An absolute essential for effective water resource management and ecological restoration is knowing the temporal and spatial variation of runoff. The objective of this study was to determine the spatial and temporal changes in runoff at the main hydrological stations along the Minjiang River and the Dadu River between 1961 and 2016 using the non-parametric Mann-Kendall test and annual variation analysis. Canonical correspondence analysis and regression analysis were used to determine the contribution of anthropogenic disturbance, vegetation, and climatic conditions to runoff change. The runoff at each station of the Minjiang River showed a clear decreasing trend, whereas the decreasing trend of the Dadu River was not significant. Moreover, the discharge at the Shawan (SW) station upstream of the Dadu River and the Gaochang (GC) station downstream of the Minjiang River have changed significantly during the flood and non-flood seasons since 2000, while the discharge at other stations has not changed significantly. The average annual runoff in the non-flood season at SW and GC in 2011–2016 increased by approximately 26.21 and 36.47%, respectively, compared with 1961–2010. Anthropogenic disturbance, vegetation, and climatic conditions in the Minjiang River Basin accounted for 76.24, 13.62, and 10.14%, respectively, of the runoff change in the basin. Water consumption and total reservoir capacity were the specific factors most affecting runoff change in the basin, accounting for 15.10 and 13.94%, respectively, of the changes in runoff. The research can provide important support for the ecological restoration of Minjiang River Basin and Yangtze River Basin.

  • A mutation occurred around 2010, hereafter the annual average flow at the Gaochang station increased by approximately 36.47% in the non-flood season.

  • A significant downward trend in runoff from 1960 to 2016 occured in Minjiang.

  • Hydropower projects and water resources utilization may have contributed greatly to changes in runoff.

Graphical Abstract

Graphical Abstract
Graphical Abstract

The distribution of water resources in time and space is widely affected by climate change and human activities (Ngongondo et al. 2011; Kale et al. 2016; Xu et al. 2018; Bombelli et al. 2021). Precipitation and temperature are vital climate variables proven to affect the hydrological environment (Nadal-Romero et al. 2008; Zhang et al. 2011; Li et al. 2016). Precipitation and evaporation are the main meteorological factors affecting watershed runoff (Cheng et al. 2011; Chang et al. 2016; Wu et al. 2020; Nazeer et al. 2022). Human activities (e.g., land use/cover variations, hydroelectric exploitation, afforestation, urbanization, water use, etc.) have obvious effects on the water cycle, resulting in great changes in the spatial-temporal distribution of water resources (Xu et al. 2008; Guo et al. 2018a, 2018b; Korkanç 2018; Lu et al. 2020). Many researchers have conducted studies on the impact of climate and human activities on runoff and employed various methods to distinguish the contributions of climate change and human activities to runoff changes (Cui et al. 2011; Maalim et al. 2013; Ejder et al. 2016; Xu et al. 2020). Trends in runoff, temperature, and precipitation in the Qom Rood Watershed in Iran during 1979–2016 were studied at monthly and annual time scales using the Mann-Kendall test, as well as the double-mass curve and a multiple regression model (Callow & Smettem 2009; Arslan et al. 2020; Yang et al. 2022). According to Yaghmaei et al. (2018), the results proved that natural driving forces showed no effect on runoff changes, and the observed decreasing tendency in runoff resulted from dam operations. Wu et al. (2020) used integrated methods of hydrological modelling and numerical analysis to show that runoff characteristics responded to climatic variations and hydroelectric exploitation in the Jinsha River Basin. Additionally, Chang et al. (2016) reported that climate changes and human activities accounted for ∼20 and ∼80%, respectively, of runoff changes. Hydroelectric exploitation has a major effect on the hydrological environment (Zhai et al. 2010; Tuset et al. 2016).

The Minjiang River is the fifth largest river in China. It is also the largest tributary of the Yangtze River and therefore generates important considerations and concerns regarding runoff. Located in the upper reaches of the Yangtze River, it plays an important role in the ecological security of the Yangtze River Basin. Additionally, the Minjiang River is one of the tributaries with the most fragile ecosystem, the most frequent human activities, and the largest engineering development intensity in the Yangtze River Basin. At present, the Minjiang River Basin (MRB) is facing prominent ecological and environmental problems, such as water reduction or even river cutoff, water pollution, biodiversity reduction, frequent geological disasters, and so on (Guo et al. 2018a, 2018b; Fan et al. 2021). Due to overdevelopment of water diversion along the Minjiang River (especially in the upper reaches), runoff of the Jinma River (a tributary of the Minjiang River) has been seriously reduced, and even river cutoff has occurred, potentially causing damage to the ecological environment of the MRB, as well as loss of fish and other resources (Wei et al. 2021). With the operation of the Zipingpu Hydropower Station on the Minjiang River and strict environmental management, both water reduction and river cutoff have been effectively alleviated (Liu et al. 2020).

The middle and lower reaches of the Minjiang River are the most economically and socially developed areas in Sichuan Province, especially in the Chengdu Plain where many large-scale irrigation areas are located for grain production (e.g., Dujiangyan and Tongjiyan). A large amount of water consumption in these areas has a strong impact on the distribution of water resources in the MRB. The water resources of the Minjiang River are not as rich as those of the Dadu River, however, for economic and social development, there is an extremely large water demand, resulting in a great disparity between water supply and water demand. The development rate of water resources in the MRB is high, and more reservoirs and dams are planned along the mainstream portions of the Minjiang River, the Dadu River, and other tributaries. According to comprehensive planning for the MRB, the runoff change of the entire basin will be regulated by hydropower and water diversion projects in the future and will have a great influence on the runoff of the Yangtze River. Since the implementation of the Yangtze River Protection Law, ecological and environmental protection in the Yangtze River Basin has attracted much more attention.

At present, many researchers have studied runoff changes in the Yangtze River (Zhang et al. 2006; Yan et al. 2011; Naveed et al. 2022). Some researchers have also studied runoff changes in the upper reaches of the MRB. Several studies have confirmed a downward trend in annual flow and sediment discharge for the Yangtze River over the past decades (Liu et al. 2007; Yan et al. 2011). Hou et al. (2018) demonstrated that the effects of human activities and climate change on runoff variation accounted for ∼77 and 23%, respectively, of annual runoff variation in the upstream portion of the MRB. Zhang et al. (2020) reported that the contributions of climate variation and land use/land cover change to runoff depth were at the ratio of 5:1 in the upper reaches of the MRB.

Some researchers have conducted studies on runoff changes at the Zhenjiangguan (ZJG) station in the upstream region of the Minjiang River, revealing the contributions of climate and forest vegetation to the runoff change (Nie et al. 2020). The results provided a good reference for our research. However, the runoff variation of the entire river basin has always been ignored. Economic and social development along the Minjiang River is highly developed, and the Dadu River is an important hydropower development base in China. Consequently, their impacts on the runoff change of the MRB should not be ignored. Decision makers need to know the flow variation of the entire MRB and the roles of variables affecting runoff. Taking the Minjiang River and its main tributary (the Dadu River) as the research objects, the temporal and spatial runoff changes in the most recent 60-year period were studied. The specific objectives of the study were to: (1) determine the change rule of the temporal and spatial water resources in the MRB; (2) determine the variation characteristics of inter-annual and annual runoff; (3) discuss the impact of artificial disturbance, precipitation, and vegetation on runoff; and (4) provide technical support for ecological flow guarantees and ecological restoration of the MRB and the Yangtze River Basin.

Study area

The Minjiang River, with a total length of 735 km and an average annual flow of 3022 m3/s, is one of the primary tributaries on the north bank of the Yangtze River. The main tributaries of the MRB are the Dadu River and the Qingyi River. The Dadu River, with a length of 1062 km and an average annual flow of 1670 m3/s, is the largest tributary of the Minjiang River. There is a high degree of development and utilization in the MRB, and the current utilization ratio of water resources is approximately 11.4%. Historically, the disordered development of diversion power stations has seriously damaged the ecological environment. So far, 10 hydropower stations have been built along the Minjiang River. Except for the Zipingpu Reservoir hydropower station with seasonal regulation performance, other stations are diversion hydropower stations. According to official statistics, excessive diversion development in the upper reaches of the Minjiang River before 2005 led to intermittent river cutoff over approximately 80 km. In addition, 15 hydropower stations have been built along the Dadu River. The control point for the Dadu River is the Pubugou hydropower station. Zipingpu and Pubugou are control reservoirs, and have a total storage capacity of 1.112 billion m3 and 5.39 billion m3, respectively, and were put into operation in 2006 and 2010, respectively. In addition, there is another control project under construction named the Shuangjiangkou Reservoir which is an annual regulation reservoir.

Data source

Monthly streamflow at six hydrological stations was collected for analysis. Those stations were Zhenjiangguan (ZJG), Dujiangyan (DJY), Gaochang (GC) located on the Minjiang River, and Dajin (DJ), Luding (LD), and Shawan (SW) located on the Dadu River. ZJG is the first hydrological station on the upper reaches of the Minjiang River. DJY is located downstream of the Zipingpu Hydropower Station, which is the representative station in the middle reaches of the Minjiang River. GC is located at the end of the lower reaches of the Minjiang River and is the gateway station of the MRB. DJ is the hydrological station on the upper reaches of the Dadu River, LD is the hydrological station on the middle reaches of the Dadu River, and SW is the gateway station of the Dadu River. The monthly runoff data (1957–2016) at ZJG, DJY, GC, DJ, LD, and SW were provided by the Yangtze River Water Commission and the Hydrological Bureau of Sichuan Province. Irrigation area, afforestation area, population, water consumption, number of reservoirs, and total storage capacity of reservoirs were obtained from the Sichuan Provincial Bureau of Statistics. Annual precipitation data were obtained from the Sichuan Meteorological Bureau. Normalized difference vegetation index (NDVI) was derived from remote sensing interpretation (MODIS MOD13).

Methods

(1) Analysis of runoff variation

The nonparametric rank correlation test (hereafter referred to as the M–K method) based on the correlation between the ranks of a time series and their time order (Mann 1945; Kendall 1975) was used to detect when changes in runoff and sediment loading occurred. This method has been commonly applied to assess the significance of trends in water quality, streamflow, temperature, and precipitation (Yue et al. 2002; Gocic & Trajkovic 2013; Kisi & Ay 2014; Nyikadzino et al. 2020). For a time-series of n observations, the Mann-Kendall test statistic (s) is defined as
formula
(1)
formula
(2)
where n is the number of observations, xi is the rank for ith observation, and xj is the rank for jth observation.
Assuming that all data are normally distributed, the mean E(s) and variance Var(s) are given as:
formula
(3)
formula
(4)
If there is a possibility of a tie in the value of x, Var(s) is computed as
formula
(5)
where m is the number of tied groups.
The standardized variable u is calculated using Equation (6):
formula
(6)
(2) Analysis of the driving factors of runoff variation

Eight indexes, including irrigation area, afforestation area, population, water consumption, number of reservoirs, the total storage capacity of reservoirs, NDVI, and annual precipitation, were selected as potential driving factors to analyze their influence on average annual runoff and runoff in the non-flood season at hydrological stations. We used Canoco 4.5 and SPSS software to conduct canonical correspondence analysis and regression analysis to study the influence of each driving factor on runoff change in the basin.

Interannual variation of runoff

Monthly discharge variation at the hydrological stations mentioned above from 1961 to 2016 is shown in Figure 1. There were obvious differences in the temporal and spatial runoff distribution between the Minjiang River and the Dadu River.
Figure 1

Variation of annual discharge at the main hydrological stations in the Minjiang River Basin, China, from 1961 to 2016. Hydrological station abbreviations are defined in Figure 11.

Figure 1

Variation of annual discharge at the main hydrological stations in the Minjiang River Basin, China, from 1961 to 2016. Hydrological station abbreviations are defined in Figure 11.

Close modal

Spatial variation of runoff

In general, the runoff of the Minjiang River increased from upstream to downstream. The average annual runoff values at ZJG and DJY were 1.7 and 14.4 billion m³, respectively, and the discharges were 55 and 464 m³/s, respectively. Additionally, the average annual runoff values at DJ, LD, and SW were 16.5, 27.5, and 46.1 billion m³, respectively, and the discharges were 522, 885, and 1,481 m³/s, respectively. The Dadu River is the largest tributary of the Minjiang River and accounts for approximately 55% of the total runoff of the Minjiang River. After the Dadu River feeds into the Minjiang River, the runoff at GC increases greatly, producing an average annual runoff of 84 billion m³ and discharge of 2693 m³/s.

Temporal variation of runoff

From the inclination rates in Figure 1 and Table 1, it can be seen that the total runoff of the Minjiang and Dadu rivers exhibited a downward trend from 1961 to 2016. Additionally, the reduction extent increased gradually from upstream to downstream. The monthly average discharges at ZJG, DJY, SW, and GS were selected to undergo an M–K trend test (Figure 2). If the UFk and UBk curves exceed the critical boundary, then the upward or downward trend is significant. If the UFk and UBk curves intersect and the intersection appears between the critical lines, the time corresponding to the intersection is the time when the change begins. The interannual variation of runoff at each station was not significant, and the same result can be seen in Figure 2. However, the runoff at each station changed significantly, both in the flood season (May–October) and in the non-flood season (November–April) (Table 1). From this analysis, the inclination rates of runoff variations in flood and non-flood seasons were found to be significantly greater than those of average annual variations.
Table 1

Discharge at the main hydrological stations in the Minjiang River Basin, China, from 1961 to 2016

RiverHydrological stationAverage annual runoff (billion m³)Average annual discharge (m3/s)Inclination rate of runoff variation over the past 60 years
Average annual dischargeAverage discharge in the flood seasonAverage discharge in the non-flood season
Minjiang River ZJG 1.7 55 −0.1955 −0.2777 −0.1178 
DJY 14.6 464 −0.5373 −1.8706 1.0583 
GC 84.7 2693 −6.3652 −14.556 2.9449 
Dadu River DJ 16.5 522 −0.7317 −1.4154 −0.0635 
LD 27.9 885 −0.0139 0.3473 −0.2344 
SW 46.7 1481 −0.4197 −2.7265 2.5322 
RiverHydrological stationAverage annual runoff (billion m³)Average annual discharge (m3/s)Inclination rate of runoff variation over the past 60 years
Average annual dischargeAverage discharge in the flood seasonAverage discharge in the non-flood season
Minjiang River ZJG 1.7 55 −0.1955 −0.2777 −0.1178 
DJY 14.6 464 −0.5373 −1.8706 1.0583 
GC 84.7 2693 −6.3652 −14.556 2.9449 
Dadu River DJ 16.5 522 −0.7317 −1.4154 −0.0635 
LD 27.9 885 −0.0139 0.3473 −0.2344 
SW 46.7 1481 −0.4197 −2.7265 2.5322 

Hydrological station abbreviations are defined in Figure 11.

Figure 2

M–K test results for annual discharge at the main hydrological stations in the Minjiang River Basin, China, from 1961 to 2016. Hydrological station abbreviations are defined in Figure 11.

Figure 2

M–K test results for annual discharge at the main hydrological stations in the Minjiang River Basin, China, from 1961 to 2016. Hydrological station abbreviations are defined in Figure 11.

Close modal

The inclination rates of runoff variation at ZJG and DJY were −0.1955 and −0.5373, respectively, indicating that the runoff at ZJG did not change significantly, while the runoff at DJY showed a slight reduction. The inclination rates at ZJG in the flood and non-flood seasons were consistent with the inclination rate of the average annual variation, whereas the inclination rate of DJY in the flood and non-flood seasons was significantly greater than the inclination rate of the average annual variation, showing a downward trend in the flood season and an upward trend in the non-flood season.

The inclination rate of the average annual variation at LD was −0.0139, indicating that it has changed insignificantly over the past 60 years. The inclination rate at SW was 0.4197, showing a slight downward trend. The inclination rates at LD in the non-flood and flood seasons were consistent with that of average annual variation, whereas the inclination rates at SW in the flood and non-flood seasons were significantly greater than that of the average annual variation, indicating that the discharge at SW changed significantly in flood and non-flood seasons, showing a downward trend in flood seasons and an upward trend in non-flood seasons.

After the Dadu River drains into the Minjiang River, the runoff at GC was observed to change with an inclination rate of −6.3652 over the past 60 years, showing an obvious downward trend. Additionally, GC was affected by the runoff change in the upper and middle reaches of the Minjiang and Dadu Rivers. The inclination rates of runoff variation in the flood and non-flood seasons were significantly greater than that of average annual variation. In recent years, runoff in the flood season has decreased significantly and increased significantly in the non-flood season.

Variation of annual runoff

Because runoff has changed significantly both in the flood and non-flood seasons over the past 60 years, we used only one set of data of monthly average discharges at ZJG, DJY, SW, and GS in the non-flood season to conduct an M–K trend test (Figure 3). Runoff in the non-flood season at ZJG showed a downward trend from 1961 to 2016. From the variation of the UFk curve, we noted that runoff at DJY in the non-flood season showed a downward trend from 1970 to 2009, although the decrease was not significant, and this downward trend was also observed at ZJG. However, since 2009, the runoff change at DJY was different from that at ZJG, and runoff in the non-flood season showed a significant upward trend.
Figure 3

M–K test results for monthly discharge in the non-flood season at the main hydrological stations in the Minjiang River Basin, China, from 1961 to 2016. Hydrological station abbreviations are defined in Figure 11.

Figure 3

M–K test results for monthly discharge in the non-flood season at the main hydrological stations in the Minjiang River Basin, China, from 1961 to 2016. Hydrological station abbreviations are defined in Figure 11.

Close modal

The discharge at SW in the non-flood season showed a slight downward trend before the 1990s and thereafter was seen to increase slowly. A large change was observed around 2010 when discharge at SW increased significantly.

The runoff change at GC in the non-flood season was affected by the superposition of runoff changes at DJY and SW and was similar to that at DJY and SW before 2010. There was a large change around 2010 when runoff in the non-flood season showed a significant upward trend.

Analysis of driving factors

Canonical correspondence analysis was conducted on the relationship between average annual runoff and runoff in the non-flood season and the main driving factors at hydrological stations in the Minjiang River Basin, and the results are shown in Figures 4 and 5, and Table 2. The results showed that the runoff-driving factor correlations were 0.874 for axis 1 and 0.596 for axis 2, indicating that the correlation coefficients for axis 1 and axis 2 can fully reflect the relationship between runoff at the hydrological stations and driving factors in the Minjiang River Basin. The correlations between axis 1 and T_SC, T_WSPU, NDVI, N_POPU, A_ICL, and N_RESE were 0.7616, 0.7147, 0.6254, 0.5654, 0.5171, and 0.3962, respectively, indicating that axis 1 mainly reflects the influence of anthropogenic disturbances and vegetation (such as total reservoir capacity, water consumption, and NDVI) on runoff. The correlation coefficient between axis 2 and T_PREC was the highest (0.3489), indicating that axis 2 mainly reflects the impact of precipitation on runoff. Along axis 1 (from left to right), the intensity of anthropogenic disturbance gradually increased and vegetation status changed. Along axis 2 (from bottom to top), precipitation gradually increased. Annual runoff values from 1961 to 2016 were divided into three groups according to the relationship between runoff changes and driving factors. Excluding some abnormal data, the first group contained the runoff data from 1961 to 1979; the second group contained the runoff data from 1980 to 2003, and the third group contained the runoff data from 2004 to 2016. The sampling sites in the first group were mainly distributed in the upper left corner of Figure 4, indicating that annual runoff from 1961 to 1979 was mainly affected by precipitation. The sampling sites in the second group were mainly distributed in the lower-left corner of Figure 4, indicating that annual runoff from 1980 to 2003 was affected by precipitation, anthropogenic disturbance, and vegetation. The sampling sites in the third group were mainly distributed on the right side of Figure 4, indicating that annual runoff from 2004 to 2016 was mainly affected by anthropogenic disturbance and vegetation.
Table 2

Correlation coefficients between environmental factors and four axes of canonical correspondence analysis

Driving FactorCorrelation coefficient with each axis
Axis 1Axis 2Axis 3Axis 4
A_ICL 0.5171 −0.2499 0.2852 0.0036 
A_AFFO 0.0786 −0.1011 0.5080 −0.0347 
N_POPU 0.5654 0.2077 0.2444 −0.0169 
T_WSPU 0.7147 −0.1109 0.1946 0.0675 
N_RESE 0.3962 −0.2464 0.2001 −0.0946 
T_SC 0.7616 −0.0115 0.1518 −0.0154 
NDVI 0.6254 0.0157 0.2087 0.1777 
T_PREC −0.3408 0.3489 0.0266 0.0809 
Runoff-driving factor correlation coefficient 0.874 0.596 0.547 0.419 
Cumulative percentage of the variance of runoff-driving factor relationship 77.5 89.8 94.6 96.9 
Driving FactorCorrelation coefficient with each axis
Axis 1Axis 2Axis 3Axis 4
A_ICL 0.5171 −0.2499 0.2852 0.0036 
A_AFFO 0.0786 −0.1011 0.5080 −0.0347 
N_POPU 0.5654 0.2077 0.2444 −0.0169 
T_WSPU 0.7147 −0.1109 0.1946 0.0675 
N_RESE 0.3962 −0.2464 0.2001 −0.0946 
T_SC 0.7616 −0.0115 0.1518 −0.0154 
NDVI 0.6254 0.0157 0.2087 0.1777 
T_PREC −0.3408 0.3489 0.0266 0.0809 
Runoff-driving factor correlation coefficient 0.874 0.596 0.547 0.419 
Cumulative percentage of the variance of runoff-driving factor relationship 77.5 89.8 94.6 96.9 

A_ICL, irrigation area; A_AFFO, afforestation area; N_POPU, population; T_WSPU, water consumption; N_RESE, number of reservoirs; T_SC, total reservoir capacity;.

NDVI, normalized difference vegetation index; T_PREC, annual precipitation.

Figure 4

Groups of each arbor plot for runoff in the Minjiang River Basin, China, and their relationships with environmental factors.

Figure 4

Groups of each arbor plot for runoff in the Minjiang River Basin, China, and their relationships with environmental factors.

Close modal
Figure 5

Relationships between runoff samples at different hydrological stations in the Minjiang River Basin, China, and environmental factors. Hydrological station abbreviations are defined in Figure 11.

Figure 5

Relationships between runoff samples at different hydrological stations in the Minjiang River Basin, China, and environmental factors. Hydrological station abbreviations are defined in Figure 11.

Close modal

The numbers 1–56 represent the years from 1961 to 2016. That is, the number 1 represents 1961, the number 2 represents 1962, and so on, and the number 56 represents 2016; A_ICL: irrigation area; A_AFFO: afforestation area; N_POPU: population; T_WSPU: water consumption; N_RESE: number of reservoirs; NDVI: normalized difference vegetation index; T_PREC: annual precipitation; T_SC: total reservoir capacity

From the distribution diagram of runoff samples and driving factors (Figure 5), it can be seen that the right part of the diagram mainly reflects the influence of human activities and vegetation types on runoff, while the left part reflects the combined influence of precipitation and human activities on runoff. The samples on the right side of the diagram are KF_GC, KF_SW, and KF_DJY (i.e., runoff samples at GC, SW, and DJY in the non-flood season). Runoff values were positively correlated with total reservoir capacity and the vegetation situation in the basin. Runoff at SW was most significantly affected by the total storage capacity, followed by runoff at GC, and finally by runoff at DJY. The Pubugou Reservoir is upstream of SW and is a control project on the Dadu River. The Zipingpu Reservoir is upstream of DJY and is a control project on the Minjiang River. GC is located downstream of the Zipingpu and Pubugou Reservoirs. Runoff in the non-flood season was also significantly affected by vegetation. On the left side of Figure 5, there are runoff samples in the non-flood season at ZJG and DJ, as well as the annual runoff samples at all stations in the basin. Because ZJG and DJ are located upstream of regulatory reservoirs, the runoff changes at the two stations were mainly affected by both precipitation and human activities.

F_ZJG: average annual runoff at ZJG; F_DJY: average annual runoff at DJY; F_GC: average annual runoff at GC; F_DJ: average annual runoff at DJ; F_LD: average annual runoff at LD; F_SW: average annual runoff at SW; KF_ZJG: runoff at ZJG in non-flood season; KF_DJY: runoff at DJY in non-flood season; KF_GC: runoff at GC in non-flood season; KF_DJ: runoff at DJ in non-flood season; KF_LD: runoff at LD in non-flood season; KF_SW: runoff at SW in non-flood season.

According to the analysis, water consumption and total reservoir capacity were the most important factors affecting runoff changes in this river basin. The influences of each factor followed the order of: water consumption (15.10%) > total reservoir capacity (13.94%) > population (13.88%) > irrigated area (13.87%) > NDVI (13.62%) > reservoir number (11.75%) > annual precipitation (10.14%) > afforestation area (7.71%) (Figure 6). Moreover, the effects of anthropogenic disturbance, vegetation, and precipitation on the runoff variation accounted for 76.24, 13.62, and 10.14%, respectively.
Figure 6

Contribution percentage of driving factors to runoff variation in the Minjiang River Basin, China. A_ICL: irrigation area; A_AFFO: afforestation area; N_POPU: population; NDVI: normalized difference vegetation index; T_WSPU: water consumption; T_PREC: annual precipitation; N_RESE: number of reservoirs; T_SC: total reservoir capacity.

Figure 6

Contribution percentage of driving factors to runoff variation in the Minjiang River Basin, China. A_ICL: irrigation area; A_AFFO: afforestation area; N_POPU: population; NDVI: normalized difference vegetation index; T_WSPU: water consumption; T_PREC: annual precipitation; N_RESE: number of reservoirs; T_SC: total reservoir capacity.

Close modal

Impacts of reservoirs on streamflow variations

Previous research has shown that the operation of reservoirs will have a significant impact on river runoff (Cui et al. 2011; He et al. 2021). Over the past 60 years, the runoff process at each hydrological station in the MRB had an obvious and regular runoff pattern in both flood and non-flood seasons. Some results showed that the runoff changes in the upper reaches of the MRB were mainly influenced by changes in vegetation, precipitation, land use, etc. However, the impact of reservoirs on runoff changes was not analyzed (Zhang et al. 2012; Hu et al. 2014; Hou et al. 2018). The large hydropower stations with regulation performance that have been built in the MRB are the Zipingpu and the Pubugou Reservoirs which began operation in 2006 and 2010, respectively. In order to further study the effect of reservoirs on runoff variation, we carried out a study at different periods based on the construction times of the hydropower stations. Based on the variation process and the changes of runoff at each station in the MRB, and taking the operation time of the Zipingpu and Pubugou Reservoirs as demarcation points, the data record was divided into two parts: 1961–2007 and 2008–2016 for ZJG and DJY, and 1961–2010 and 2011–2016 for GC and SW, in order to analyze the annual runoff variation process (Figure 7).
Figure 7

Average monthly flow distribution at (a) ZJG, (b) DJY, (c) GC, and (d) SW in the Minjiang River Basin, China, before and after 2010. Hydrological station abbreviations are defined in Figure 11.

Figure 7

Average monthly flow distribution at (a) ZJG, (b) DJY, (c) GC, and (d) SW in the Minjiang River Basin, China, before and after 2010. Hydrological station abbreviations are defined in Figure 11.

Close modal

ZJG is located upstream of the Zipingpu Reservoir, so the operation of the reservoir has no impact on runoff, and annual runoff variation at ZJG was generally consistent in the two periods. In contrast, DJY (located downstream of the reservoir) was affected by its operation, so the annual runoff processes during the two periods showed obvious differences: runoff increased significantly in the non-flood season and decreased in the flood season. The average annual runoff at DJY in the non-flood season was approximately 46.64% higher in 1961–2007 than in 2008–2016. In the flood season, the average annual runoff was approximately 8.23% lower in 1961–2007 than in 2008–2016.

The average annual runoff at SW was influenced by the operation of the Pubugou Reservoir. Compared with 1961–2010, the average annual runoff at SW in 2011–2016 increased by approximately 26.21% in the non-flood season and decreased by approximately 15.20% in the flood season.

The average annual runoff at GC was jointly influenced by the operation of the Zipingpu and Pubugou Reservoirs. Compared with 1961–2010, the average annual runoff at GC in 2011–2016 increased by approximately 36.47% in the non-flood season and decreased by approximately 11.77% in the flood season.

The Shuangjiangkou Hydropower Station located upstream of the Dadu River Basin is a control project of cascade development, with a dam height of 314 m, and serves as an annual regulation reservoir.

At present, the Shuangjiangkou Hydropower Station is still under construction and will have a future significant effect on runoff changes in the Dadu and Minjiang Rivers.

From the interannual runoff variation at each station in the non-flood season (Figure 8), it can be seen that after the Zipingpu Reservoir began operation, the discharge at DJY increased significantly from 2008 to 2016. After the operation of the Pubugou Reservoir, the discharge at SW and GC increased significantly (SW: from 616.1 to 843.2 m3/s; GC: from 1074.5 to 1357.9 m3/s, respectively), indicating that the Zipingpu and Pubugou Reservoirs with seasonal regulation capacity had significant impacts on the annual runoff distribution of the Minjiang and Dadu Rivers. In general, interannual runoff is mainly influenced by dam regulations (Zhai et al. 2016; Shi et al. 2019; Bombino et al. 2021). Our results are following the findings of He et al. (2006) and Zuo et al. (2011).
Figure 8

Variation of annual dry season flow at each station in the non-flood season at (a) DJY, (b) SW, and (c) GC in the Minjiang River Basin, China. Hydrological station abbreviations are defined in Figure 11.

Figure 8

Variation of annual dry season flow at each station in the non-flood season at (a) DJY, (b) SW, and (c) GC in the Minjiang River Basin, China. Hydrological station abbreviations are defined in Figure 11.

Close modal

Influence of water use on streamflow variation

The Minjiang River is the main water source for Chengdu City, the Chengdu Plain, and the DJY irrigation area, and is therefore the lifeblood of regional economic and social development. Approximately 82% of the water in the MRB is drawn from the Minjiang River, approximately 9% from the Dadu River, and approximately 9% from the Qingyi River. According to comprehensive planning for the MRB, the middle and lower reaches of the basin are the main water supply areas, where the water supply below DJY accounts for approximately 92% of the total water supply of the Minjiang River, and the water supply below LD accounts for approximately 79% of the total water supply of the Dadu River. As can be seen from Figure 9, water consumption from the Minjiang River in 2018 was mainly used for industry, followed by agricultural and domestic uses. The agricultural water consumption in the Dadu River Basin was larger than the industrial water consumption. The middle and lower reaches of the Minjiang River are rich in solar radiation and thermal resources and have high cultivated land reclamation rates and superior irrigation conditions. These areas are the main grain-producing areas in Sichuan Province. Additionally, there are three large-scale irrigation areas (e.g., DJY, Tongjiyan, and Yuxihe) and 29 medium-scale irrigation areas.
Figure 9

Water consumption from the Minjiang River and the Dadu River in 2018 and 2030.

Figure 9

Water consumption from the Minjiang River and the Dadu River in 2018 and 2030.

Close modal

Due to irrigation in the middle and lower reaches of the Minjiang River and economic and social development, the increasing water demand of these areas has resulted in a decrease in annual runoff at DJY and GC over the last 60 years. The water supply in the MRB will continue to increase due to economic and social development. According to comprehensive planning for the MRB, by 2030 the water supply of the Minjiang River will increase from 633,985 to 810,161 × 104 m3, and the water supply of the Dadu River will increase from 72,350 to 113,943 × 104 m3 (Figure 9). With the construction of the Tianfu New District and the expansion of the DJY irrigated area, the disparity between the water supply and demand around the Chengdu Plain has become increasingly prominent. At present, water diversion projects are planned to alleviate the disparity, including a water diversion project from the Dadu River to the Minjiang River. By 2030, the temporal and spatial water distribution in the basin will be further affected by the utilization of water resources.

Influence of precipitation and vegetation on streamflow variation

Several studies have confirmed that annual runoff is affected by not only the operation of hydropower stations but also by precipitation and vegetation, as well as water supply and demand (Syed et al. 2003; Tatsumi & Yamashiki 2015; Valverde et al. 2015; Bombelli et al. 2021). In order to verify the contribution of precipitation and water resource development and utilization to river runoff, the last 60 years were divided into periods of ‘before 2000’ and ‘after 2000’. Then the fitting analysis between average annual precipitation in the basin and average annual discharge at GC was carried out (Figure 10). It can be seen from the results that the R2 values between precipitation and runoff in 1961–2016, 1961–2000, and 2008–2016 were 0.6316, 0.6593, and 0.4670, respectively. These results showed that there was a close correlation between precipitation and runoff before 2000, but the correlation decreased significantly from 2001 to 2016. Runoff variation in the past was mainly affected by precipitation. However, in recent years, runoff variation has been increasingly affected by the development and utilization of water resources, for instance, industrial and agricultural water consumption. Some researchers who carried out studies in other regions also reached similar conclusions (Kale et al. 2018; Sönmez & Kale 2020; Wang et al. 2021; Ye et al. 2021).
Figure 10

Correlation analysis between average annual flow and precipitation at different time intervals at the GC station in the Minjiang River Basin, China.

Figure 10

Correlation analysis between average annual flow and precipitation at different time intervals at the GC station in the Minjiang River Basin, China.

Close modal
Figure 11

The Minjiang River Basin, China, and the location of the Zipingpu and Pubugou Reservoirs. Hydrological stations are Zhenjiangguan (ZJG), Dujiangyan (DJY), Gaochang (GC), Dajin (DJ), Luding (LD), and Shawan (SW).

Figure 11

The Minjiang River Basin, China, and the location of the Zipingpu and Pubugou Reservoirs. Hydrological stations are Zhenjiangguan (ZJG), Dujiangyan (DJY), Gaochang (GC), Dajin (DJ), Luding (LD), and Shawan (SW).

Close modal

Impacts of ecological restoration on streamflow variation

Continuous investigation and monitoring of the ecological water environment in the MRB and scientifical evaluation of the impact of water conservancy and hydropower project development on the ecosystem (especially the adverse impact on important sensitive targets and protected objects) are conductive to optimizing the dispatching mode of cascade stations and subsequent hydropower development schemes.

Runoff variation will have an adverse impact on the aquatic ecosystem of the basin. Based on the identification of the impact of hydropower cascade development, combined with the need for aquatic ecological protection, cascade joint dispatching should be studied and implemented to effectively reduce the impact of human activities.

The development of the MRB has produced a great demand for water. The Dadu River is rich in water resources, but the water demand is low at present. The further increase in water demand will have an adverse impact on the water resources in the MRB (Xu et al. 2019) by 2030. On the one hand, measures of prioritizing water-savings should be implemented in the MRB to improve the irrigation assurance rate and reduce industrial water consumption. On the other hand, consideration should be given to the potential diversion of water from the Dadu River which has low utilization of abundant water resources in order to optimize the allocation of water resources in the basin, build and improve the water network system in the Sichuan region, and rationally allocate and utilize water resources (Xu et al. 2021; Cheng et al. 2022).

The development and utilization rate of water resources in the MRB is high, and the ecological environment has seriously deteriorated. On the one hand, previous studies mostly focus on the upper reaches of the Minjiang River, but ignore changes of runoff in the whole basin. On the other hand, previous studies mostly focus on driving effects of single factors such as vegetation, precipitation or dams on runoff, without analyzing the contribution of each influencing factor to runoff change. In this study, runoff variations of the Minjiang River and its main tributary (the Dadu River), as well as the runoff distribution between flood and non-flood seasons over the past 60 years were analyzed. Furthermore, the impacts of reservoir operation, water consumption, precipitation, and vegetation on runoff were analyzed. Ultimately, some suggestions regarding ecological environment protection and ecological restoration were put forward. This study provides a new perspective for identifying the contribution of anthropogenic disturbance and climate change to runoff, and our results can improve the effectiveness of ecological restoration at a basin scale.

The results indicated that water resources in the MRB have gradually decreased over the past 60 years, especially in the Minjiang River. Before the 1980s, the runoff was mainly affected by precipitation and other climatic conditions. From the 1980s to 2000, both anthropogenic disturbances and climate change had significant effects on runoff, after 2000, the anthropogenic disturbance was the main factor affecting runoff. In general, the contributions of anthropogenic disturbance, vegetation, and climatic conditions to runoff change in the Minjiang River Basin were 76.24, 13.62, and 10.14% respectively. There was an obvious relationship between runoff distribution and water resource development and utilization. In the past two decades, runoff has been affected by the operation of hydropower projects and the development and utilization of water resources such as water supply and irrigation. Due to increasing water demand in the Chengdu Plain, the annual runoff of the Minjiang River has gradually decreased over the past 60 years. Therefore, it is suggested that (i) ecological environment monitoring should be strengthened, (ii) water conservation measures should be implemented, and (iii) investigations and implementation of water resource allocation projects in the MRB should be increased.

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

The authors declare there is no conflict.

Arslan
G.
,
Kale
S.
&
Sönmez
A. Y.
2020
Trend analysis and forecasting of the Gökırmak River streamflow (Turkey)
.
Oceanological and Hydrobiological Studies
49
,
230
246
.
Bombino
R.
,
Pérez-Cutillas
P.
,
D'Agostino
D.
,
Denisi
P.
,
Labate
P.
,
Martínez-Salvador
A.
,
Zema
D. A.
,
Zimbone
S. M.
&
Conesa-García
C.
2021
Comparing the hydrological response of regulated vs. unregulated forested headwaters in the Mediterranean semi-arid environment
.
Water
13
,
1275
.
Chang
J. X.
,
Zhang
H. X.
,
Wang
Y. M.
&
Zhu
Y. L.
2016
Assessing the impact of climate variability and human activities on streamflow variation
.
Hydrology and Earth System Sciences
20
,
1547
1560
.
Cheng
L.
,
Xu
Z.
,
Wang
D.
&
Cai
X.
2011
Assessing interannual variability of evapotranspiration at the catchment scale using satellite-based evapotranspiration data sets
.
Water Resources Research
47
,
W09509
.
Cheng
S.
,
Cheng
L.
,
Qin
S.
,
Zhang
L.
,
Liu
P.
,
Liu
L.
,
Xu
Z.
&
Wang
Q.
2022
Improved understanding of how catchment properties control hydrological partitioning through machine learning
.
Water Resource Research
58
,
e2021WR031412
.
Ejder
T.
,
Kale
S.
,
Acar
S.
,
Hisar
O.
&
Mutlu
F.
2016
Restricted effects of climate change on annual streamflow of Sarıçay stream (Çanakkale, Turkey)
.
Marine Science and Technology Bulletin
5
,
7
11
.
Guo
W.
,
Xu
G. H.
,
Shen
J.
&
Du
T.
2018b
Research on runoff variation trend and hydrological situation variation in Minjiang River Basin
.
Yangtze River
49
,
64
68
.
He
D. M.
,
Feng
Y.
,
Gan
S.
,
Darrin
M.
&
You
W. H.
2006
Transboundary hydrological effects of hydropower dam construction on the Lancang River
.
Chinese Science Bulletin
51
,
16
24
.
Hou
J. W.
,
Ye
A. Z.
,
You
J. J.
,
Ma
F.
&
Duan
Q. Y.
2018
An estimate of human and natural contributions to changes in water resources in the upper reaches of the Minjiang River
.
Science of the Total Environment
635
,
901
912
.
Hu
X. S.
,
Wu
C. Z.
,
Hong
W.
,
Qiu
R. Z.
,
Li
J.
&
Hong
T.
2014
Forest cover change and its drivers in the upstream area of the Minjiang River, China
.
Ecological Indicators
46
,
121
128
.
Kale
S.
,
Ejder
T.
,
Hisar
O.
&
Mutlu
F.
2016
Climate change impacts on streamflow of Karamenderes River (Çanakkale
.
Turkey). Marine Science and Technology Bulletin
5
,
1
6
.
Kale
S.
,
Hisar
O.
,
Sönmez
A. Y.
,
Mutlu
F.
&
Filho
W. L.
2018
An assessment of the effects of climate change on annual streamflow in rivers in Western Turkey
.
International Journal of Global Warming
15
,
190
211
.
Kendall
M. G.
1975
Rank Correlation Methods
.
Oxford University Press
,
Oxford
.
Li
S.
,
Liu
D. Y.
,
Wang
B. H.
,
Yu
S. H.
,
Zuo
Y. E.
&
Gu
F. T.
2016
Temporal variations characteristic of precipitation in the Three Gorges Reservoir area from 1961 to 2016
.
Journal of Water & Climate Change
13
,
1766
1775
.
Liu
J. P.
,
Xu
K. H.
,
Li
A. C.
&
Milliman
J. D.
2007
Flux and fate of Yangtze river sediment delivered to the east China Sea
.
Geomorphology
85
,
208
224
.
Liu
Q.
,
Wang
W.
,
Luo
B.
&
Wang
K.
2020
Analysis on improvement of water environment in Minjiang River Basin by ten- action plan for prevention and control of water pollution
.
Sichuan Environment
39
,
80
87
.
Mann
H. B.
1945
Nonparametric tests against trend
.
Econometrica
13
,
245
259
.
Naveed
A.
,
Wang
G. X.
,
Booij
M. J.
,
Sun
X. Y.
,
Hussain
F.
&
Nabi
G.
2022
Separation of the impact of landuse/landcover change and climate change on runoff in the upstream area of the Yangtze River, China
.
Water Resources Management
36
,
181
201
.
Ngongondo
C.
,
Xu
C. Y.
,
Gottschalk
L.
&
Alemaw
B.
2011
Evaluation of spatial and temporal characteristics of rainfall in Malawi: a case of data scarce region
.
Theoretical and Applied Climatology
106
,
79
93
.
Nie
C.
,
Ni
F. Q.
,
Deng
Y.
,
Ma
J.
&
Zhang
Y.
2020
Response of runoff to climate and land use change in Minjiang and Tuojiang River Basin
.
Journal of Water Resources and Water Engineering
31
,
110
118
.
Nyikadzino
B.
,
Chitakira
M.
&
Muchuru
S.
2020
Rainfall and runoff trend analysis in the Limpopo river basin using the Mann Kendall statistic
.
Physics and Chemistry of the Earth
117
,
102870
.
Shi
P.
,
Zhang
Y.
,
Ren
Z.
,
Yu
Y.
,
Li
P.
&
Gong
J.
2019
Land-Use changes and check dams reducing runoff and sediment yield on the loess plateau of China
.
Science of the Total Environment
664
,
984
994
.
Sönmez
A. Y.
&
Kale
S.
2020
Climate change effects on annual streamflow of Filyos River (Turkey)
.
Journal of Water and Climate Change
11
,
420
433
.
Syed
K. H.
,
Goodrich
D. C.
,
Myers
D. E.
&
Sorooshian
S.
2003
Spatial characteristics of thunderstorm rainfall fields and their relation to runoff
.
Journal of Hydrology
271
,
1
21
.
Tuset
J.
,
Vericat
D.
&
Batalla
R. J.
2016
Rainfall, runoff and sediment transport in a Mediterranean mountainous catchment
.
Science of the Total Environment
540
,
114
132
.
Valverde
P.
,
Carvalho
M.
,
Serralheiro
R.
,
Maia
R.
,
Ramos
V.
&
Oliveira
B.
2015
Climate change impacts on rainfed agriculture in the Guadiana river basin (Portugal)
.
Agricultural Water Management
150
,
35
45
.
Wei
Q.
,
Li
K. F.
,
Zhao
P. X.
,
Zhang
P.
,
Wang
K. L.
,
Liang
R. F.
&
Zheng
Y. J.
2021
Study on the influence of Tangba Navigation Power Junction Project on water environment in Minjiang River
.
Resources and Environment in the Yangtze Basin
30
,
1175
1183
.
Wu
Y. Y.
,
Fang
H. W.
,
Huang
L.
&
Ouyang
W.
2020
Changing runoff due to temperature and precipitation variations in the dammed Jinsha River
.
Journal of Hydrology
582
,
124500
.
Xu
J. J.
,
Yang
D. W.
,
Yi
Y. H.
,
Lei
Z. D.
,
Chen
J.
&
Yang
W. J.
2008
Spatial and temporal variation of runoff in the Yangtze River basin during the past 40 years
.
Quaternary International
186
,
32
42
.
Xu
G.
,
Zhang
J. X.
,
Li
P.
,
Li
Z. B.
,
Lu
K. X.
,
Wang
X. K.
,
Wang
F. C.
,
Cheng
Y. T.
&
Wang
B.
2018
Vegetation restoration projects and their influence on runoff and sediment in China
.
Ecological Indicators
95
,
233
241
.
Xu
D. D.
,
Lyon
S. W.
,
Mao
J. Q.
,
Dai
H. C.
&
Jarsjö
J.
2020
Impacts of multi-purpose reservoir construction, land-use change and climate change on runoff characteristics in the Poyang Lake basin, China
.
Journal of Hydrology
29
,
100694
.
Yaghmaei
H.
,
Sadeghi
S. H.
,
Moradi
H.
&
Gholamalifard
M.
2018
Effect of Dam operation on monthly and annual trends of flow discharge in the Qom Rood Watershed, Iran
.
Journal of Hydrology
557
,
254
264
.
Yang
L.
,
Zhao
G. J.
,
Tian
P.
,
Mu
X. M.
,
Tian
X. J.
,
Feng
J. H.
&
Bai
Y. P.
2022
Runoff changes in the major river basins of China and their responses to potential driving forces
.
Journal of Hydrology
607
,
127536
.
Zhai
H. J.
,
Hu
B.
,
Luo
X. Y.
,
Qiu
L.
,
Tang
W. J.
&
Min
J.
2016
Spatial and temporal changes in runoff and sediment loads of the Lancang River over the last 50 years
.
Agricultural Water Management
174
,
74
81
.
Zhang
Q.
,
Xu
C. Y.
,
Becker
S.
&
Jiang
T.
2006
Sediment and runoff changes in the Yangtze River basin during past 50 years
.
Journal of Hydrology
331
,
511
523
.
Zhang
M. F.
,
Wei
X. H.
,
Sun
P. S.
&
Liu
S. R.
2012
The effect of forest harvesting and climatic variability on runoff in a large watershed: the case study in the Upper Minjiang River of Yangtze River basin
.
Journal of Hydrology
464–465
,
1
11
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY-NC-ND 4.0), which permits copying and redistribution for non-commercial purposes with no derivatives, provided the original work is properly cited (http://creativecommons.org/licenses/by-nc-nd/4.0/).