Listen
graphic
View largeDownload slide

The Kuye River Basin is the main coal-producing area in northwest China, and it is also a typical arid and semi-arid area. Under the combined influence of climate change and human activities, the runoff of the basin decreases noticeably. The Mann–Kendall trend test, the Hurst index, cumulative anomaly, and wavelet analysis were used to analyze the runoff evolution characteristics of the Kuye River, and the contributions of climate change and human activities to the runoff reduction were quantitatively separated by the slope-changing ratio of the cumulative quantity. The results showed that the runoff decreased significantly from 1970 to 2020, and there was an obvious abrupt change in 1996, while the change trends in precipitation and evaporation were not significant. The contributions of climate change and human activities to runoff reduction were 22.71 and 77.29%, respectively, and human activities have become the dominant factor in runoff variations. Coal mining and the implementation of soil and water conservation measures are considered to be the main human activities contributing to the reduction in runoff. In particular, mining activities reduce groundwater recharge to rivers, which provides a profound understanding of the effect of human activities in mining areas on runoff variation.

Listen

  • The abrupt change points of runoff in the Kuye River in the past 50 years were identified and their periodic characteristics were analyzed.

  • A new method was used to quantitatively separate the contributions of climate and human activities to runoff change.

  • The results of this study provide a basis for further understanding of the effects of human activities on runoff change in mining areas.

attribution analysis, climate change, human activities, Kuye River Basin, precipitation, runoff
Listen

Along with human activities and climate change, river runoff in semi-arid and arid areas of the world has decreased significantly in recent decades (Lidén & Harlin 2000; Wang et al. 2012; Shamir et al. 2015; Ning et al. 2016). In the semi-arid and arid areas of northwest China, particularly in the coal-mining areas, the decline in runoff has been a serious threat to ecological security and water resources (Shang et al. 2016; Wu et al. 2022). Identifying the characteristics of runoff change and assessing the effects of human activities and climate change on runoff have been popular research topics in hydrology.

Climate change alters global precipitation patterns by affecting atmospheric circulation, while human activities alter the spatial and temporal distributions of water resources (Robock et al. 1993; Touchan et al. 2005). In arid and semi-arid areas, climate change and human activities are more significant for runoff change and directly affect the water resources in these areas (Chen et al. 2013; Mohammed & Scholz 2018). However, a key issue is how to quantitatively separate the contributions made by climate change and human activities to changing runoff. The main methods can be summarized into two categories: hydrological models and mathematical statistical analyses. Hydrological models provide an approximate description of the complex water cycle process and are a means of hydrological science research, which mainly include the soil water assessment tool (Zhang et al. 2012a; Lian et al. 2021), hydrologic simulation program – FORTRAN (Abdulla et al. 2009), and variable infiltration capacity (Eum et al. 2016). Mathematical statistical analyses include Budyko-type equations (Jiang et al. 2015), the double mass curve method (Zhang et al. 2012b; Luan et al. 2021), regression analysis (Miao et al. 2011; Luan et al. 2021), sensitivity analysis (Ahn & Merwade 2014), and the water balance method (Li et al. 2016b).

Based on the above research methods, many researchers have conducted extensive research on a number of river basins around the world, such as the Liaohe River (Ren et al. 2002), the Heihe River (He et al. 2008), the Shiyang River (Xue et al. 2021), the Tarim River (Ling et al. 2014), and the Nile River (Hasan et al. 2018). Their research shows that runoff is not only affected by climate but is also sensitive to human activities. However, in terms of quantitative separation methods, further research is needed. For example, the hydrological model can vividly describe the temporal and spatial changes of the hydrological cycle in the basin, but the uncertainty of the model structure and parameters also affects the accuracy of the results. Compared with model parameters, the influence of the model structure on uncertainty is more important. Therefore, it is difficult to accurately construct hydrological models with limited data (Wang et al. 2009). The quantitative evaluation method avoids the complexity of the physical model, but the noise in the long time series causes some interference to the analysis process (Sankarasubramanian et al. 2001), especially in the area where the runoff varies greatly between the dry and wet seasons. In this study, the slope-changing ratio of the cumulative quantity (SCRCQ) was used to quantitatively separate the contribution of climate and human activities to runoff reduction in high-intensity coal-mining areas. This method takes the year as the independent variable and the cumulative runoff, cumulative precipitation, and cumulative evaporation as the dependent variables, and the correlation between the year and the cumulative amount is very high. The introduction of cumulants eliminates the influence of the annual fluctuation of measured data to a certain extent.

The Kuye River Basin is the main coal-producing area in the middle reaches of the Yellow River. In recent years, due to the impact of high-intensity coal mining and climate change, runoff has been decreasing. Coal mining affects the transformation relationship between the groundwater and surface water by destroying the stratum structure and indirectly affecting the runoff. According to Li et al. (2016a) and Chi et al. (2022), the development of a water-conducting fracture zone caused by coal mining is the key factor of water resource loss. The study conducted by Lyu et al. (2014) estimated that the extraction of each ton of coal in the Kuye River Basin resulted in the destruction of approximately 2.038 m3 of water resources. It is essential to quantify the influences of human activities and climate changes on runoff in semi-arid and arid coal-mining areas.

In this study, the cumulative anomaly method (Wang et al. 2020; Li et al. 2022b) is used to determine the inflection point of hydrological and climatic series data, which effectively avoids the artificial randomness of the double accumulation curve. Morlet wavelet analysis (Li et al. 2022a) was applied to analyze the periodic variation characteristics of precipitation, evaporation, and runoff. The main objectives of this study are as follows:

  • (1) To analyze the evolution characteristics of precipitation, evaporation, and runoff time series in the Kuye River Basin over the past 50 years, to determine the abrupt turning points, and to analyze the periodic characteristics and future change trends of hydrological and climate time series in the basin.

  • (2) To quantitatively separate the effects of climate change and human activities on runoff and to explore the influence mechanism of climate change and human activities on runoff change in the coal-mining area.

Listen

Study area

Listen
One of the subbasins of the Yellow River Basin is the Kuye River Basin (109°30′–110°45′E, 38°22′–39°50′N). The total length and basin area are 242 km and 8,706 km2, respectively. The Wulanmulun River and the Beiniuchuan River are the two main tributaries of the Kuye River, with controlled drainage areas of 3,840 and 2,274 km2, respectively. The Kuye River Basin is located in the transition zone between the Mu Us Desert and the Loess Plateau, which is mainly situated in the loess hills, with high terrain in the northwest and low terrain in the southeast. The elevation ranges from 743.8 to 1,578.9 m (Figure 1). The average annual temperature and average annual precipitation are 7.9 °C and 386 mm, respectively, belonging to a typical arid and semi-arid climate. The Kuye River Basin is characterized by numerous gullies, undulating terrain, and serious soil erosion. According to a survey, the coal resource reserves in this area account for 20% of the national coal reserves. Currently, the proven coal reserves are approximately 200 billion tons, ranking among the top in the country. Due to large-scale coal mining, a series of ecological and environmental problems, such as the decline of groundwater level, the drying of spring water, and the reduction of runoff in this area, have seriously restricted the sustainable development of the local area.
Figure 1
Location of the basin and distribution of main meteorological and hydrological stations.
View largeDownload slide

Location of the basin and distribution of main meteorological and hydrological stations.

Figure 1
Location of the basin and distribution of main meteorological and hydrological stations.
View largeDownload slide

Location of the basin and distribution of main meteorological and hydrological stations.

Close modal

Data

Listen

Monthly runoff data from the Wenjiachuan hydrographic station is provided by the Hydrology Bureau of the Yellow River Conservancy Commission. Monthly precipitation data were obtained from the China Meteorological Data Sharing Service System (http://data.cma.cn/). The evaporation data were derived from the reanalysis dataset (ERA5 monthly averaged data on single levels from 1959 to the present) provided by the European Centre for Medium-Range Weather Forecasts (ECMWF). All of the data were from 1970 to 2020. The elevation data were derived from digital elevation data with a resolution of 30 m in the geospatial data cloud (https://www.gscloud.cn/).

Methodology

Listen

Mann–Kendall trend test

Listen
The Mann–Kendall trend test is a nonparametric statistical testing method that is not disturbed by a few outliers and does not require samples to follow a certain distribution. It is widely recognized as an effective approach for analyzing trends in time series data and is recommended by the World Meteorological Organization. The method has been successfully employed for the analysis of trends in precipitation, temperature, and other time series (Yue & Wang 2004). According to the Mann–Kendall trend test, the test statistic S is defined as follows:
(1)
where xi and xj are random variables, and n is the length of the selected dataset. The sign() is a sign function, which is computed as follows:
(2)
For n > 10, the Z-test statistic is as follows:
(3)
in which Var(S) is the variance of statistic S.

An increasing trend was indicated by a positive Z-value and vice versa. It is also used to test the null hypothesis H0 that there is no significant trend. A significance level of 0.05 was used to test the null hypothesis of no trend in this study (Ahn & Merwade 2014).

Hurst exponent

Listen
The rescaled range (R/S) analysis method proposed by the British hydrologist Hurst (1951) can quantify the persistence of a time series and has been widely used in meteorological, hydrological, economic, and other fields (Hurst 1951). For any positive integer h ≥ 1, and a certain meteorological and hydrological time series x(t) (t = 1,2, … , n), the mean can be defined as follows:
(4)
The cumulative deviation is calculated as follows:
(5)
The range series is calculated as follows:
(6)
The standard deviation is calculated as follows:
(7)
The rescaled range is calculated as follows:
(8)
where a is a constant and H is the Hurst exponent, which describes the autocorrelation of the stationary time series and has been widely used in hydrology, meteorology, and earthquake research. For different H-values (0 ≤ H ≤ 1): (1) for H = 0.5, the time series is completely independent, and its change is random; (2) for 0.5 < H < 1, the time series has a long-term correlation and exhibits persistence; and (3) 0 < H < 0.5 indicates anti-sustainability. In Table 1, we divided the Hurst index into five grades and quantitatively described the strength of future trends (Liu et al. 2016).
Table 1

Classification of the H-values

GradeH-valuesSustainabilityGradeH-valuesAnti-sustainability
1 0.5 < H ≤ 0.55 Weak −1 0.45 ≤ H < 0.50 Weak 
2 0.55 < H ≤ 0.65 Weaker −2 0.35 ≤ H < 0.45 Weaker 
3 0.65 < H ≤ 0.75 Stronger −3 0.25 ≤ H < 0.35 Stronger 
4 0.75 < H ≤ 0.80 Strong −4 0.20 ≤ H < 0.25 Strong 
5 0.80 < H ≤ 1.00 Very strong −5 0.00 ≤ H < 0.20 Very strong 
GradeH-valuesSustainabilityGradeH-valuesAnti-sustainability
1 0.5 < H ≤ 0.55 Weak −1 0.45 ≤ H < 0.50 Weak 
2 0.55 < H ≤ 0.65 Weaker −2 0.35 ≤ H < 0.45 Weaker 
3 0.65 < H ≤ 0.75 Stronger −3 0.25 ≤ H < 0.35 Stronger 
4 0.75 < H ≤ 0.80 Strong −4 0.20 ≤ H < 0.25 Strong 
5 0.80 < H ≤ 1.00 Very strong −5 0.00 ≤ H < 0.20 Very strong 
View Large

Cumulative anomaly

Listen
The cumulative anomaly method is commonly employed to intuitively assess the changing trends of runoff, precipitation, evaporation, and other factors based on the cumulative curve of the anomaly. The cumulative anomaly (Xt) is expressed as follows for a discrete series xi:
(9)
(10)
where is the mean value of the time series xi. The cumulative anomaly curve effectively captures the variations in amplitude and potential turning points of hydrological and meteorological series data (Wang et al. 2012).

Morlet wavelet analysis

Listen

Wavelet analysis is often used to reveal the periodicity of meteorological and hydrological processes. The wavelet functions mainly include the Mexican hat wavelet, the discrete Meyer wavelet, and the Morlet wavelet (Guo et al. 2022). Compared with the real-form wavelet, the Morlet wavelet, as a wavelet with a complex form, can effectively eliminate the oscillation of the coefficient modes in the wavelet transform process. In this study, the Morlet wavelet function was used to transform the series data of the runoff, precipitation, and evaporation. Based on the wavelet transform, continuous wavelet power spectrum (WPS) is used to analyze the periodicity and significance of the variable.

For a time series with the sampling frequency dt (xn, n = 1, 2, 3 … N), taking φ as the wavelet function, the continuous wavelet transform can be expressed as the convolution of time series xn with the wavelet scale s and the normalized wavelet function (Grinsted et al. 2004):
(11)
(12)
where is the complex conjugate of φ, and the WPS is expressed as ⁠, where j is the scale level (Torrence & Compo 1998).
The statistical significance of the wavelet power can be assessed by comparing it with the null hypothesis, which holds that the signal is generated by a static process with a given background power spectrum (Pk). Time series with red noise characteristics can be modeled by a first-order autoregressive (AR1) process. The Fourier power spectrum of AR1 is as follows (Allen & Smith 1996):
(13)
In order to explore the correlation between two periodically changing time series signals x and y on different time scales, two time series wavelet coherence functions need to be solved, which are defined as follows (Rhif et al. 2019):
(14)
where W(x) and W(y), respectively, represent the continuous wavelet transform of two time series, and the product of the two means the energy distribution of each signal in each frequency band. W(x, y) is a cross wavelet transform of two time series, reflecting the degree of linear dependence between them on each frequency band. S is a smoothing operator that can reduce noise and enhance coherence by averaging local regions in the time–frequency domain. s is the wavelet scale. Similar to traditional coherence, the wavelet coherence coefficient is between 0 and 1, and the closer to 1, the stronger the correlation.
Furthermore, the partial wavelet coherence (PWC) function can indicate the wavelet coherence between time series y and x after excluding the influence of another time series z on y (Mihanović et al. 2009):
(15)
where
(16)

Slope-changing ratio of the cumulative quantity

Listen
Wang et al. (2012) first proposed the SCRCQ, which has been successfully applied in the Yellow River Basin for quantitative analysis of runoff change. We assumed that the slopes of linear relationships between the year and cumulative runoff before and after a turning year were, respectively, SRb and SRa (108 m3/a). The linear relationships' slopes between the year and cumulative precipitation before and after a turning year, respectively, were SPb and SPa (mm/year). In addition, the linear relationships' slopes between the year and cumulative evaporation before and after a turning year, respectively, were SEb and SEa (mm/year). The changing ratios of the precipitation (Rp), evaporation (RE), and runoff (RR) can be expressed as follows:
(17)
(18)
(19)
The contributions of evaporation (CE, unit: %) and precipitation (CP, unit: %) to changes in runoff can be expressed as follows:
(20)
(21)
The contribution of human activities (CH, unit: %) to runoff change, on the basis of the water balance, can be expressed as follows:
(22)
where CG is groundwater contribution to change in runoff.

On a long time scale, the groundwater level in the basin will not increase or decrease obviously under natural conditions. The exploitation of coal resources in the basin causes the groundwater level to drop continuously, which leads to a decrease in the groundwater recharge to the rivers. This influence can be considered as the influence of human activities on runoff. Therefore, in this study, CG was set to 0.

Listen

Evolution characteristics of the runoff, precipitation, and evaporation

Listen

Interannual variation and future trends

Listen
The runoff displayed a decreasing trend on the 51-year time scale, as depicted in Figure 2. The maximum and minimum runoffs recorded were 10.19 × 108 m3 in 1979 and 1.23 × 108 m3 in 2011, respectively. Since 2000, the runoff has decreased significantly, but the precipitation and evaporation have not changed significantly. The precipitation and evaporation exhibited a consistent pattern, with the maximum values recorded in 1995 (750.35 mm for precipitation) and 1990 (538.59 mm for evaporation), respectively. Conversely, the minimum values were observed in 1974 (272.54 mm for precipitation) and 2000 (281.28 mm for evaporation), respectively.
Figure 2
Interannual variation characteristics of the runoff, precipitation, and evaporation.
View largeDownload slide

Interannual variation characteristics of the runoff, precipitation, and evaporation.

Figure 2
Interannual variation characteristics of the runoff, precipitation, and evaporation.
View largeDownload slide

Interannual variation characteristics of the runoff, precipitation, and evaporation.

Close modal

The Mann–Kendall trend test results indicated that the Z-values for runoff, precipitation, and evaporation were −5.08, −0.39, and −0.54, respectively (Table 2). Runoff showed a significant decreasing trend with a Z-value passing the significance test at α = 0.05, while precipitation and evaporation exhibited nonsignificant decreasing trends.

Table 2

The Z-values of runoff, precipitation, and evaporation according to the Mann–Kendall trend test, 1970–2020

ElementsZTrendPasses the significance test (α = 0.05)
Runoff (108m3) −5.08 Decreasing Yes 
Precipitation (mm) −0.39 Decreasing No 
Evaporation (mm) −0.54 Decreasing No 
ElementsZTrendPasses the significance test (α = 0.05)
Runoff (108m3) −5.08 Decreasing Yes 
Precipitation (mm) −0.39 Decreasing No 
Evaporation (mm) −0.54 Decreasing No 
View Large

The Hurst exponent was used to assess the sustainability of meteorological and hydrological time series, and the fitting curves of R/S − τ were plotted (Figure 3). The exponential fitting relation reveals that the H-values for runoff, precipitation, and evaporation are 0.32, 0.27, and 0.38, respectively, all of which fall below the threshold value of 0.5. This outcome implies an inverse future trajectory for runoff, precipitation, and evaporation in comparison to historical patterns. As a result, runoff, precipitation, and evaporation are likely to increase in the future. Table 1 shows that the continuous intensities of precipitation and runoff were stronger, but that of evaporation was weaker according to Hurst's classification.
Figure 3
The fitting curves of R/S analysis for runoff (a), precipitation (b), and evaporation (c).
View largeDownload slide

The fitting curves of R/S analysis for runoff (a), precipitation (b), and evaporation (c).

Figure 3
The fitting curves of R/S analysis for runoff (a), precipitation (b), and evaporation (c).
View largeDownload slide

The fitting curves of R/S analysis for runoff (a), precipitation (b), and evaporation (c).

Close modal

Mutation point identification

Listen
The cumulative anomaly curves for runoff, precipitation, and evaporation in the Kuye River Basin during the period 1970–2020 are depicted in Figure 4. The runoff anomaly value exhibited an initial increase before 1996, followed by a rapid decrease, indicating a pronounced shift in the runoff dynamics during that year. The occurrence of an abrupt change in the runoff in 1996 can be inferred. The precipitation and evaporation cumulative anomaly curves exhibited similar variation patterns, with two notable inflection points observed in 1996 and 1998. The abrupt change in the time of evaporation is 1–2 years behind that of runoff and precipitation.
Figure 4
The cumulative anomaly curve of runoff, precipitation, and evaporation.
View largeDownload slide

The cumulative anomaly curve of runoff, precipitation, and evaporation.

Figure 4
The cumulative anomaly curve of runoff, precipitation, and evaporation.
View largeDownload slide

The cumulative anomaly curve of runoff, precipitation, and evaporation.

Close modal

Periodic characteristics

Listen
The Morlet wavelet analysis method was applied to illustrate the periodic variation characteristics of the precipitation, evaporation, and runoff. Based on the results of the Morlet wavelet analysis, the WPS and global wavelet spectrum were plotted (Figure 5). The runoff time series shows discontinuous high energy in the period of 2–4 years. Specifically, the energy signal of runoff is mainly concentrated from 1970 to 1992. After the 1990s, the energy signal of runoff weakened, and there was no significant periodic variation in the confidence interval (Figure 5(a) and 5(b)). During 1981–2002, the precipitation time series shows high energy in the period of 2–4 years, while, in 1970–2020, its periodicity is more significant and continuity is better in a 6-year period (Figure 5(c) and 5(d)). The evaporation time series shows significant periodicity only in the local period (after 1980) (Figure 5(e) and 5(f)). As a result, the periodicity of runoff does not coincide perfectly with that of precipitation and evaporation over the entire time scale. The difference in the periodicity of the runoff time series and the precipitation and evaporation time series also indicates that runoff is not completely controlled by meteorological conditions, which makes the change in the runoff more complicated.
Figure 5
WPS and global wavelet spectrum for runoff (a, b), precipitation (c, d), and evaporation (e, f).
View largeDownload slide

WPS and global wavelet spectrum for runoff (a, b), precipitation (c, d), and evaporation (e, f).

Figure 5
WPS and global wavelet spectrum for runoff (a, b), precipitation (c, d), and evaporation (e, f).
View largeDownload slide

WPS and global wavelet spectrum for runoff (a, b), precipitation (c, d), and evaporation (e, f).

Close modal
In order to study whether the periodic change of runoff is related to the periodic change of evaporation and precipitation, the wavelet coherence function and PWC function of evaporation and runoff, precipitation, and runoff are calculated. The results show that, in the period of 4–8 years, evaporation and runoff are coherent, but the coherence is unstable (Figure 6(a)). Even after the influence of precipitation on runoff is excluded, the coherence between evaporation and runoff is still discontinuous (Figure 6(c)). However, in the period of 1–6 and 6–12 years, the coherence between precipitation and runoff is relatively good, with a coherence coefficient greater than 0.7 and a confidence level higher than 95% (Figure 6(b)). Especially after the interference of evaporation on runoff is excluded, the coherence between precipitation and runoff becomes higher and more continuous (Figure 6(d)). The above shows that the variation of runoff is affected by the periodic variation of precipitation, but it presents the characteristics of non-periodic dynamic variation due to the influence of other factors such as evaporation or humans. The periodic changes of climate may be caused by periodic changes such as solar activity and the El Niño/Southern Oscillation (Huang & Xie 2015).
Figure 6
Wavelet coherence between evaporation and runoff (a), precipitation and runoff (b), partial wavelet coherence between evaporation and runoff (c), and precipitation and runoff (d).
View largeDownload slide

Wavelet coherence between evaporation and runoff (a), precipitation and runoff (b), partial wavelet coherence between evaporation and runoff (c), and precipitation and runoff (d).

Figure 6
Wavelet coherence between evaporation and runoff (a), precipitation and runoff (b), partial wavelet coherence between evaporation and runoff (c), and precipitation and runoff (d).
View largeDownload slide

Wavelet coherence between evaporation and runoff (a), precipitation and runoff (b), partial wavelet coherence between evaporation and runoff (c), and precipitation and runoff (d).

Close modal

Effects of climate variability and human activities on runoff

Listen
The cumulative anomaly curve of runoff (Figure 4) reveals distinct stage variations in the Kuye River Basin on a 51-year time scale. Therefore, we considered 1996 as the dividing point between the base and variation periods. Specifically, the base period spanned from 1970 to 1996 (T1R), while the variation period covered the years from 1997 to 2020 (T2R). The runoff of the Kuye River is primarily influenced by climate change during the base period. However, during the variation period, as the intensity of human activities increases, both climate change and human activities have an impact on the changes in runoff. The distribution of data points and the linear relationship between the year and the accumulation of runoff, precipitation, and evaporation are depicted in Figure 7. The relationships between the year and each factor are highly significant, with R-values of >0.99. In addition, the confidence level of the P-values is <0.0001.
Figure 7
Relationships between the year and accumulation of runoff (a), precipitation (b), and evaporation (c).
View largeDownload slide

Relationships between the year and accumulation of runoff (a), precipitation (b), and evaporation (c).

Figure 7
Relationships between the year and accumulation of runoff (a), precipitation (b), and evaporation (c).
View largeDownload slide

Relationships between the year and accumulation of runoff (a), precipitation (b), and evaporation (c).

Close modal

The slopes (SR) of the relationship between the year and cumulative runoff during periods T1R (1970–1996) and T2R (1997–2020) were 5.66 × 108 and 2.23 × 108 m3/a, respectively (Table 2). Compared with the base period (T1R), the slope of the linear relationship between the year and cumulative runoff in the variation period (T2R) was 3.43 × 108 m3/a (60.60%) lower (Table 3). Similarly, compared to T1R, the slope of the linear relationship between the year and cumulative precipitation was 28.86 mm/a (5.36%) lower, and the slope of the linear relationship between the year and cumulative evaporation was 37.60 mm/a (8.40%) lower. Assuming that there is no influence of human activities in period T2R, the rates of change of the runoff and meteorological elements (precipitation and evaporation) should be the same, and the CP + CE should be 100%. According to the calculation results of Equation (22), in comparison with period T1R, precipitation and evaporation contribute to a 22.71% reduction in the runoff, while human activities account for 77.29% of the runoff reduction.

Table 3

The slopes and rates of change of the runoff, precipitation, and evaporation in different periods

PeriodSlopes
Rates of change (%)
SR (108 m3/year)SP (mm/year)SE (mm/year)SRSPSE
T1R (1970–1996) 5.66 537.96 447.65 − 60.60 − 5.36 − 8.40 
T2R (1997–2000) 2.23 509.1 410.05 
PeriodSlopes
Rates of change (%)
SR (108 m3/year)SP (mm/year)SE (mm/year)SRSPSE
T1R (1970–1996) 5.66 537.96 447.65 − 60.60 − 5.36 − 8.40 
T2R (1997–2000) 2.23 509.1 410.05 
View Large

Listen

The change in runoff is influenced by a multitude of factors, encompassing both natural and anthropogenic influences. In terms of interannual patterns of runoff, precipitation, and evaporation, it can be seen that in the past 51 years, the precipitation and evaporation in the Kuye River Basin have been relatively stable, and they have not exhibited significant increasing or decreasing trends, while the runoff has continuously decreased, especially after 1996. This difference between climate factors and runoff demonstrated that climate change is not solely responsible for runoff changes in the Kuye River Basin. According to the periodic characteristics of the runoff, precipitation, and evaporation, the periodic change in the runoff on the 11–15-year time scale was relatively stable before the 1990s. However, the periodic change was not obvious after the 1990s, which was speculated to be closely linked to increased human activity after the 1990s (Li et al. 2022a, 2022b). Gu et al. (2014) analyzed the runoff characteristics in the Kuye River Basin from 1956 to 2012 and found that the vegetation, soil, slope, and land use status in the area have been significantly altered due to coal-mining activities and implemented measures for soil and water conservation, resulting in a substantial impact on runoff. Li et al. (2020) conducted a systematic study of the causes of water and sediment changes in the Kuye River Basin, and their results revealed that human activities, mainly vegetation restoration and coal mining, were the important reasons for the reductions in the runoff and sediment transport in the basin.

Influence of coal mining on water resources

Listen
The Kuye River Basin, which is characterized by arid and semi-arid conditions, serves as the principal coal-mining region in China. Subsequent to 1996, extensive coal extraction activities have been undertaken within this basin. The annual coal production in the Kuye River Basin has experienced a rapid increase since 1996, surpassing 2.50 × 108 t/a by 2012. However, with the introduction of the policy on coal resource integration, the annual coal production has gradually decreased. After the end of coal resource integration in 2016, coal production exhibited a significant increasing trend, and since 2018, coal production has been close to 3.50 × 108t/a (Figure 8) (Lei 2021). The coal seam in the Kuye River Basin has a shallow burial depth (approximately 100 m), and the caving zone and fracture zone generated by coal mining can easily extend to the bottom of the quaternary aquifer and even to the surface. Surface water resources leak due to ground cracks, leading to a reduction in runoff. In addition, in the process of coal mining, a large amount of groundwater is drained, resulting in a reduction of groundwater resources. According to statistics, in 2010, the mine drainage in the Kuye River Basin was close to 0.5 × 108 m3 (Wu et al. 2002). Coal mining also caused the groundwater level to decrease and altered the mutual transformation relationship between groundwater and surface water, resulting in a decrease in the groundwater recharge to the Kuye River. Before 1995, the Muhegou spring in the upper reaches of the Kuye River maintained a flow rate of 0.05 m3/s. With increasing coal-mining intensity, the flow rate decreased to 0.01 m3/s in 2000, ultimately ceasing completely in 2002. The large-scale exploitation of coal resources has seriously damaged the natural occurrence conditions of the groundwater resources and has caused irreversible impacts on the hydrogeological environment in the basin.
Figure 8
Plot of the runoff and coal production in the Kuye River Basin, 1980–2020.
View largeDownload slide

Plot of the runoff and coal production in the Kuye River Basin, 1980–2020.

Figure 8
Plot of the runoff and coal production in the Kuye River Basin, 1980–2020.
View largeDownload slide

Plot of the runoff and coal production in the Kuye River Basin, 1980–2020.

Close modal

According to Lyu et al. (2014), 2.038 m3 of groundwater is consumed per ton of coal produced in this area. Based on this calculation, the consumption of groundwater resources for coal mining in the Kuye River Basin in 2018 was approximately 7.13 × 108 m3. Only a small amount of this water is available for treatment and reuse, and most of it is difficult to recover.

Influence of soil and water conservation measures on runoff

Listen

Since the 1980s, a series of soil and water conservation measures have been implemented in the Kuye River Basin, and the area of forest and grassland has been increased to a certain extent. By 1998, the policy of returning farmland to forest and grassland was implemented on a large scale. Since then, the Kuye River Basin's cultivated land, forestland, and grassland have changed significantly. In 2010, the soil erosion control area in the Kuye River Basin measured 3,739.90 km2, with a total of 1,548 silting dams having been constructed. Soil and water conservation measures in the Kuye River Basin primarily encompass terraced fields, afforestation, grass planting, and silting dam construction (Zhao et al. 2010), all of which influence the spatiotemporal variations in runoff by altering the characteristics of the underlying surface within the basin. Zhao et al. (2010) conducted a calculation on the water storage capacity of various soil and water conservation measures implemented in the Kuye River Basin during the 1980s. The afforestation, grass planting, and terraced fields were found to have respective water storage capacities of 1953.2 × 104, 394.21 × 104, and 134.4 × 104 m3. The implementation of soil and water conservation measures in the basin has significantly mitigated sediment production, albeit at the expense of reduced water resources.

Comparison with previous research results

Listen

In recent years, the majority of studies on runoff changes in the Kuye River Basin have primarily focused on qualitative analysis of the factors influencing reductions in runoff, with relatively limited research conducted on quantifying the contribution rates of these factors affecting runoff alterations. Zhang et al. (2011) employed an improved multiple regression model to quantitatively assess the impact of human activities on decreased runoff in the Wulanmulun River Basin located in the upper reaches of the Kuye River Basin. The results showed that from 1961 to 2007, the influence of human activities on runoff reached 89.5%. The elastic coefficient method and the double accumulation curve method were employed by Li et al. (2020) to quantitatively assess the respective contributions of climate change and human activities to the observed decline in runoff within the Kuye River Basin. The results indicated that between 1997 and 2015, they achieved growth rates of 18.53 and 81.47%, respectively. Our research results were basically consistent with those of Zhang et al. (2011) and Li et al. (2020), showing that the influence of human activities on runoff accounts for about 80%, and the small differences are mainly caused by different research methods and time and spatial scales. The research results of Zhang et al. (2011) showed that human activities contributed more to the reduction of runoff. This is related to subjectivity in determining variable weights in multiple regression models and the selection of the study area. Li et al. (2020) considered the period from 1997 to 2015 as the variation period, and their calculated contribution of human activities to runoff reduction was slightly greater than that obtained in this study. Considering the trend of precipitation change since 2011, there has been a slight increase in runoff with increasing precipitation, and climate factors have been relatively enhanced, while human activities have been relatively weaker.

Listen

In this study, the Kuye River Basin in the arid and semi-arid area of northwest China was taken as the research area, and the Mann–Kendall trend test, Hurst exponent, cumulative anomaly, and wavelet analysis methods were applied to analyze the variation trend, abrupt change characteristics, and periodic characteristics of the meteorological and hydrological elements from 1970 to 2020. The contributions of climate change and human activities to runoff reduction were quantitatively assessed using the SCRCQ method. The conclusions of this study are summarized below:

  • (1) From 1970 to 2020, the runoff in the Kuye River Basin decreased significantly, and an obvious abrupt change occurred in 1996. In contrast, the variation trends of precipitation and evaporation were not significant. The H-values for runoff, precipitation, and evaporation are 0.32, 0.27, and 0.38, respectively. They are all below the threshold value of 0.5. As a result, an increase in runoff, precipitation, and evaporation is likely in the future.

  • (2) Although the runoff variation is affected by precipitation, there is a significant difference between the two periodic variations, which indicates that human activities have a significant impact on the dynamic variation of runoff in addition to meteorological factors.

  • (3) In the Kuye River Basin, the contributions of climate change and human activities to the runoff reduction were 22.71 and 77.29%, respectively, and the contribution of human activities to the runoff reduction was much greater than that of climate change. Coal mining and the implementation of soil and water conservation measures were the main human activities contributing to the decrease in runoff.

  • (4) This study quantitatively separated the contribution of climate change and human activities to runoff reduction in coal-mining areas. However, human activities include coal mining, soil and water conservation measures, and changes in land use types. Future studies need to distinguish, in detail, the influence processes of different human activities on runoff to promote the protection of water resources.

Listen

This research was financially supported by the National Natural Science Foundation of China (41831289) and the Natural Science Foundation of Hebei Province (E2021403001).

Listen

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

Listen

The authors declare there is no conflict.

Abdulla
F.
,
Eshtawi
T.
 &
Assaf
H.
2009
Assessment of the impact of potential climate change on the water balance of a semi-arid watershed
.
Water Resources Management
23
(
10
),
2051
–
2068
.
Google Scholar
Crossref
Search ADS
 
Chi
M.
,
Cao
Z.
,
Wu
B.
,
Li
Q.
,
Zhang
Y.
,
Wu
Y.
,
Wang
L.
,
Yang
Y.
,
Zhang
B.
 &
Li
H.
2022
Prediction of water source and water volume of underground reservoir in coal mine under multiple aquifers
.
Water Supply
22
(
4
),
4067
–
4081
.
Google Scholar
Crossref
Search ADS
 
Eum
H. I.
,
Dibike
Y.
 &
Prowse
T.
2016
Comparative evaluation of the effects of climate and land-cover changes on hydrologic responses of the Muskeg River, Alberta, Canada
.
Journal of Hydrology: Regional Studies
8
,
198
–
221
.
Google Scholar
 
Grinsted
A.
,
Moore
J. C.
 &
Jevrejeva
S.
2004
Application of the cross wavelet transform and wavelet coherence to geophysical time series
.
Nonlinear Processes in Geophysics
11
(
5/6
),
561
–
566
.
Google Scholar
Crossref
Search ADS
 
Gu
X.
,
Chen
L.
 &
Dong
G.
2014
Study on hydrological regime changes during the past 50 years in Kuye River
.
Yellow River
36
(
12
),
25
–
27
(in Chinese)
.
Google Scholar
 
Hasan
E.
,
Tarhule
A.
,
Kirstetter
P. E.
,
Clark
R.
 &
Hong
Y.
2018
Runoff sensitivity to climate change in the Nile River Basin
.
Journal of Hydrology
561
,
312
–
321
.
Google Scholar
Crossref
Search ADS
 
He
H.
,
Zhou
J.
 &
Zhang
W.
2008
Modelling the impacts of environmental changes on hydrological regimes in the Hei River Watershed, China
.
Global and Planetary Change
61
(
3
),
175
–
193
.
Google Scholar
 
Hurst
H. E.
1951
Long-term storage capacity of reservoirs
.
Transactions of the American Society of Civil Engineers
116
(
1
),
770
–
799
.
Google Scholar
Crossref
Search ADS
 
Lei
Y.
2021
Study on the Influence of Coal Mining on Water Resources in Kuye River Basin of Shaanxi Province
.
MS Thesis
,
Northwest University
,
China
(in Chinese)
.
Google Scholar
 
Li
S.
,
Xiong
L.
,
Li
H. Y.
,
Leung
L. R.
 &
Demissie
Y.
2016b
Attributing runoff changes to climate variability and human activities: Uncertainty analysis using four monthly water balance models
.
Stochastic Environmental Research and Risk Assessment
30
(
1
),
251
–
269
.
Google Scholar
Crossref
Search ADS
 
Li
H.
,
Shi
C.
,
Ma
X.
 &
Liu
W.
2020
Quantification of the influencing factors of runoff and sediment discharge changes of the Kuye River catchment in the middle reaches of the Yellow River
.
Resources Science
42
(
3
),
499
–
507
(in Chinese)
.
Google Scholar
 
Li
L.
,
Wu
X.
,
Zhou
X.
,
Dong
Y.
,
Zhang
F.
,
Zhang
Y.
 &
Liu
J.
2022a
Evolution characteristics of river runoff in coal mining area based on Morlet wavelet theory
.
Water Resources and Power
40
(
4
),
37
–
40
(in Chinese)
.
Google Scholar
 
Li
Z.
,
Sun
Z.
,
Sun
L.
,
Liu
J.
,
Xiong
W.
,
Dong
H.
 &
Zheng
H.
2022b
Hydrological variation and hydro-sediment interrelation of the Luozha River in the Lancang River Basin
.
Water Supply
22
(
5
),
4839
–
4851
.
Google Scholar
Crossref
Search ADS
 
Lian
Y.
,
Sun
M.
,
Wang
J.
,
Luan
Q.
,
Jiao
M.
,
Zhao
X.
 &
Gao
X.
2021
Quantitative impacts of climate change and human activities on the runoff evolution process in the Yanhe River Basin
.
Physics and Chemistry of the Earth, Parts A/B/C
122
,
102998
.
Google Scholar
Crossref
Search ADS
 
Lidén
R.
 &
Harlin
J.
2000
Analysis of conceptual rainfall-runoff modelling performance in different climates
.
Journal of Hydrology
238
(
3
),
231
–
247
.
Google Scholar
 
Liu
Y.
,
Yuan
Z.
 &
Sun
B.
2016
R/S analysis of major climatic factors in Xi'an in recent 15 years
.
Journal of Xi'an University(Natural Science Edition)
19
(
1
),
73
–
78
.
Google Scholar
 
Lyu
X.
,
Wang
S.
,
Yang
Z.
,
Bian
H.
 &
Liu
Y.
2014
Influence of coal mining on water resources: A case study in Kuye River Basin
.
Coal Geology & Exploration
42
(
2
),
54
–
57
(in Chinese)
.
Google Scholar
 
Miao
C.
,
Ni
J.
,
Borthwick
A. G. L.
 &
Yang
L.
2011
A preliminary estimate of human and natural contributions to the changes in water discharge and sediment load in the Yellow River
.
Global and Planetary Change
76
(
3
),
196
–
205
.
Google Scholar
 
Mihanović
H.
,
Orlić
M.
 &
Pasrić
Z.
2009
Diurnal thermocline oscillations driven by tidal flow around an island in the Middle Adriatic
.
Journal of Marine Systems
78
,
S157
–
S168
.
Google Scholar
Crossref
Search ADS
 
Ren
L.
,
Wang
M.
 &
Zhang
W.
2002
Impacts of human activity on river runoff in the northern area of China
.
Journal of Hydrology
261
(
1
),
204
–
217
.
Google Scholar
 
Rhif
M.
,
Ben Abbes
A.
,
Farah
I. R.
,
Martínez
B.
 &
Sang
Y.
2019
Wavelet transform application for/in non-stationary time-series analysis: A review
.
Applied Sciences
9
(
7
),
1345
.
Google Scholar
Crossref
Search ADS
 
Robock
A.
,
Turco
R. P.
,
Harwell
M. A.
,
Ackerman
T. P.
,
Andressen
R.
,
Chang
H. S.
 &
Sivakumar
M. V. K.
1993
Use of general circulation model output in the creation of climate change scenarios for impact analysis
.
Climatic Change
23
(
4
),
293
–
335
.
Google Scholar
Crossref
Search ADS
 
Sankarasubramanian
A.
,
Vogel
R. M.
 &
Limbrunner
J. F.
2001
Climate elasticity of streamflow in the United States
.
Water Resources Research
37
(
6
),
1771
–
1781
.
Google Scholar
Crossref
Search ADS
 
Shamir
E.
,
Megdal
S. B.
,
Carrillo
C.
,
Castro
C. L.
,
Chang
H. I.
,
Chief
K.
,
Corkhill
F. E.
,
Eden
S.
,
Georgakakos
K. P.
,
Nelson
K. M.
 &
Prietto
J.
2015
Climate change and water resources management in the Upper Santa Cruz River, Arizona
.
Journal of Hydrology
521
,
18
–
33
.
Google Scholar
Crossref
Search ADS
 
Torrence
C.
 &
Compo
G. P.
1998
A practical guide to wavelet analysis
.
Bulletin of the American Meteorological Society
79
(
1
),
61
–
68
.
Google Scholar
Crossref
Search ADS
 
Touchan
R.
,
Xoplaki
E.
,
Funkhouser
G.
,
Luterbacher
J.
,
Hughes
M. K.
,
Erkan
N.
,
Akkemik
Ü.
 &
Stephan
J.
2005
Reconstructions of spring/summer precipitation for the Eastern Mediterranean from tree-ring widths and its connection to large-scale atmospheric circulation
.
Climate Dynamics
25
(
1
),
75
–
98
.
Google Scholar
Crossref
Search ADS
 
Wang
G.
,
Xia
J.
 &
Chen
J.
2009
Quantification of effects of climate variations and human activities on runoff by a monthly water balance model: A case study of the Chaobai River basin in northern China
.
Water Resources Research
45
,
W00A11
.
Google Scholar
 
Wu
Q.
,
Dong
D.
,
Fu
Y.
,
Bai
X.
 &
Sun
Z.
2002
Water environment problems induced by coal mining
.
Journal of China University of Mining & Technology
1
,
22
–
25
(in Chinese)
.
Google Scholar
 
Xue
D.
,
Zhou
J.
,
Zhao
X.
,
Liu
C.
,
Wei
W.
,
Yang
X.
,
Li
Q.
 &
Zhao
Y.
2021
Impacts of climate change and human activities on runoff change in a typical arid watershed, NW China
.
Ecological Indicators
121
,
107013
.
Google Scholar
Crossref
Search ADS
 
Zhang
S.
,
Ma
C.
 &
Zhang
L.
2011
Influencing factors of runoff attenuation of Wulan Mulun River in Daliuta Mining area, Yulin
.
Acta Scientiae Circumstantiae
31
(
4
),
889
–
896
(in Chinese)
.
Google Scholar
 
Zhang
A.
,
Zhang
C.
,
Fu
G.
,
Wang
B.
,
Bao
Z.
 &
Zheng
H.
2012a
Assessments of impacts of climate change and human activities on runoff with SWAT for the Huifa River Basin, Northeast China
.
Water Resources Management
26
(
8
),
2199
–
2217
.
Google Scholar
Crossref
Search ADS
 
Zhao
X.
,
Wang
S.
 &
Fan
X.
2010
Analysis on the change trend of runoff and influence factors in Kuye River Basin from 1954 to 1993
.
Journal of Water Resources & Water Engineering
21
(
05
),
32
–
36
(in Chinese)
.
Google Scholar
 

Author notes

†

The author contributed equally to this work as co-first author.

© 2024 The Authors
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).