Human Activities Impacts on Runoff and Ecological Flow in the Huangshui River of the Yellow River Basin, China

Abstract

This study analyzed 61 years of hydrological data from the Minhe and Xiangtang Hydrological Stations (1956–2016) to examine hydrological changes and ecological flow assurance rates in the Huangshui River Basin, China. Using the Mann–Kendall trend test, IHA/RVA method, and ecological flow calculation methods, the study revealed the following results: (1) After 1994, increased human activity in the Datong River led to a measured runoff decrease compared to natural runoff. Although human activities in the Huangshui River’s main stream were present before 1972, after 1972, these activities intensified, resulting in a more pronounced decrease in the measured runoff. (2) Ecological flow analysis indicated that the main stream of the Huangshui River and the Datong River have ecological flow assurance rates of 100% for all but a few months, where the rates are 98%. The water volume is sufficiently abundant to meet ecological water demands.

1. Introduction

Climate change and the intensification of human activities are causing a decrease in river runoff worldwide, posing significant challenges to water resource management and ecological development. Activities such as agricultural irrigation, water conservancy projects, and land-use changes can lead to substantial hydrological alterations. Scholars globally are conducting quantitative research on how these factors affect river runoff. Climate change is expected to modify the behavior of droughts within watersheds, thereby threatening drought monitoring and early warning systems [1]. For instance, Sadio et al. [2] utilized the GR2M model to assess the impact of climate change on runoff characteristics in watersheds in Senegal and Guinea. The interplay between climate change and watershed characteristics can influence runoff evolution to varying extents. Based on the Budyko hypothesis, Lv et al. [3] analyzed runoff and meteorological data from the Huangshui River Basin, concluding that while runoff showed a non-significant decreasing trend, changes in watershed characteristics were the primary contributors to the reduction in runoff. Previous research indicates that human activities are the predominant factor affecting changes in watershed runoff [4]. Li et al. [5] investigated the impact of human activities on hydrological and drought dynamics in the Xilin River Basin, revealing that these activities were the main cause of the significant reduction in runoff, accounting for 68% of the change.

Ensuring ecological flow is essential for sustaining a healthy aquatic ecosystem, especially amid changes in river hydrological conditions. Hydrological rhythms serve as a vital reference for establishing reasonable ecological flow rates. Among the various methods for calculating ecological flow, the hydrological method is the most widely used, including techniques such as the Tennant method [6] and the 7Q10 method [7]. The data needed for the Tennant method can be derived from hydrological monitoring stations without the need for on-site measurement. In cases where no monitoring station exists for the river under study, hydrological technology can be employed to obtain the necessary data quickly and efficiently. However, this method has its limitations, as it does not account for rivers with high sediment content, pronounced seasonality, significant flow variability, or the substantial influence of geometric shapes on flow.

Hydrological research encompasses a range of methods, such as wavelet analysis (WA) [8,9], hydrological change indicators, the hydrological change range method [10,11], and various hydrological models [12,13,14,15]. The Indicators of Hydrologic Alteration (IHA), introduced by Richter et al. [10] in 1996, represent the most widely adopted indicator system. To identify indicators that more precisely describe the extent of hydrological changes, Richter et al. [11] proposed the Range of Variability Approach (RVA) in 1997, building upon the IHA system. The combination of IHA and RVA offers a more systematic and comprehensive assessment of river hydrological condition changes through multi-indicator analysis. IHA is predominantly utilized in hydrological research to evaluate temporal changes in hydrological conditions due to factors such as human activities, climate change, and land transformation [16,17,18], as well as to assess the ecological response to current or past hydrological conditions in relation to specific ecological variables [19]. Kannan et al. [20] also explored the use of IHA to assess the overall health status of rivers.

The Huangshui River is the largest and most important tributary of the upper reaches of the Yellow River, and also a concentrated area for economic activities. In recent years, the level of human activity in the basin has increased, and multiple water conservancy projects have been constructed, such as the Datong River to Huangshui River Water Diversion Project and the Yellow River to Xining Water Diversion Project. In this situation, it is necessary to increase attention to the hydrological changes and ecological flow guarantee of the long sequence of the Huangshui River and Datong River to provide scientific theoretical guidance for the construction of water conservancy projects.

Therefore, this study primarily employs the IHA/RVA method and related ecological water demand methods to analyze the hydrological condition changes in the Huangshui River Basin and the ecological flow guarantee rates of the Datong River and the Huangshui River. The aim is to identify the influencing factors and offer targeted recommendations.

2. Study Area

The Huangshui River and the Datong River both originate within Qinghai Province. The Datong River, originating on Muli Mountain, flows from northwest to southeast and merges with the Huangshui River at Xiangtangxia, on the border of Gansu and Qinghai provinces. The Huangshui River, the largest tributary of the upper Yellow River, originates on Daban Mountain. Its upper reaches, known as the Bawutu River, flow from north to south before Haiyan, then gradually turn from west to east. Exiting Minhe in Qinghai, the Huangshui River meets the Datong River in Haishiwan, Gansu Province, and is collectively referred to as the Huangshui River. It eventually flows into the Yellow River near the Bapan Gorge in Gansu Province.

The Huangshui River basin has a total controlled area of 32,863 km². The watershed area monitored by the Minhe Hydrological Station is 15,342 km², and that of the Datong River at the Xiangtang Hydrological Station is 15,126 km². Together, the Minhe and Xiangtang stations control a watershed area of 30,468 km². Established by the Hydrological and Water Resources Survey Bureau of the Yellow River Commission, these stations observe various parameters including water level, flow rate, suspended sediment transport rate, sediment concentration, water temperature, and ice conditions. Since their establishment in 1950, the stations have undergone reorganization and now offer excellent control conditions for observation. The site of Huangshui River Basin is shown in Figure 1.

For this study, hydrological data from the Minhe and Xiangtang Hydrological Stations spanning 61 years, from 1956 to 2016, were selected to analyze the hydrological changes and the ecological flow assurance rate in the Huangshui River Basin. These data include measured runoff, natural runoff, and precipitation.

3. Study Methods

3.1. Mann–Kendall Mutation Test Method

The Mann–Kendall mutation test is a non-parametric hypothesis testing method used to examine trend changes in time series. This testing method detects monotonic trends (upward, downward, or no trend) in time series by comparing the size of each data point with its previous data points.

In the Mann–Kendall test, the null hypothesis H₀ is a time series (x₁, …, x_n) consisting of n independent, identically distributed samples of random variables, without a certain upward or downward trend, the alternative hypothesis H₁ is a bilateral test, and for all k, j ≤ n, and k ≠ j, the distributions of x_k and x_j are not the same. The test statistic S is calculated as follows:

$$\mathrm{S}={\displaystyle {\sum }_{\mathrm{k}=1}^{\mathrm{n}-1}}{\displaystyle {\sum }_{\mathrm{j}=\mathrm{k}+1}^{\mathrm{n}}}\mathrm{s}\mathrm{g}\mathrm{n}({\mathrm{x}}_{\mathrm{j}}-{\mathrm{x}}_{\mathrm{k}})$$

(1)

where, S represents a statistical quantity that follows a normal distribution.

$$\mathrm{s}\mathrm{g}\mathrm{n}({\mathrm{x}}_{\mathrm{j}}-{\mathrm{x}}_{\mathrm{k}})=\left\{\begin{array}{c}1,{\mathrm{x}}_{\mathrm{j}}-{\mathrm{x}}_{\mathrm{k}}>0\\ 0,{\mathrm{x}}_{\mathrm{j}}-{\mathrm{x}}_{\mathrm{k}}=0\\ -1,{\mathrm{x}}_{\mathrm{j}}-{\mathrm{x}}_{\mathrm{k}}<0\end{array}\right.$$

(2)

where, sgn(·) represents the sign function.

When n > 10, the standard normal system variable is calculated by the following formula:

$$\mathrm{Z}=\left\{\begin{array}{ll}{\displaystyle \frac{\mathrm{S}-1}{\sqrt{\mathrm{V}\mathrm{a}\mathrm{r}\left(\mathrm{S}\right)}}}& \mathrm{S}>0\\ 0& \mathrm{S}=0\\ {\displaystyle \frac{\mathrm{S}+1}{\sqrt{\mathrm{V}\mathrm{a}\mathrm{r}\left(\mathrm{S}\right)}}}& \mathrm{S}<0\end{array}\right.$$

(3)

where, Z represents the standard normal distribution statistic and Var(S) represents variance.

$$\mathrm{V}\mathrm{a}\mathrm{r}(\mathrm{S})={\displaystyle \frac{\mathrm{n}(\mathrm{n}-1)(2\mathrm{n}+5)}{18}}$$

(4)

In this way, in the bilateral trend test, if |Z| ≥ Z_1−α/2 at a given alpha confidence level, the null hypothesis is unacceptable. This shows that there is a significant upward or downward trend in the time series data at the α confidence level. The trend type is determined by the sign of the statistic Z. Z > 0 indicates an upward trend and Z < 0 indicates a downward trend. Whether this trend is significant is then determined by the magnitude of the Z value.

The steps for using the M-K mutation algorithm for mutation testing are as follows:

① For the time series x_i, construct a rank sequence r_i to represent the cumulative number of samples where x_i > x_j (1 ≤ j ≤ i), and define s_k as:

$${\mathrm{s}}_{\mathrm{k}}={\displaystyle {\sum }_{\mathrm{i}=1}^{\mathrm{k}}}{\mathrm{r}}_{\mathrm{i}}{\mathrm{r}}_{\mathrm{i}}=\left\{\begin{array}{c}1\\ 0\end{array}\right.\begin{array}{c}{\mathrm{x}}_{\mathrm{i}}>{\mathrm{x}}_{\mathrm{j}}\\ \mathrm{e}\mathrm{l}\mathrm{s}\mathrm{e}\end{array}\mathrm{j}=1,2,\dots ,\mathrm{i}$$

(5)

② Assuming the time series is randomly independent, the statistical variable UF_k is defined as [21]:

$${\mathrm{U}\mathrm{F}}_{\mathrm{k}}={\displaystyle \frac{{\mathrm{s}}_{\mathrm{k}}-\mathrm{E}\left({\mathrm{s}}_{\mathrm{k}}\right)}{\sqrt{\mathrm{V}\mathrm{a}\mathrm{r}\left({\mathrm{s}}_{\mathrm{k}}\right)}}} (\mathrm{k}=2,3,\dots ,\mathrm{n})$$

(6)

where, UF_k—standard normal distribution statistic; E(s_k)—the mean value of s_k; and $\sqrt{\mathrm{V}\mathrm{a}\mathrm{r}\left({\mathrm{s}}_{\mathrm{k}}\right)}$—the variance of s_k.

Among them, UF₁ = 0, and at a given significance level α, if |UF_k| ≥ U_α, this indicates a significant trend change in the sequence.

③ Arrange the time series x in reverse order, repeat ① and ②, and make:

$${\mathrm{U}\mathrm{B}}_{\mathrm{k}}=-{\mathrm{U}\mathrm{F}}_{\mathrm{k}} (\mathrm{k}=\mathrm{n},\mathrm{n}-1,\dots ,1)$$

(7)

Among them, UB₁ = 0. By analyzing the statistical sequences UB_k and UF_k, we can further analyze the time nodes of sequence x mutations and display the mutation regions. If UF_k > 0, it indicates that the sequence is showing an upward trend; conversely, UF_k shows a downward trend. When they exceed any critical line, this indicates a significant upward or downward trend. If the UB_k and UF_k lines intersect and the intersection point is between two critical lines, then the moment corresponding to the intersection point is the moment when the mutation begins.

3.2. Indicators of Hydrologic Alteration (IHA)

To quantitatively assess the effects of human activities on hydrological conditions, Richter et al. (1996) introduced the Indicators of Hydrologic Alteration (IHA). This comprehensive framework comprises 33 distinct indicators categorized into five ecologically significant components. These components reflect various aspects of hydrological change, including variability in flow magnitude, frequency of occurrence, timing of events, duration of hydrologic events, and the rate of change in these conditions (

Table 1. Indicators of Hydrologic Alteration and their characteristics.

IHA Parameter Groups	Characteristics	Indicators
Average monthly runoff	Runoff Time	Average monthly runoff value
Annual extreme runoff	Runoff Duration	Annual maximum and minimum average runoff (1, 3, 7, 30, 90 d) Period of river interruption Base flow coefficient
Extreme runoff occurrence time	Time	Annual maximum and minimum 1-day runoff occurrence time
Frequency and duration of high and low runoff	Runoff Frequency Duration	Number of occurrences of high and low runoff per year Average duration of high and low runoff per year
Rate and frequency of runoff changes	Rate of change Frequency	Average rate of runoff increase and decrease Number of runoff reversals

).

3.3. Range of Variability Approach (RVA)

In 1997, Richter proposed the Range of Variability Approach (RVA) based on the IHA index system. The core of this method is to construct the RVA target by adding or subtracting the standard deviation of the average value of each index before the impact, or the values of 75% and 25% of the frequency of occurrence of each index as the upper and lower limits of each index. The degree of hydrological change is evaluated by the proportion of each index value falling within the RVA target range after the impact and the difference before the impact. The calculation formula for hydrological change degree is:

$${\mathrm{D}}_{\mathrm{i}}={\displaystyle \frac{{\mathrm{N}}_{\mathrm{i}\mathrm{o}}-{\mathrm{N}}_{\mathrm{i}\mathrm{e}}}{{\mathrm{N}}_{\mathrm{i}\mathrm{e}}}}\times 100\%$$

(8)

where, N_ie = rN_T,D_i is the hydrological change degree of the i-th IHA indicator. When D_i is a positive value, this indicates that the i-th indicator falls within the RVA target after being affected by human activities, and when D_i is a negative value, this indicates that the i-th indicator falls outside the RVA target after being affected by human activities. N_io is the number of years in which the i-th indicator still falls within the RVA target after being affected by human activities. N_ie is the expected number of years for the i-th indicator to fall within the RVA target after being affected by human activities. r is the proportion of the i-th indicator falling within the RVA target before being affected by human activities. N_T is the total number of years of hydrological series affected by human activities. In order to objectively evaluate the degree of hydrological change in IHA indicators, this method divides hydrological change into three levels: 0 ≤ |D_i| < 33% is a low degree change, 33% ≤ |D_i| < 67% is a moderate change, and 67% ≤ |D_i| ≤ 100% is a high degree change.

3.4. Ecological Water Demand Method

3.4.1. Tennant Method

This method considers 10% of the annual average flow as the minimum ecological water demand of the river, 30% of the annual average flow as the optimal flow to meet the survival of aquatic organisms, and 60% to 100% as the ecological flow to maintain the original natural river ecosystem. The calculation formula is:

$${\mathrm{Q}}_{\mathrm{T}}={\sum }_{\mathrm{i}}^{12}{\mathrm{Q}}_{\mathrm{i}}\times {\mathrm{Z}}_{\mathrm{i}}$$

(9)

where, Q_T is the ecological water demand of the river channel (m³), Q_i is the average annual flow rate for the i-th month within a year (m³), and Z_i corresponds to the recommended base flow percentage (%) for the i-th month.

The Tennant method has 8 levels, and the recommended ecological environment water demand in the river is divided into the general water use period (from October to March of the following year) and the fish spawning and juvenile period (from April to September). The recommended value is based on the percentage of runoff. The recommended runoff of the Tennant method is shown in

Table 2. Recommended ecological flow by Tennant method.

Qualitative Description of Habitats	Recommended Base Flow Standard (Percentage of Annual Average Flow)
	General Water Use Period (from October to March of the Following Year)	Fish Spawning and Juvenile Period (from April to September)
Maximum	200	200
Optimum flow	60~100	60~100
Excellent	40	60
Very good	30	50
Good	20	40
Become vestigial	10	30
Poor or minimum	10	10
Extremely poor	<10	<10

.

3.4.2. Ecological Flow Assurance Rate and Evaluation Standard

The ecological flow of rivers and lakes is evaluated based on the monthly average satisfaction level. The degree of ecological flow satisfaction is defined as the ratio of the number of months where the monthly flow value is greater than the monthly ecological flow value to the total number of months corresponding to the long-term runoff year. The calculation formula is:

D_i = (T_bi/T_i) × 100%, 1 ≤ i ≤ 12

(10)

where, D_i represents the ecological flow guarantee rate for the i-th month, T_bi is the number of months that satisfy the ecological flow in the i-th month of the calculation year, and T_i is the total number of months in the i-th month of the calculation year.

The evaluation criteria for the ecological flow satisfaction index are shown in

Table 3. Evaluation criteria for ecological flow satisfaction index.

Index	Evaluation Criteria (%)
	Excellent	Good	Medium	Poor	Inferior
Ecological flow assurance rate	100	95~100	90~95	80~90	<80

.

4. Result Analysis

4.1. Hydrological Change Trends and Impact Analysis of Huangshui River and Datong River

Measured runoff refers to the amount of water that passes through a certain cross-section of a river during a certain period of time. Natural runoff refers to the amount of water that has been reverted from the measured river runoff, which generally refers to the measured runoff plus the utilization of water above the measured cross-section. A Mann–Kendall trend test analysis was conducted on the measured annual runoff data of the Huangshui River’s Minhe section and the Datong River’s Xiangtang section from 1956 to 2016 (Figure 2). At the Xiangtang section, there were four intersections between the UF and UB curves during this period, occurring in the years 2000, 2007, 2011, and 2013. These intersections indicate significant shifts in the measured runoff of the Datong River. Prior to 1973, positive UF values suggested an increasing trend in runoff. Between 1973 and 1987, negative UF values indicated a decreasing trend. From 1987 to 1995, positive UF values again pointed to an increasing trend. Post-1995, negative UF values with a strengthening decreasing trend marked a pronounced reduction in runoff.

In the Minhe section, three intersections between the UF and UB curves were observed in 1963, 1964, and 1965, indicating abrupt changes in the main stream of the Huangshui River during those years. Before 1965, positive UF values reflected an increasing trend in runoff. After 1965, negative UF values indicated a decreasing trend that intensified from 1965 to 1980 and then moderated around 1980. The decreasing trend picked up again after 1989, and by around 2004, it began to subside.

Figure 3 shows the natural and measured runoff variations for the Xiangtang and Minhe sections from 1956 to 2016. The deviation rate of measured runoff, denoted as α, is calculated using the formula α = (measured runoff − natural runoff)/natural runoff. Figure 4 shows the variations in the deviation rate.

At the Xiangtang section, prior to 1994, the absolute value of the measured runoff deviation rate remained under 1%, indicating that the river was predominantly in a natural state with minimal human influence. Between 1994 and 2016, the absolute deviation rate increased significantly, ranging from 2% to approximately 30%.

For the Minhe section, before 1972, the absolute deviation rate of measured runoff was generally under 20%. The linear regression equation y = −0.0014x + 2.6807 reflected minimal human impact during this period. However, between 1973 and 2016, the absolute deviation rate consistently increased, typically exceeding 20%, with a peak reaching 45%. The linear regression equation for this period was y = −0.0029x + 5.4149, featuring a slope twice that of the pre-1972 period, indicating a significant increase in human activity levels.

Both the Minhe and Xiangtang sections exhibited a rising trend in the absolute deviation rate of measured runoff annually, suggesting an escalating impact of human activities on the hydrological conditions of these rivers.

Therefore, the Xiangtang section was considered to be in a natural state before 1994, and after 1994, it transitioned into a phase influenced by both human activities and climate change. For the Minhe section, it can be approximated as being in a natural state stage before 1972, and after 1972, it entered a stage influenced by human activities and climate change.

Regarding the Xiangtang section, the runoff deviation rates from March to November showed a decreasing trend under the influence of human activities and climate change, with rates less than zero, compared to the natural state. However, from December to February of the following year, the deviation rate was positive, indicating an increasing trend. Except for February, the runoff in other months displayed moderate to high variability due to human activities and climate change. The degree of change in hydrological indicators in Xiangtang is shown in

Table 4. Degree of change in hydrological indicators in Xiangtang.

Indicator (Average Runoff)	1956—1993 (Approximate Natural State)			1994—2016 (Human Activities + Climate Change)
	Mean Value	RVA Target Value (m3/s)		Mean Value	Deviation Rate (%)	Di (%)
		Lower Limit	Upper Limit
January	19.52	16.95	22.30	21.29	9.07	−52.17 (M)
February	19.26	17.15	22.55	21.14	9.75	−30.43 (L)
March	25.97	22.80	28.45	25.66	−1.17	−43.48 (M)
April	54.00	45.85	61.90	46.50	−13.89	−65.22 (M)
May	99.19	75.79	118.00	62.65	−36.84	−73.91 (H)
June	131.19	90.91	168.00	91.16	−30.52	−52.17 (M)
July	213.18	174.51	241.00	172.20	−19.23	−60.87 (M)
August	200.42	152.50	258.49	182.89	−8.75	−34.78 (M)
September	171.60	116.51	199.50	167.51	−2.38	−43.48 (M)
October	92.98	71.85	107.00	80.83	−13.07	−43.48 (M)
November	45.25	38.51	52.41	43.43	−4.00	−65.22 (M)
December	26.98	23.15	29.90	30.82	14.23	−69.57 (H)

.

For the Minhe section, the runoff deviation rates from February to December were negative, reflecting a decreasing trend following an increase in human activity levels, compared to earlier stages of lower human activity. June and September showed low variability, while other months exhibited moderate variability. Consequently, it is evident that there was a significant increase in human activity levels in the main stream of the Huangshui River from January to May, July to August, and October to December after 1972. The degree of change in hydrological indicators in Minhe is shown in

Table 5. Degree of change in hydrological indicators in Minhe.

Indicator (Average Runoff)	1956—1972 (Approximate Natural State)			1973—2016 (Human Activities + Climate Change)
	Mean Value	RVA Target Value (m3/s)		Mean Value	Deviation Rate (%)	Di (%)
		Lower Limit	Upper Limit
January	24.02	19.00	28.80	26.47	10.19	−43.18 (M)
February	26.92	21.70	32.16	26.25	−2.48	−36.36 (M)
March	28.16	22.95	32.85	20.65	−26.66	−65.90 (M)
April	33.51	18.30	44.10	28.96	−13.55	−36.36 (M)
May	49.62	22.80	57.55	28.53	−42.51	−43.18 (M)
June	46.76	19.00	73.36	42.28	−9.58	−25.00 (L)
July	85.97	54.51	116.51	71.52	−16.80	−36.36 (M)
August	115.88	63.84	171.50	87.39	−24.59	−38.64 (M)
September	110.31	63.25	171.49	94.74	−14.11	−15.91 (L)
October	81.52	58.36	97.24	72.74	−10.77	−40.91 (M)
November	44.36	33.20	49.34	37.47	−15.53	−52.27 (M)
December	32.48	25.80	37.04	31.98	−1.54	−34.09 (M)

.

The observed decrease in measured runoff compared to the natural state is attributed to the combined effects of human activities and climate change. Interestingly, this analysis also noted an increase in runoff during specific months: from December to February at the Xiangtang Station and from January at the Minhe Station. The following analysis explores the reasons for these increases.

The primary source of runoff in the Huangshui River basin is atmospheric precipitation, with rainwater as the predominant source and snow as the secondary source. The year can be divided into distinct hydrological periods: the spring flood season from May to June, replenished by the melting of ice and snow in the upstream areas and rainfall; the summer and autumn flood season from July to early September, mainly supplemented by large-scale precipitation; the autumn normal water period from October to December, primarily supported by groundwater recharge and river channel storage; and the winter dry season from January to April of the following year, mainly sustained by groundwater with a small and stable water volume.

After increases in human activity levels, the months when runoff at both Xiangtang and Minhe Stations increases correspond to the river’s dry season. While not entirely excluding the influence of climate change, considering the ‘flood storage and dry discharge’ operation mode of reservoirs, it can be inferred that the increase in runoff during the dry season is likely related to the discharge practices of the upstream reservoirs relative to the hydrological stations.

The difference between the annual natural runoff and the annual measured runoff at Xiangtang Station is used to estimate the volume of water extracted by human activities in the Datong River. Similarly, the difference at Minhe Station can be used to estimate water extraction in the main stream of the Huangshui River. The changes in these values from 1956 to 2015 are depicted in Figure 5 and Figure 6.

Before 1994, human activity levels in the Datong River Basin were relatively low, resulting in small volumes of water being taken from the river, with natural runoff closely aligning with measured runoff. Since 1994, the volume of water extracted by human activities in the Datong River Basin has gradually increased, with a notable spike around 2004. Concurrently, there were significant increases in human activity levels in 1999 and 2014. The Mann–Kendall (M-K) mutation test identified breakpoints that suggest human activities as the primary cause of the sudden changes in the Datong River runoff.

The significant increases in water withdrawal at Xiangtang Station in 1999 and 2014 are attributed to specific developments. The first phase of the Datong-to-Huangshui River Diversion Project, which includes the Heiquan Reservoir and the first phase of the Huangshui North Main Canal, began construction in 1996. The Heiquan Reservoir started intercepting the river flow in 1997 and completed its main project, allowing for water storage, in 2001. The second phase of the project, featuring the Shitouxia Reservoir and the second phase of the Huangshui North Main Canal, concluded at the end of 2014 with the completion of the Shitouxia Hydropower Station. The construction and operation of these projects have significantly impacted water withdrawal, as evidenced by the abrupt increases in 1999 and 2014.

Although human activities in the main stream of the Huangshui River were present before 1972, water withdrawal from the river exhibited only slight fluctuations, remaining relatively stable. However, since 1972, there has been a gradual increase in the level of human activity. Between 1972 and 1989, the volume of water extracted through human activities saw a gradual rise. From 1989 to 2004, human activity levels escalated further, but there was a notable decrease around 1999. Post-2004, water intake has gradually decreased.

The Mann–Kendall (M-K) mutation test indicates that the changes in human activities within the Huangshui River Basin generally correspond to various breakpoints and trends identified in the analysis. These findings suggest that human activities have had a significant and evolving impact on water withdrawal patterns over the years.

The significant decrease in human activities in the Huangshui River Basin around 1999 can be attributed to several factors. The Hehuang Valley, primarily shaped by the Huangshui River, is a region of intense human activity on the Qinghai–Tibet Plateau and an early center of human activity within the Yellow River Basin. According to the Qinghai Provincial Water Resources Bulletin, irrigation water consumption for farmland in the Huangshui River Basin constitutes the largest portion of total water consumption, ranging from 45% to 55%. This is followed by water consumption for forestry, animal husbandry, fishing, and other livestock-related activities. The primary sources of irrigation water in the basin are river withdrawal and natural precipitation, with climate factors exerting a significant influence on irrigation needs.

Analyzing precipitation data, the Huangshui River Basin experienced relatively high levels of rainfall from 1997 to 1999, contrasting with both the preceding and subsequent three-year periods. In other years, higher precipitation was often an isolated event. The sustained high precipitation over three consecutive years likely led to a reduction in the amount of water extracted from rivers for agricultural irrigation in 1999. Additionally, in 1998, nationwide heavy rainfall and catastrophic floods in major river basins severely impacted farmland, which in turn affected agricultural production and led to a decrease in irrigation water usage in the following year.

The annual variation in precipitation at the Minhe Station, which provides insight into these climatic influences, is illustrated in Figure 7.

As depicted in Figure 8, the total amount of human water withdrawal at the Xiangtang and Minhe Stations indicates a gradual increase in human activity levels in the Huangshui River Basin since 1972. The activity level has remained relatively stable since 1980. However, a significant surge occurred in 1994, followed by a decline after 2007. Notably, there was a sharp increase in 2014, after which the withdrawal rate gradually decreased.

This pattern suggests that while human activity in the basin has been generally consistent since the 1980s, there have been notable fluctuations, particularly in the late 1990s and early 2000s. These changes likely reflect shifts in water management practices, agricultural demands, and possibly climate-related factors affecting water availability and demand.

4.2. Analysis of Ecological Flow Assurance Rate

The Tennant method is used to determine the ecological flow at the Xiangtang and Minhe sections by designating 30% of the natural average flow from 1956 to 2016 as the ecological flow for each section. As per the Implementation Rules of the Yellow River Water Dispatching Regulations, the minimum flow requirement for a 95% assurance rate at the Xiangtang Station of the Datong River is set at 10 m³/s. Based on this method, the annual average ecological flow at the Xiangtang Station is calculated to be 28.42 m³/s, with an annual ecological water demand of 896 million m³.

Similarly, for the Minhe section in the main stream of the Huangshui River, 30% of the average natural flow from 1956 to 2016 is used to calculate the ecological flow. With the warning flow rate of 8 m³/s as stipulated in the same regulations, the annual average ecological flow for the Minhe section is determined to be 19.98 m³/s, with an annual ecological water demand of 631 million m³.

The monthly ecological flow calculation results of the two sections are shown in

Table 6. Ecological flow rate of cross-section Unit: m³/s.

Month	Hydrology Stations
	Xiangtang	Minhe
January	10	8.32
February	10	8.65
March	10	10.13
April	17.8	22.16
May	28.87	18.57
June	37.80	20.72
July	60.87	28.43
August	58.54	31.28
September	52.19	31.14
October	29.95	31.13
November	15	18.43
December	10	10.21

. These calculations ensure that the ecological needs of the river systems are met while also considering the regulatory requirements for water flow management.

The ecological flow assurance rate for the Xiangtang section of the Datong River is 98% in January, classifying the ecological flow satisfaction as ‘good’ for that month. For all other months, the rate is 100%, indicating an ‘excellent’ ecological flow satisfaction level. Similarly, the Minhe section of the Huangshui River has ecological flow assurance rates of 98% in March, August, and September, with a ‘good’ satisfaction degree for those months, and 100% in other months, reflecting ‘excellent’ satisfaction. The specific results are shown in

Table 7. Evaluation of ecological flow assurance rate.

Month	Ecological Flow Assurance Rate (%)		Ecological Flow Satisfaction
	Datong River	Huangshui River	Datong River	Huangshui River
January	98	100	Good	Excellent
February	100	100	Excellent	Excellent
March	100	98	Excellent	Good
April	100	100	Excellent	Excellent
May	100	100	Excellent	Excellent
June	100	100	Excellent	Excellent
July	100	100	Excellent	Excellent
August	100	98	Excellent	Good
September	100	98	Excellent	Good
October	100	100	Excellent	Excellent
November	100	100	Excellent	Excellent
December	100	100	Excellent	Excellent

.

These data indicate that the Datong River and the main stream of the Huangshui River have sufficient water volumes to meet ecological water demands throughout the year, with only a slight decrease in assurance rates during specific months. The ecological health of both rivers is generally well-maintained, demonstrating a strong capacity to support their respective ecosystems.

5. Conclusions and Suggestions

Since 1994, there has been an increase in human activity levels in the Datong River, resulting in a measured runoff that has shown a decreasing trend when compared to the natural runoff. Although human activities were present in the main stream of the Huangshui River before 1972, after 1972, these activities intensified, leading to a more pronounced decrease in measured runoff compared to the natural runoff.

Ecological flow analysis indicates that the ecological flow assurance rates for both the main stream of the Huangshui River and the Datong River are 100% for all months except for a few, where the rates are 98%. This suggests that the water volume is relatively abundant and capable of meeting the ecological water demand.

The Datong-to-Huangshui River diversion Project marked significant milestones with successful water supply in the main canal by the end of 2015 and in the North and West main canals in September 2023. These achievements underscore the importance of monitoring the hydrological changes and ecological flow of the Datong and Huangshui rivers, influenced by the project, to ensure the stability and health of the river ecosystems.

Ensuring stable water supply and ecological flow assurance rates involves a combination of management strategies, technological solutions, and regulatory measures. Here are some specific measures that can be taken to maintain these aspects:

Water Allocation Management: Implementing a comprehensive water allocation plan that prioritizes ecological needs while balancing human consumption and agricultural demands.

Flow Regulation: Adjusting the operation of reservoirs and dams to maintain minimum ecological flows, especially during dry seasons. This can involve releasing water from reservoirs to supplement natural flows.

River Connectivity: Ensuring that rivers and their tributaries remain connected to facilitate the movement of aquatic species and maintain ecological integrity.

Monitoring Systems: Establishing real-time monitoring systems to track water levels, flow rates, and water quality. These data can be used to make informed decisions about water management.

Climate Resilience Planning: Developing strategies to adapt to climate change impacts, such as increased variability in precipitation and higher temperatures, which can affect the flow of rivers and water availability and quality.

Stakeholder Engagement: Involving local communities, industries, and other stakeholders in water management decisions to ensure that all perspectives are considered and that solutions are sustainable and equitable.

Regulatory Frameworks: Enforcing regulations that limit water extraction and protect water resources, ensuring that ecological flow requirements are met.

Restoration Projects: Undertake river restoration projects to improve habitat conditions, reconnect fragmented habitats to ensure the continuity of flow, and enhance the overall health of river ecosystems.

Education and Awareness: Raising public awareness about the importance of maintaining ecological flows and the role of water in supporting biodiversity and ecosystem services.

Research and Innovation: Supporting research into new technologies and methods for water management, such as artificial intelligence for predicting water needs and optimizing water allocation.

By integrating these measures, it is possible to ensure that water supply and ecological flow assurance rates remain stable, supporting both human needs and the health of river ecosystems.