because regulation the normal began fromflow 2009,regulation the impact began offrom water 2009, compensation the impact of in water 2008 compensation on the lake volume in 2008 was on the not lake as obvious volume as was the not ones as atobvious the other as the time. ones The at volumetric the other differencestime. The duringvolumetric the fourdifferences periods during were obviously the four differentperiods were with eachobviously other duedifferent to the variationwith each in upstreamother due inﬂow to the and variation water compensation. in upstream There inflow was muchand water more compensatedcompensation. water There when was amuch dry year more occurred, compensated such as water 2011. when The total a dry compensated year occurred, water such was as about2011. 5.7The× total108 mcompensated3 during the water dry seasons was about in 2011, 5.7 which× 108 m lead3 during to approximate the dry seasons 0.22 m in water 2011, level which increase. lead to approximate 0.22 m water level increase.

Water 2018,, 10,, 1011x FOR PEER REVIEW 12 of 17

3 3 Volume increase (a) Volume decrease (b) ) 3 m

9 2 2 10 

1 1

0 0 Lake volume variation ( Lake volume variation

-1 -1 2008/1/25 2008/2/19 2008/3/15 2008/4/9 2008/12/31 2009/1/31 2009/3/3 2009/4/3 3 3 (d) ) (c) 3 m 9 2 2 10 

1 1

0 0 Lake volume variation ( volume Lake variation

-1 -1 2009/12/31 2010/1/31 2010/3/3 2010/4/3 2010/12/31 2011/1/31 2011/3/3 2011/4/3 Date Date

Figure 8. Volumetric differences during dry seasons with and without the TGR (2008–2011). (a–d) Figure 8. Volumetric differences during dry seasons with and without the TGR (2008–2011). (a–d) denote the water compensation periods of each year from 2008 to 2011. denote the water compensation periods of each year from 2008 to 2011. 4. Discussion 4. Discussion The water exchange between the mainstream and the Dongting Lake is an important hydrological The water exchange between the mainstream and the Dongting Lake is an important process to keep the balance of the river-lake system. The flow increase in the mainstream during the hydrological process to keep the balance of the river‐lake system. The flow increase in the period of water compensation from the TGR changed the hydraulic interaction, and finally reached mainstream during the period of water compensation from the TGR changed the hydraulic a new balance in the system. Figure9 presents the relationship between the coefficient of water interaction, and finally reached a new balance in the system. Figure 9 presents the relationship exchange and the outflow from the TGR. As shown in the Figure9, there is a good agreement between between the coefficient of water exchange and the outflow from the TGR. As shown in the Figure 9, the coefficient of water exchange and outflow from the TGR, which indicates that the outflow from the there is a good agreement between the coefficient of water exchange and outflow from the TGR, TGR has a direct influence on the water exchange between the mainstream and the Dongting Lake. which indicates that the outflow from the TGR has a direct influence on the water exchange between With the increase in the outflow from the TGR, the coefficient of water exchange increases in the form the mainstream and the Dongting Lake. With the increase in the outflow from the TGR, the of a power function. Therefore, the outflow increase of the TGR resulted in the variation in the water coefficient of water exchange increases in the form of a power function. Therefore, the outflow exchange between the mainstream and the Dongting Lake, which contributed to relieving the extreme increase of the TGR resulted in the variation in the water exchange between the mainstream and the drought to some extent. Dongting Lake, which contributed to relieving the extreme drought to some extent.

Water 20182018,, 1010,, 1011x FOR PEER REVIEW 13 of 17 Water 2018, 10, x FOR PEER REVIEW 13 of 17

0.16 Natural process 0.16 Flow regulation Natural process 0.14 Power function fitting in without TGR Flow regulation  0.14 Power function fitting in  without TGR Power function fitting in  with TGR 0.12 Power function fitting in  with TGR 0.12 0.10 0.10 3.56 out =5.502E-16×Q 0.08 

/Q 2 3.56 out

in =5.502E-16×QR =0.818 0.08  Q

/Q 2 in 0.06 R =0.818 Q 0.06 0.04  =3.507E-19×Q4.36 0.04  2  =3.507E-19×QR =0.814 4.36 0.02  R2=0.814 0.02 0.00 0.00 2000 4000 6000 8000 10000 12000 14000 2000 4000The discharge 6000 8000in Yichang,Q 10000 (m3/s) 12000 14000 The discharge in Yichang,Q (m3/s)

Figure 9. Relationship between the coefficient of water exchange and the outflow discharge of the Figure 9.9. RelationshipRelationship between between the the coefficient coefficient of of water water exchange exchange and and the the outflow outflow discharge discharge of the of TGR the TGR during the dry seasons (December to the following April) from 2008 to 2011. and during the dry seasons (December to the following April) from 2008 to 2011. η and η denote the TGRdenote during the coefficients the dry seasonsof water (Decemberexchange with to theand followingwithout the April) TGR. from 2008 1to 2011.2 and coefficientsdenote the coefficients of water exchange of water with exchange and without with and the without TGR. the TGR. The outflow from the TGR increased during the water compensation from December to the followingThe outflowoutflow April, which from thechanged TGR increasedthe hydraulic during interaction the water between compensation the rivers from and December the lake to area. the followingTherefore, April, the changed which hydraulic changed the theinteraction hydraulic may interaction interaction lead to some between between variations the the rivers rivers in the and and water the the level lake lake alongarea. area. Therefore,the channels. the The changed relationship hydraulic between interaction the flow may difference lead to at some the Yichang variations station in the and water the levelwater along level thedifference channels. at The the Chenglingjirelationship stationbetween with the flow flowand difference without the at the TGR Yichang is shown station in Figure and the 10. water With level the differenceincrease of atat flow the the Chenglingji difference,Chenglingji stationthere station is with an with obvious and and without withoutlinear the increase TGRthe TGR is shownin isthe shown inwater Figure inlevel 10Figure .difference. With 10. the With increase When the ofincreasecompared flow difference, of with flow the difference, there unregulated is an there obvious flow, is an there linear obvious will increase belinear an in increaseincrease the water about in levelthe 0.49 water difference. m inlevel the difference.water When level compared Whenat the withcomparedChenglingji the unregulated with station the whenunregulated flow, the there flow willflow, increase be there an increaseis will up beto 1000an about increase m 0.493/s during m about in the the 0.49 water dry m seasons.in level the atwater theThe Chenglingji levelwater at level the stationChenglingjiat the lowest when station the outlet flow when of increase the the Dongting flow is up increase to 1000Lake mis reflects 3up/s to during 1000 the themmagnitude3/s dry during seasons. theof lakedry The seasons. watervolume. level The Therefore, at water the lowest level the outletatvolume the oflowest change the Dongting outlet in the of lake Lake the areaDongting reflects was the the Lake magnitude response reflects to of thethe lake variationmagnitude volume. in Therefore, ofthe lake hydraulic volume. the volume interaction Therefore, change of the in thevolumeriver lake‐lake change area system. was in the the response lake area to was the the variation response in the to the hydraulic variation interaction in the hydraulic of the river-lake interaction system. of the river‐lake system.

1.1 1.1

0.6 0.6

H=0.0005Q-0.0104 2 H=0.0005Q-0.0104R =0.96 R2=0.96

0.1

Monthly mean stage difference in Chenglingji,H(m) in difference mean stage Monthly 0.1 Monthly mean stage difference in Chenglingji,H(m) in difference mean stage Monthly -500 100 700 1300 1900 -500 100Monthly mean 700 flow difference in 1300 Yichang,Q(m3/s) 1900 Monthly mean flow difference in Yichang,Q(m3/s)

-0.4 . -0.4 . Figure 10. Relationship between the water level difference at the Chenglingji station and the outflow outﬂow differenceFigure 10. Relationshipat the YichangYichang between stationstation withthewith water andand without withoutlevel difference thethe TGR.TGR. at the Chenglingji station and the outflow difference at the Yichang station with and without the TGR.

Water 2018, 10, 1011 14 of 17 Water 2018, 10, x FOR PEER REVIEW 14 of 17

InIn fact, fact, the the stage stage changes changes in in the the lakelake areaarea werewere closely related related to to the the inflow inflow and and outflow outflow of ofthe the DongtingDongting Lake Lake [41 [41].]. Figure Figure 11 presents 11 presents the relationshipthe relationship between between the outflow the outflow at the at Chenglingji the Chenglingji station andstation the hydraulic and the gradient hydraulic and gradient discharge and at discharge the Yichang at thestation. Yichang The increase station. inThe flow increase discharge in flow at the Yichangdischarge station at the made Yichang nearly station no difference made nearly to the no outflow difference of theto Dongtingthe outflow Lake of the when Dongting the hydraulic Lake gradientwhen thewas hydraulic between 0.000043~0.000048. gradient was between While 0.000043~0.000048. the hydraulic gradient While the was hydraulic smaller than gradient 0.000043 was or largersmaller than than 0.000048, 0.000043 the or larger outflow than increased 0.000048, greatly the outflow with increased the flow greatly increase with at the the flow Yichang increase station. at However,the Yichang the hydraulic station. gradientHowever, played the hydraulic an important gradient role in played the whole an important changing role processes in the of whole outflow. changing processes of outflow. The outflow variation directly resulted in the changes of water level The outflow variation directly resulted in the changes of water level and water volume in the Dongting and water volume in the Dongting Lake. Therefore, in order to contribute to the ecological Lake. Therefore, in order to contribute to the ecological restoration and environmental improvement restoration and environmental improvement in the river-lake system, much more attention should in the river-lake system, much more attention should be paid to the hydraulic gradient variation, be paid to the hydraulic gradient variation, which greatly affects the hydraulic interaction between which greatly affects the hydraulic interaction between the rivers and the Dongting Lake. When there is the rivers and the Dongting Lake. When there is a constant outflow from the TGR, i.e., the discharge a constantat the Yichang outflow station from theis kept TGR, the i.e., same, the discharge the outflow at thefrom Yichang the Dongting station isLake kept to the the same, Yangtze the outflowRiver fromcan the be Dongting adjusted Lakeby the to hydraulic the Yangtze gradient River can to some be adjusted extent. by Consequently, the hydraulic a gradient reasonable to some hydraulic extent. Consequently,gradient definitely a reasonable contributes hydraulic to the gradientimprovement definitely of water contributes volume of to the the Dongting improvement Lake during of water volumethe extremely of the Dongting dry seasons. Lake during the extremely dry seasons.

FigureFigure 11. 11.Outﬂow Outflow processes processes with with the the discharge discharge and and hydraulic hydraulic gradient gradient during during the the dry dry seasons seasons (December to the following April) from 2008 to 2011. The outflow of the Dongting Lake was (December to the following April) from 2008 to 2011. The outﬂow of the Dongting Lake was observed observed at the Chenglingji station; the hydraulic gradient was between the Zhicheng and at the Chenglingji station; the hydraulic gradient was between the Zhicheng and Chenglingji stations. Chenglingji stations.

5. Conclusions5. Conclusions HydrologicalHydrological changes changes occurred occurred increasingly increasingly frequently frequently around around the world the due world to the due widespread to the influenceswidespread of anthropogenic influences of anthropogenic activities. Water activities. compensation Water compensation of the TGR, as of one the ofTGR, the importantas one of the flow regulationimportant measures, flow regulation greatly measures, contributed greatly to the contributed changes of to hydrological the changes of regime hydrological in the downstream regime in river-lakethe downstream system. river-lake Based on system. the hydrodynamic Based on the model hydrodynamic simulation model and simulation data analysis, and data the interesting analysis, andthe important interesting conclusions and important have conclusions been drawn, have as been follows: drawn, as follows: (1) Water exchange between the Yangtze River and the Dongting Lake was further strengthened (1) Water exchange between the Yangtze River and the Dongting Lake was further afterstrengthened the TGR operation. after the TGR Flow operation. diversion Flow from diversion the Yangtze from Riverthe Yangtze to the River Dongting to the Lake Dongting increased Lake to differentincreased degrees to different during thedegrees dry seasons during thefrom dry 2008 seasons to 2011. from Furthermore, 2008 to 2011. the coefficient Furthermore, of water the exchangecoefficient also of increased water exchange during thealso water increased compensation during the periods water compensationof the TGR, which periods implied of the that TGR, there waswhich a less implied increase that in there the outflow was a less from increase the Dongting in the outflow Lake. from the Dongting Lake. (2)(2) There There was was no no good good agreement agreement inin lakelake levellevel changeschanges at at the the lake lake gauging gauging stations. stations. When When the the timetime stage stage percentage percentage was was larger larger than than 62%, 62%, the lake the level lake at level the Chenglingji at the Chenglingji outlet markedly outlet markedly increased dueincreased to the level dueincrease to the level in the increase mainstream. in the mainstream. The lake level, The withlake level, a possibility with a possibility of 95% at theof 95% Chenglingji at the outlet,Chenglingji increased outlet, by 0.26increased m under by 0.26 the m water under compensation the water compensation from the TGR from than the theTGR one than without the one the

Water 2018, 10, 1011 15 of 17

TGR. Consequently, the lake volume also increased to different degrees during the dry seasons from 2008 to 2011. (3) The changes in inflow and outflow of the Dongting Lake were the determining factors of the hydrological regime of lake area. The water exchange between the Yangtze River and the Dongting Lake principally resulted from the outflow from the TGR and the hydraulic gradient between the inlets and the outlet of the Dongting Lake. When the water compensation is conducted at the proper hydraulic gradient range, the lake level and water volume will be effectively improved in the Dongting Lake, consequently, the ecological environment will also be recovered to some extent during the dry seasons.

Author Contributions: J.Z. and M.D. conceived the article structure and collected the data; D.W. and W.X. analyzed the data; T.Y. edited the manuscript. L.F. contributed materials/analysis tools; L.C. wrote the paper. Funding: This research was funded by National Key R&D Program of China (2017YFC0405900); the National Natural Science Foundation of China (51509273, 51679094, 51409085); the National Key Research and Development Plan (2016YFC0400901); Open Foundation of State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering (2015491111); the Open Research Fund of State Key Laboratory of Simulation and Regulation of Water Cycle in River Basin (China Institute of Water Resources and Hydropower Research), Grant NO: IWHR-SKL-201607, 2016TS07; the Technology Demonstration Projects of Ministry of Water Resources (SF2016-10); the Research and Development Projects on Application Technology of Heilongjiang Province (GZ16B011) and Fundamental Research Funds for the Central Universities (2017KFYXJJ194, 2016YXZD048). Acknowledgments: The authors gratefully acknowledge the anonymous reviewers and editors for their careful reviews and suggestions, which significantly contributed to the manuscript improvement. Conflicts of Interest: The authors declare no conflicts of interest. The founding sponsors had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; nor in the decision to publish the results.

References

1. Omer, A.M. Energy, environment and sustainable development. Renew. Sust. Energy Rev. 2008, 12, 2265–2300. [CrossRef] 2. Fang, C.; Bao, C.; Huang, J. Management implications to water resources constraint force on socio-economic system in rapid urbanization: A case study of the Hexi Corridor, NW China. Water Resour. Manag. 2007, 21, 1613–1633. [CrossRef] 3. Chen, L.; Singh, V.P.; Huang, K.D. Bayesian Technique for the Selection of Probability Distributions for Frequency Analyses of Hydrometeorological Extremes. Entropy 2018, 20, 117. [CrossRef] 4. González-Fernández, R.A.; Oviedo-Sanabria, R.E.; Leite da Silva, A.M. Generating capacity reliability assessment of the Itaipu hydroelectric plant via sequential Monte Carlo simulation. In Proceedings of the Power Systems Computation Conference (PSCC), Wroclaw, Poland, 18–22 August 2014; pp. 1–7. 5. Postel, S.; Richter, B. Rivers for Life: Managing Water for People and Nature; Island Press: Washington, DC, USA, 2012. 6. Tecle, A. Downstream Effects of Damming the Colorado River. Int. J. Lakes Rivers 2017, 10, 7–33. 7. Mendelssohn, I.A.; Andersen, G.L.; Baltz, D.M.; Caffey, R.H. Oil impacts on coastal wetlands: Implications for the Mississippi River Delta ecosystem after the Deepwater Horizon oil spill. Bioscience 2012, 62, 562–574. [CrossRef] 8. Chen, L.; Singh, V.P.; Guo, S.; Hao, Z.; Li, T. Flood coincidence risk analysis using multivariate copula functions. J. Hydrol. Eng. 2011, 17, 742–755. [CrossRef] 9. Chen, L.; Singh, V.P.; Guo, S.; Zhou, J.; Zhang, J. Copula-based method for multisite monthly and daily streamﬂow simulation. J. Hydrol. 2015, 528, 369–384. [CrossRef] 10. Grizzetti, B.; Pistocchi, A.; Liquete, C.; Udias, A.; Bouraoui, F.; van de Bund, W. Human pressures and ecological status of European rivers. Sci. Rep. 2017, 7, 205. [CrossRef][PubMed] 11. Duc, N.A.; Nguyen, L.T.; Thai, T.H.; Khan, A.; Rautenstrauch, K.; Schmidt, C. Assessing Cumulative Impacts of the Proposed Lower Mekong Basin Hydropower Cascade on the Mekong River Floodplains and Delta–Overview of Integrated Modeling Methods and Results. J. Hydrol. 2018, in press. 12. Wisser, D.; Frolking, S.; Douglas, E.M.; Fekete, B.M. The signiﬁcance of local water resources captured in small reservoirs for crop production—A global-scale analysis. J. Hydrol. 2010, 384, 264–275. [CrossRef] Water 2018, 10, 1011 16 of 17

13. Rahel, F.J.; Olden, J.D. Assessing the effects of climate change on aquatic invasive species. Conserv. Biol. 2008, 22, 521–533. [CrossRef][PubMed] 14. Liu, Z.; Guo, S.; Guo, J.; Liu, P. The impact of Three Gorges Reservoir refill operation on water levels in Poyang Lake, China. Stoch. Environ. Res. Risk Assess. 2017, 31, 879–891. [CrossRef] 15. Huang, Z.; Wu, B. Ecological and Environmental Monitoring After Reservoir Impoundment. In Three Gorges Dam; Springer: Myrtle Beach, SC, USA, 2018; pp. 279–306. 16. Chen, L.; Singh, V.P. Entropy-based derivation of generalized distributions for hydrometeorological frequency analysis. J. Hydrol. 2018, 557, 699–712. [CrossRef] 17. Mao, J.; Zhang, P.; Dai, L.; Dai, H.; Hu, T. Optimal operation of a multi-reservoir system for environmental water demand of a river-connected lake. Hydrol. Res. 2016, 47, 206–224. [CrossRef] 18. Yang, S.; Milliman, J.D.; Xu, K.; Deng, B.; Zhang, X.; Luo, X. Downstream sedimentary and geomorphic impacts of the Three Gorges Dam on the Yangtze River. Earth-Sci. Rev. 2014, 138, 469–486. [CrossRef] 19. Song, Y.; Cheng, F.; Murphy, B.R.; Xie, S. Downstream effects of the Three Gorges Dam on larval dispersal, spatial distribution, and growth of the four major Chinese carps call for reprioritizing conservation measures. Can. J. Fish. Aquat. Sci. 2017, 75, 141–151. [CrossRef] 20. Gao, X.; Lin, P.; Li, M.; Duan, Z.; Liu, H. Impact of the Three Gorges Dam on the spawning stock and natural reproduction of Chinese sturgeon in Changjiang River, China. Chin. J. Oceanol. Limnol. 2016, 34, 894–901. [CrossRef] 21. Feng, L.; Hu, C.; Chen, X.; Zhao, X. Dramatic inundation changes of China’s two largest freshwater lakes linked to the Three Gorges Dam. Environ. Sci. Technol. 2013, 47, 9628–9634. [CrossRef][PubMed] 22. Huang, Q.; Sun, Z.D.; Opp, C.; Lotz, T.; Jiang, J.H.; Lai, X.J. Hydrological Drought at Dongting Lake: Its Detection, Characterization, and Challenges Associated with Three Gorges Dam in Central Yangtze, China. Water Resour. Manag. 2014, 28, 5377–5388. [CrossRef] 23. Li, S.; Xiong, L.; Dong, L.; Zhang, J. Effects of the Three Gorges Reservoir on the hydrological droughts at the downstream Yichang station during 2003–2011. Hydrol. Process. 2013, 27, 3981–3993. [CrossRef] 24. Lai, X.; Jiang, J.; Yang, G.; Lu, X.X. Should the Three Gorges Dam be blamed for the extremely low water levels in the middle-lower Yangtze River? Hydrol. Process. 2014, 28, 150–160. [CrossRef] 25. Wang, J.; Sheng, Y.; Wada, Y. Little impact of the Three Gorges Dam on recent decadal lake decline across China’s Yangtze Plain. Water Resour. Res. 2017, 53, 3854–3877. [CrossRef] 26. Zhang, W.; Yang, Y.; Zhang, M.; Li, Y.; Zhu, L.; You, X.; Wang, D.; Xu, J. Mechanisms of suspended sediment restoration and bed level compensation in downstream reaches of the Three Gorges Projects (TGP). J. Geogr. Sci. 2017, 27, 463–480. [CrossRef] 27. Chen, X.; Yan, Y.; Fu, R.; Dou, X.; Zhang, E. Sediment transport from the Yangtze River, China, into the sea over the Post-Three Gorge Dam Period: A discussion. Quat. Int. 2008, 186, 55–64. [CrossRef] 28. Chen, L.; Singh, V.P.; Guo, S.; Zhou, J.; Zhang, J.; Liu, P. An objective method for partitioning the entire flood season into multiple sub-seasons. J. Hydrol. 2015, 528, 621–630. [CrossRef] 29. Chen, L.; Guo, S.; Yan, B.; Liu, P.; Fang, B. A new seasonal design flood method based on bivariate joint distribution of flood magnitude and date of occurrence. Hydrol. Sci. J. 2010, 55, 1264–1280. [CrossRef] 30. Liu, X.; Chen, L.; Zhu, Y.; Singh, V.P.; Qu, G.; Guo, X. Multi-objective reservoir operation during flood season considering spillway optimization. J. Hydrol. 2017, 552, 554–563. [CrossRef] 31. Three Gorges Corporation. Available online: http://www.ctg.com.cn/ (accessed on 15 May 2018). 32. Danish Hydraulic Institute (DHI). A Modelling System for Rivers and Channels. In Reference Manual; DHI Water & Environment: Aarhus N, Denmark, 2011. 33. Danish Hydraulic Institute (DHI). MIKE 21 Flow Model: Hydrodynamic Module Scientific Documentation. In User Mannual; DHI: Horsholm, Denmark, 2009. 34. DHI. MIKE FLOOD 1D-2D Modelling. In User Manual; DHI: Horsholm, Denmark, 2007; pp. 5–20. 35. Zhao, D.; Shen, H.; Tabios, G., III; Lai, J.; Tan, W. Finite-volume two-dimensional unsteady-flow model for river basins. J. Hydraul. Eng. 1994, 120, 863–883. [CrossRef] 36. Sleigh, P.; Gaskell, P.; Berzins, M.; Wright, N. An unstructured finite-volume algorithm for predicting flow in rivers and estuaries. Comput. Fludids 1998, 27, 479–508. [CrossRef] 37. Cunge, J.A. On the subject of a flood propagation computation method (Musklngum method). J. Hydraul. Res. 1969, 7, 205–230. [CrossRef] Water 2018, 10, 1011 17 of 17

38. Zhang, J.; Chen, L.; Singh, V.P.; Cao, H.; Wang, D. Determination of the distribution of flood forecasting error. Nat. Hazards 2015, 75, 1389–1402. [CrossRef] 39. Chen, L.; Singh, V.P.; Lu, W.; Zhang, J.; Zhou, J.; Guo, S. Streamflow forecast uncertainty evolution and its effect on real-time reservoir operation. J. Hydrol. 2016, 540, 712–726. [CrossRef] 40. Jiang, J.H.; Lai, X.J.; Huang, Q. The characteristics of flood responses to the restoration of polders on Dongting Lake, China. Hydrol. Sci. J. 2007, 52, 671–685. 41. Dai, M.L.; Wang, J.; Zhang, M.B.; Chen, X. Impact of the Three Gorges Project operation on the water exchange between Dongting Lake and the Yangtze River. Int. J. Sediment. Res. 2017, 32, 506–514. [CrossRef]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).