In the wake of global and regional climate change and heightened human activities, runoff from some rivers in the world, especially in the arid and semi-arid regions, has significantly decreased. To reveal the varying characteristics leading to the change in runoff, detecting the influencing factors has been important in recent scientific discussions for water resources management in drainage basins. In this paper, an investigation into attributing the runoff response to climate change and human activities were conducted in two catchments (Wushan and Shetang), situated in the upper reaches of Weihe River in China. Prior to the identification of the factors that influenced runoff changes, the Mann-Kendall test was adopted to identify the trends in hydro-climate series. Also, change-points in the annual runoff were detected through Pettitt test and the precipitation-runoff double cumulative curve method. It is found that both catchments presented significant negative trend in annual runoff and the detected change-point in runoff occurs in 1993. Hence, the baseline period and human-inducing period are defined before and after 1993, respectively. Then, runoff response to climate change and human activities was quantitatively evaluated on the basis of hydrologic sensitivity analysis and hydrologic model simulation. They provided similar estimates of the percentage change in mean annual runoff for the human-induced period over the considered catchments. It is found that the decline in annual runoff over both catchments can be mainly attributed to the human activities, the reduction percentages due to human activities range from 59% to 77%. The results of this study can provide a reference for the development, utilization and management of the regional water resources and ecological environment protection.
Key words: climate change; human activities; Weihe River; change point; hydrologic sensitivity analysis; hydrologic model
1. Introduction
Changes in runoff can be attributed to the combined effects of climate, land cover and human activities in the basin. With the global warming on the rise, as well as, the overexploitation of water resources, decreases in streamflow has appeared in a large number of rivers (e.g.Zhang et al, 2012; Wang et al, 2012; Chen et al, 2012). The reduction in runoff could impact the river functions, and further induce severe ecological and environmental problems. Hence, the investigations on the attributions to the change in runoff have recently drawn considerable concerns.
During the past decades, due to the strong climate warming and the significant regional precipitation variation coupled with the drastic agricultural and industrial development in China (Piao et al., 2010), more attention has been paid to the assessment of the impacts of climate change and human activities on runoff change. Owing to the impacts varying from place to place, it is usually investigated at a local scale. Wang et al (2010) indicated the percentage of the runoff change due to climate change is found to be 89 % followed by 66 % and 56 % in 1970s, 1980s and 1990s, respectively in Baimasi River (sub-basin of the Yellow River). Li et al (2009) pointed out that climate variability influenced the surface hydrological process more significantly than land use change during the 1981-2000 in an agricultural catchment of the Loess Plateau, a tributary of Jinghe River. Hao et al (2008) suggested the impact of human activities on the decrease of surface runoff in the main stream is 41.59 % (in 1970s), 63.77% (in 1980s) and 75.15 % (in 1990s) in Tarim River. Jiang et al (2011) quantitatively analyzed the effects of climate variability and human activities on runoff from the Laohahe basin in northern China. Wang et al (2012) concluded that human activities could be mainly responsible for the runoff reduction in three sub-basins (Luanhe River catchment, Chaohe River catchment and Zhanghe River catchment) of the Huaihe River Basin.
For a certain river basin, the following steps can be adopted to quantify the effects of climate variability and human activities. Firstly, the change points are determined by the advanced statistical methods (e.g. Mann-Kendall (MK) trend test (Kendall，1975), the Pettitt test (Pettitt, 1979; Kiely et al., 1998) or DDC method (Huo et al., 2008), and the period before the change is regarded as a baseline period. Therefore the period that shows impacts of climate variability and human activities are separated from the baseline or natural period. This is followed by application of methods that summarize the effect of climate variability.The remainder of the effect is then attributed to other factors such as human activities (Zhao et al., 2010).
There have been a great number of methods used to separate the impacts of climate change and human activities on runoff. However, the hydrological model simulation method is traditionally the most widely used. For instance, Zhang et al (2012) and Fan et al (2010) employed soil and water assessment tool to measure the effect of natural and human factors on the water cycle in Huifa River Basin, the variable infiltration capacity model is applied by Jiang et al (2011) in Laoha River basin. The similar studies have also been conducted on the basis of monthly water balance model (eg: Wang et al, 2009; Wang et al, 2011). In addition, some new methods have been developed and widely used in many regions to estimate the effect of climate variability or human activities, such as regression analysis method and hydrological sensitivity analysis method. For example, Tian et al (2009) and Zhang et al (2009) employed regression analysis to estimate the impact of human activities on streamflow in the Hutuo River basin, and the lower Xijiang, respectively. The hydrological sensitivity method, developed by Dooge et al (1999) and Milly and Dunne (2002) to describe first-order effect of changes in precipitation and potential evaporation on streamflow, has been successfully used to evaluate the effects of climate variability and human activities on the hydrologic cycle (e.g. Jones et al., 2006; Zhao et al.,2010).
The Weihe River, the largest tributary of the Yellow River, plays an important role in the Yellow River's ecological and environmental improvement. The key area of the runoff yield in the Weihe River basin is situated in the upper reaches. In the past decades, the annual runoff in the upper reaches has decreased significantly. For example, compared with that of 1981-1990, the annual runoff obtained from Linjiacun gauging stations in 1991-2000 decreased by 53.9 %, the sharply decreasing runoff leads to the reduction in the contribution of Wei River's discharge flux into the Yellow River, further having an impact on the ecology and environment in the Yellow River. To date, there have been some studies reporting that the streamflow for the upper reaches in the Weihe River basin has dramatically decreased during recent years. Liang et al (2012) concluded that the inter-annual variation of runoff in upper reach is large, and the decrease in runoff is significant during the period 1960-2000. Wang et al (2006）argued that human activities were the main reasons behind runoff reduction since 1980s, and the extent of the influence is intensifying gradually. However, systematically quantifying the effects of climatic variability and human activities on runoff change in the upper reaches in the Weihe River basin has not been reported. Therefore, the objectives of this study are to (1) identify change trends and change points in annual runoff and (2) separate the effects of climatic variability and human activities on runoff with two methods. This paper is organized as follows: First, a brief description of the study area and data sources is given. Next, details of the methods used are provided. This is followed by a presentation of the results, including change points determination, inter-comparison of models in the estimation of the influence of climate variability and human activities on runoff. Finally, discussions and conclusions of the study are given.
2. STUDY AREA AND DATA
2.1 Study area
The Weihe River, the largest tributary of the Yellow River, originates from the north of the Niaoshu Mountains at an altitude of 3485 m, runs across 818 km through the provinces of Gansu and Shanxi and feeds into the Yellow River (Figure 1). The upper reaches of the Weihe River refer to the catchment above the Linjiacun gauging stations and extend from 104.00︒E to 107.00︒E longitude, and 34.25︒N to 36.25︒N latitude, and covers an area of approximately. The elevation within the basin ranges from 647 to 3635m above the sea level. It is characterized by semi-arid continental climate. The average annual temperature is between 9 and 13°C, the annual rainfall ranges from 315 to 664 mm, the main flooding season usually occurs in July, August and September, accounting for nearly 60%-70% of annual total runoff over the upper reaches of Weihe River. In addition, the average annual evaporation in the study area is up to 1400 mm or so.
As a result of serious soil and water loss, a great number of soil and water conservation projections have been constructed in recent 30 years. The detailed soil and water conservation measures include terraced fields, warping dams and afforestation (Wang et al, 1994., Ma et al., 2002). Unfortunately, no information is available on historical land use in the basin, apart from some contextual data indicating the increasing intensity in soil and water conservation projections since middle 1980s. For instance, the control degree in soil and water loss achieves 23.9% in 1989, and increased control degree occurred in 1996, up to 31.4% (Ma et al., 2002).
2.2 Data
The Wushan (in the period 1975-2007) and Shetang (in the period 1972-2007) gauging station were selected to examine the annual runoff variations of the Wushan and Shetang River basins, respectively. Daily precipitation data in the consistent period were obtained from the hydrologic year book. The same-period time series of daily streamflow in Wushan and Shetang hydrologic station were prepared. The meteorological data, including daily mean air temperature, wind speed, relative humidity and sunshine duration, are not available for the two considered catchments. The closest available observations are for the meteorological station of Huajialing, approximately 50 km to the east of the Wushan river basin. Similarly, the meteorological data from Tianshui station are applied to the simulation in the Shetang river basin. The Penman-Monteith equation recommended by Food and Agriculture Organization of the United Nations was used to calculate Potential Evapotranspiration (Allen et al., 1998).
3. METHODOLOGY
3.1 Trend test and change point analysis method
3.1.1 Trend test
Mann-Kendall test. The rank-based Mann-Kendall test (Kendall, 1975) was used to detect trends in the hydro-climatic series in this study. The method, recommended by the World Meteorological Organization and widely used (e.g. Zhang et al., 2008; Wang et al., 2011a, b), is usually adopted to estimate the significance of monotonic trends in hydrological and meteorological series.
For a time series, where n>10, the standard normal statistic Z is estimated as follows:
Where
In which is the extent of any given time.
The statistic Z follows the standard normal distribution. At a 10 % significance level, the null hypothesis of no trend is rejected if . Similarly, the threshold values of can reach up to 1.96 (at a 5 % significance level) and 2.58(at a 1 % significance level). A positive value of Z indicates an increasing trend, and the opposite corresponds to a decreasing trend. The pre-whitening technique (Yue and Wang, 2002) was adopted to eliminate the effects of the serial correlation on the MK test.
3.1.2 Change point analysis method
(1) Pettitt's test. The Pettitt's test (Pettitt, 1979) is a non-parametric approach to determine the occurrence of a change point. It has been commonly used to detect changes in the hydrological as well as climatic series (e.g. Verstraeten et al., 2006). This approach considers a time series as two samples represented by and . The Pettitt indices can be calculated from the following formula (Kiely et al., 1998):
(t=1,…, n) (5)
Then, when the largest appears is just the change point year.
(2) Double mass curve method. The double mass curve (DMC) is the plot of the accumulated values of one variable against the accumulated values of another related variable for a concurrent period (Searcy and Hardison, 1960). DMC between precipitation and runoff has recently become an effective tool for detecting the changes of hydrological regime due to anthropogenic disturbances (e.g. Huo et al., 2008). Normally the DMC between precipitation and runoff is a straight line, a change in the gradient of the curve may present that the original relationship between variables was broken. In this study, the DMC will be utilized to identify the change point of the runoff series as a confirmation of the change points detected by Pettitt's test.
Through trend and change-point analysis, the runoff series will be divided into a natural period series and a human-induced period series. On the basis of the divided periods, the impacts of climate variability and human activities on runoff can be separated by using the following methods.
3.2 Hydrologic sensitivity analysis method
Hydrological sensitivity can be described as the percentage change in mean annual runoff in response to the change in mean annual precipitation and potential evapotranspiration. The water balance for a basin can be described as
(6)
Where P is precipitation, E is actual evapotranspiration (AET), Q is streamflow, and ΔS is the change in soil water storage. For a long period (i.e. 10 years or more), ΔS can be assumed as zero.
Following a simple model (called Zhang's curve) developed by Zhang et al (2001).
Long-term mean annual AET can be estimated as follows:
(7)
Where PET is the potential evapotranspiration and w is the plant-available water coefficient related to vegetation type (Zhang et al., 2001). The details of the relationship can be found in Zhang et al (2001). In this study, the parameter w is calibrated by comparing the long-term annual AET from Equation (2) and (3).
Perturbations in both precipitation and PET can lead to changes of water balance. It can be assumed that a change in mean annual runoff caused by climate variability is determined as follows expression (Milly and Dunne, 2002):
(8)
where, ΔP and ΔPET denote changes in runoff, precipitation and PET, respectively, and and are the sensitivity coefficients of runoff to precipitation and PET, which are expressed as (Li et al., 2007)
(9)
(10)
in which x is the mean annual index of dryness (equal to PET/P).
3.3 Hydrologic model simulation method
For the sake of evaluating the impacts of climate change and human activities on runoff variation, method of reconstructing nature runoff based on the hydrological models was used. The hydrological model was first calibrated based on observed runoff in the natural period, and natural runoff during the human-induced period is reconstructed by changing only meteorological input without any change in the calibrated parameter and consideration of local human activities.
Then the impact of human activities on runoff can be calculated as follows:
(11)
where represents the change in mean annual runoff due to the effect of human activities， denotes the observed runoff of the human-induced period;expresses the reconstructed runoff for the human-induced period.
In this study, the limited information and the available data sets fail to meet the minimal requirements for physical-based hydrological model. Instead, a simple lumped hydrological model is used to estimate the effects of climate variability and human activities on annual streamflow.
The HBV model (Seibert 1998) uses daily precipitation and air temperature and monthly potential evaporation as inputs to simulate runoff processes. The model includes three main modules: snow accumulation and melt, soil moisture routing, and river routing and response modules (Abebe et al., 2010). Snow melt is simulated with the degree-day method (Eq. 12).
(12)
Three parameters (FC, SM, and BETA) are involved in the calculation of soil moisture. FC is the maximum soil storage capacity, and parameter SM is the water content of the soil box in the basin, BETA determines the contribution to runoff from rainfall and snow melt (P) depends on the relation between SM and FC (Eq. 13). The parameter LP donates SM threshold for reduction of evaporation, ranging from 0 to 1. When SM/(FC-LP) is above 1, actual evaporation equals the potential evaporation. While a linear reduction is used when SM/(FC-LP) is below 1 (Eq. 14).
(13)
(14)
The response routing consists of two reservoirs (the upper groundwater box (SUZ (mm) and the lower groundwater box (SLZ (mm)), which is connected by constant percolation rate PERC (). Runoff from the groundwater boxes is simulated as the sum of the two or three outflow equations (, and ()) controlled by SUZ is above or below a threshold value, UZL (mm) (Eq. 15). Finally, the groundwater runoff is transformed by a triangular weighting function defined by the parameter MAXBAS (d) Eq. 16 and Eq. 17) to simulate the total runoff ().
(15)
(16)
(17)
The Nash-Sutcliffe coefficient (NSCE) and Relative Bias (BIAS), defined by Equations (18), (19), are used to evaluate the model performance.
(18)
(19)
Where is the observed runoff (mm/month) at time step , is the simulated runoff (mm/month) at the corresponding time, is the mean value of the observed values (mm/month), and is the number of data points.
3.4 Estimating the contribution of climate change and human activities to runoff
Observed runoff during the nature period is taken as benchmark value, the difference between it and observed runoff in the human-induced period was assumed to the result of both climate variability and human activities (Ma et al. 2009; Liu et al. 2010). In this study, the impact of climate change and human activities on runoff variation can be separated following the formulas:
(20)
(21)
(22)
(23)
where is the total change of runoff, denotes the measured runoff of the natural period, and are defined as previously, and and show the impact in percentage of human activities and climate variability on runoff, respectively.
4. RESULTS AND DISCUSSION
4.1 Trend and change-point analysis of PET, precipitation and runoff series
Historical trends of hydro-meteorology factors can help to understand the effects of
Climate change on water resources systems. Mann-Kendall trend test method was applied to identify the change in trends of annual precipitation, PET and runoff depth in Wushan and Shetang catchments of Weihe River. The statistical results on the basis of the MK test are shown in table 1. Combined with the Figure 2, it is found that in despite of decreased annual precipitation in both the catchments, the statistically significant trends cannot be identified in either. However, annual runoff in two catchments showed remarkable negative trends at the rates of 2.64mm (Wushan River, confidence levels of 99%) and 0.97 mm (Shetang River, confidence levels of 90%) every year. In addition, the significant increasing trend is shown for PET at the rates of 1.72 mm (Wushan River, confidence levels of 99%) and 1.16 mm (Shetang River, confidence levels of 90%) every year.
The Pettitt and DCC tests were applied to identify the change-point of the annual runoff series. The results of Pettitt's test are shown in Figure 3, it is concluded that 1993 could be the detected significant change points reflecting the effect of human activities on runoff for the two catchments.
Furthermore, we use the DMC method to detect the change points in runoff series. Figure 4 shows the cumulative annual precipitation and runoff over the two catchments. It shows that the relationships between cumulative annual precipitation and cumulative annual runoff can be expressed nearly with two straight lines in different slopes before and after 1993, suggesting the characteristics of precipitation or runoff changed after 1993.
Overall, the change points detected in annual runoff occurred in early 1990s (listed in Table 2). Then, the study period for both catchments can be divided into the natural period and human-induced period of the change points.
In order to better understand the characteristics of the runoff change, firstly, the differences between the means of the annual runoff during the nature and human-induced periods are analyzed using the T-test. The significant differences can be found at 90% confidence level for all catchments, inferring the differences in the factors influencing runoff for the nature and human-induces periods. In addition, the average monthly precipitation and runoff for the two periods (figure 5) are compared to further understand the impacts of climate and other factors on runoff during the two periods. For both catchments, the changes in mean monthly precipitation were not obvious for the two periods. However, the dramatic reductions are presented in average monthly runoff during 1994-2007, compared with that for the natural period. And the greatest decline is shown in flood seasons (July, August, and September). As an example, the slightly increasing precipitation and consistent largely decreased runoff happens in July at both catchments.
Hence, to some degree, the reduction in runoff during 1993-2007 may be due to basin-related human activities.
4.2 Calibration and validation of different methods
In the hydrological sensitivity analysis method, three calibrated parameters including w, (the runoff sensitivity coefficients to precipitation) and (the runoff sensitivity coefficients to PET) are shown in Table 3. The obtained w values are 1.94 (Wushan River) and 1.45 (Shetang River). For both catchments, the absolute value of (the sensitivity coefficients of runoff to precipitation) is larger than (the sensitivity coefficients of runoff to PET), revealing that the change in runoff was more sensitive to precipitation than to PET.
For the hydrologic model simulation method, the HBV model was calibrated by Monte Carlo method during the period of 1975-1984 at Wushan station and 1972-1984 at Shetang station, the corresponding validation periods at both stations are 1985-1992. The score skills for the HBV model during the calibrated and validated periods are summarized in Table 4. During the calibration period, the NSCE coefficient reaches up to 0.81 (Wushan station) and 0.85 (Shetang station), and the absolute values of BIAS in two stations are both lower than 10%. In comparison with the calibration period, the lower score skills for HBV model during the validation are shown, but the model results are overall acceptable. Meanwhile, good agreement between monthly observed and simulated runoff at both stations are presented in Figure 6, during nature period. Thereafter, natural runoff during the human-induced period is reconstructed with the calibrated hydrologic model and the actual meteorological and hydrologic data. The reconstructed runoff series during the human-induced period and the corresponding observed runoff series provide the opportunity to quantitatively estimate the effects of climate variability and human activities on runoff.
4.3 Effects of climate variability and human activities on runoff
With the simulated results of the two different estimation methods, the evaluated effects of climate variability and human activities on runoff are shown in Table 5. For catchments, the hydrologic model simulation method and hydrologic sensitivity analysis method provided approximate estimations of change in mean annual runoff for human-induced period induced by climate variability and human activities. Concurrently, it offers confidence in the methods which are applied to separate the effects of climate change and human activities. The runoff reduction during the post-change period (form1993 to 2007) should be mainly attributed to human activities for Wushan and Shetang catchments. Human activity should be responsible for 59% and 71% runoff change computed by hydrological model and hydrological sensitivity analysis in Wushan River. For the Shetang River, the 66% (detected by hydrological model) and 77% (detected by hydrological sensitivity analysis) of reduction in runoff for human-induced period are induced by human activities. Figure 7 presents the time series of for the human-induces period, on the basis of two methods. Overall, the series computed by different methods were comparable, inferring that they are capable of simulating the effects of climate variability and human activities. Moreover, a positive effect of climate variability on runoff can be easily found when the annual precipitation is high. This phenomenon was also found in some earlier studies in another area of China (e.g. Laohahe River catchment (Jiang et al., 2011); Huaihe River basin (Wang et al, 2012)). In order to understand the effect of larger human activities in the human - induced period, four pairs of measured runoff under about equal amount of evaporation are shown in Table 6. From the table one can get the conclusion that when under the same evaporation and the different underlying surface condition, the higher annual precipitation during human-induced period results in the lower observed annual runoff in both catchments. Moreover, 1991 and 2001 are selected as two sample years to identify the difference in the runoff response to precipitation at Wuhan sub-basin (with the similar annual evaporation) before and after 1993 (Figure 8a-8b). It is concluded that the relationship between precipitation and runoff become weaker in human-induced period (in the year of 2001). In flood seasons (from June to August), the heavy precipitation just resulted in lower flow within the Wushan sub-basin. Also the similar result can be achieved in Shetang River basin from figure 8c-8d. Combined figure 8 with table 6, it is suggested that the runoff was dramatically affected by human activities during the post-change period.
5. DISCUSSION
In despite of no clear trends in annual precipitation, significant decreasing trends in runoff can be found in two catchments of the upper reaches of the Weihe River basin. This infers, to some degree, that runoff in the considered catchments may be affected by other factors (human activities) apart from the climate change. Usually, related human activities are considered as the reasons leading to the sharp decline in runoff and they include agricultural irrigation, industry development, dam construction as well as soil and water conservation mechanical measures. The change points of runoff in the two catchments happened in the early 1990s, which corresponds to the fact that the constructions of soil and water conservation projection have been gradually increasing since 1985 (Wang et al, 2006). For example, the increase in area of terraced fields reaches up to 74.70during 1994-2000 in upper reaches located in Gansu province including the two catchments. Therefore, the change in land cover due to the constructions of terraced field sand as well as soil and water conservation projection may be the main driving factors of runoff decline. The upper reaches of the Weihe River basin are a large heavily sediment and runoff-laden catchment. Owing to serious soil and water loss, a great number of soil and water conservation projection have been constructed with data. At the same time, it changes the land cover, which distinctly decreases runoff and sediment in the watershed outlet at upper reaches. The significantly decreased annual runoff in upper reaches can influence living, production and ecological use of water resource in lower reaches，but also can decrease the water discharge to the midstream and downstream of the Yellow River, and intensified the water shortage in the Yellow River Basin. Thus, proposing orderly human activities is important for this regional water resources' sustainable development. The measures can be summarized as the reasonable layout for the soil and water conservation, as well as moderately returning farmland to forest or grassland, in term of increasing the produced runoff.
Usually, distributed physically based hydrological model may be preferred for hydrological effect study (Legesse et al., 2003). However, the limitations, such as its complexity in model setup as well as data set requirements involving topography, vegetation and soil hydraulic properties (Wei and Zhang, 2010), lie when its application at basin scale. In this study, a simple lumped hydrological model was selected for making hydrological simulation. The simple lumped hydrological model is not expected to provide spatial information about hydrological processes, whereas our results show that it did not affect the performance of the simple water balance model in terms of quantifying the climate and anthropogenic effects on runoff.
It should be noted that some uncertainties lie in assessing effects of climate variability and human activities on runoff. First, uncertainty may arise from the limited hydro-meteorological observation data. Because no meteorological station exists in Wushan and Shetang catchment, meteorological data such as daily mean temperature, wind speed and relative humidity from a nearby meteorological station are used, which may limit the accuracy of the calculated PET and simulated runoff. Second, the hydrological sensitivity analysis denotes the response of runoff to the annual precipitation and PET, However, runoff can be influenced by changes in other influencing factors, as an example, when under the same underlying surface condition and annual precipitation, the higher the percentage of the precipitation during flood season the larger the simulated runoff (Zhang et al, 2012). The absence of these aspects may affect the accuracy of the hydrological sensitivity analysis method (Zhao et al., 2010). Moreover, uncertainty in model parameters can also inevitably affect the simulation results (Jiang et al., 2011). More work should be conducted in future studies to quantify and reduce these uncertainties.
6. CONCLUSION
With global or regional climate change and enhanced human activities, decrease in a large number of river streamflow especially in the arid and semi-arid areas has appeared. It is hence useful to investigate the attributions to the changes in runoff. In this study, hydrological model method and hydrological sensitivity method are applied to quantitatively estimate the impacts of climate variability and human activities on runoff in two catchments (i.e. the Wushan catchment and Shetang catchment) located in the upper reaches of the Weihe River basin. The main conclusions are shown as follows:
Significant decreasing trends in runoff can be found in both catchments, especially in Wushan basin, which are dominated by significant decreasing trends at 99% confidence level, whereas no significant trend in precipitation is found in either catchment. For both catchments, the detected change points in annual runoff series occurred in 1993, on the basis of Pettitt and DMC tests. Accordingly, the annual runoff series can be divided into two periods named baseline and human-induced periods. Compared with the baseline period, reductions in mean annual runoff range from -42.6% to 52.9% during 1993-2007.
(2) Similar estimates of the impacts of climate variability and human activities on runoff in 1993-2007 are obtained, by means of the hydrological model simulation method and hydrological sensitivity analysis method. Human activities should be mainly responsible for the runoff reduction in the Wushan catchment (accounting for 59% and 71% by hydrological model method and hydrological sensitivity method respectively), and Shetang catchment (accounting for 66% and 77%, respectively).
(3) The results of the present study can supply a reference to regional water resources management and planning. At the same time, a practically possible proposition in term of increasing the produced runoff has been put forward for local managers to reasonably arrange the local actions, synthetically considering the sustainable development in the regional water resource and ecological environment.
ACKNOWLEDGEMENTS
This study was supported by the National Natural Science Foundation of China (NO. 41130639, NO. 51179045) and the Research and Innovation Program for Graduate of Universities in Jiangsu Province China (NO. CXZZ12_0241).
