Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
Contents lists available at ScienceDirect
Environmental Research journal homepage: www.elsevier.com/locate/envres
Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin Jing Zhao n, Qiang Huang, Jianxia Chang, Dengfeng Liu, Shengzhi Huang, Xiaoyu Shi State Key Laboratory Base of Eco-Hydraulic Engineering in Arid Area, Xi’an University of Technology, Xi’an 710048, China
art ic l e i nf o
a b s t r a c t
Article history: Received 26 September 2014 Received in revised form 19 December 2014 Accepted 31 December 2014
The Wei River is the largest tributary of the Yellow River in China. The relationship between runoff and precipitation in the Wei River Basin has been changed due to the changing climate and increasingly intensified human activities. In this paper, we determine abrupt changes in hydro-climatic variables and identify the main driving factors for the changes in the Wei River Basin. The nature of the changes is analysed based on data collected at twenty-one weather stations and five hydrological stations in the period of 1960–2010. The sequential Mann–Kendall test analysis is used to capture temporal trends and abrupt changes in the five sub-catchments of the Wei River Basin. A non-parametric trend test at the basin scale for annual data shows a decreasing trend of precipitation and runoff over the past fifty-one years. The temperature exhibits an increase trend in the entire period. The potential evaporation was calculated based on the Penman-Monteith equation, presenting an increasing trend of evaporation since 1990. The stations with a significant decreasing trend in annual runoff mainly are located in the west of the Wei River primarily interfered by human activities. Regression analysis indicates that human activity was possibly the main cause of the decline of runoff after 1970. & 2015 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Keywords: Precipitation The Mann–Kendall test method Change points Trends The Wei River Basin
1. Introduction Climate change is an important factor affecting hydrological cycle, and the hydrological regime could be severely modified in response to the anticipated changes in temperature and precipitation during the present century. The increasing average surface air temperature intensified global hydrological cycles during the 20th century is a general consensus (Huntington, 2006; Milliman et al., 2008; Déry et al., 2009). Air temperature near the surface of the Earth increased by 0.74 °C (IPCC, 2007) over the last century, and the observed air temperature in China has increased by 1.1 °C since the 1980s at an increasing rate of 0.22 °C/decade, which is higher than the global average (Kundzewicz et al., 2009; Piao et al., 2012). Rising global temperatures have been accompanied by changes in weather and climate. Many regions have seen significant changes in precipitation which have led to more floods, droughts, as well as more frequent and severe heat waves (IPCC, 2007; Shrestha et al., 2012; Kang et al., 2007; Dai, 2013). Streamflow as an important hydrological variable is a combination of what happens to precipitation, temperature and other components of hydrological cycle within a large basin scale (Z 100,000 km2); as such, streamflow can be used as an indicator of n
Corresponding author. Fax: þ 86 29 82312797. E-mail address: [email protected] (J. Zhao).
hydrological responses to climate change (Hao et al., 2008; Fu et al., 2007; Sankarasubramanian and Vogel, 2003; Lioubimtseva and Henebry 2009). The increasing shortage of water in most river basins is mainly due to human activities (Vorosmarty et al., 2000). Runoff, an important source of replenishment for surface and groundwater storage, is also declining in most river basins. Therefore, in recent decades, analyzing the change trend and identifying the driving factors on runoff has been the focus of hydrological research (Labat et al., 2004; Liu and Xia, 2004; Fraiture, 2007). The traditional regression method and the hydrological models are usually used to identify the magnitude of impacts of climate change on runoff (e.g., Yao et al., 2008). Additionally, the methods such as the Crammer's method, moving Ttest (MTT), Yamamoto method, Mann–Kendall method and Pettitttest method also can be used to detect abrupt changes in hydrometeorological phenomena. Wei River Basin, the largest tributary of the Yellow River, is an important foodstuff base in China. Natural streamflow as the main source of surface water plays a key role in agricultural irrigation, socio-economic development and local eco-environmental conservation in the basin. Precipitation and temperature are primary atmospheric factors affecting the streamflow. Annual and seasonal precipitation in the Wei river show decreasing trends, especially in summer precipitation, and the temperature shows an increasing trend of 0.3 °C per decade in recent 50 years, obviously exceeding the warming amplitude of global average temperature. Studies
http://dx.doi.org/10.1016/j.envres.2014.12.028 0013-9351/& 2015 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
2
have shown that the runoff of the Wei River has decreased significantly in recent decades. The shortage of water resources in the Wei River Basin is becoming more and more serious. The annual per-capita water availability is far below the recognized standard (IPCC, 2001). The descent rate of precipitation and runoff are extremely inconsistent. This does not only limit socio-economic development in the region, but also poses considerable threat to the health of the ecosystem and the environment at the downstream of the catchment (Luo and Guo, 2004). The variability of streamflow in the Wei River Basin has attracted attentions and interests from academic circles and local government. The objective of this paper is to comprehensively investigate the variability of long-term historical records of climate and hydrological data for the longest time-series and highest number of stations available in the Wei River Basin, to determine runoff slope break by means of the sequential Mann-Kendall test, and then to analyze the main driving factors in related period of abrupt runoff decline.
2. Study area and data 2.1. Study area The Wei River Basin, which originates from the north side of Niaoshu Mountain, is the largest tributary of the Yellow River. With its length of 818 km and the basin area of 13.5 10 km2, the Wei River flows from east to west through the Loess Plateau in East Gansu, the Tianshui Basin, the Baoji Vally, the Guanzhong Basin, and makes the confluence with Yellow River at Tongguan County in Shaanxi Province (Fig 1). The length of the channel is approximately 502 km and the basin area is approximately 6.24 104 km2 in Shaanxi Province before the confluence. As the largest branch of the Yellow River, the average annual runoff of Wei River is 8.09 billion m3 at Huaxian Station. The drainage basin lies between 103.5°E–110.5°E and 33.5°N–37.5°N. Located in the continental monsoon climate zone, the Wei River Basin is characterized by abundant precipitation and high temperature in summer and by rare precipitation and very low temperature in winter. The annual precipitation of the basin is approximately 559 mm (Huang et al., 2014a, 2014b). Topographically, the altitude decreases from the highest northwest mountainous areas to the lowest Guanzhong Plain in the southeast and south portion of the basin. It is worth mentioning that the Guanzhong Plain is designated as a state key economic development zone, acting as a great stimulus to the economic development of surrounding area. Therefore, the Guanzhong
Plain's economic development will directly affect the sustainable development of Shaanxi province's economic society. In recent years, with the development of economy and population growth, the Guanzhong region water demand increases greatly. At present, it is difficult for the local water of Guanzhong region to satisfy the requirement of economic and social development. Given the significance of water security in the Guanzhong plain, a deep understanding of the change in climate and runoff is of great significance in the Wei River Basin. 2.2. Data The following hydro-meteorological data were used in this study. Daily precipitation and temperature data (1960–2010) of 21 National Meteorological Observatory (NMO) stations were provided by the National Climatic Centre of China. The potential evaporation data were estimated by the Penman–Menteith method, which was recommended by the World's Food and Agriculture Organization (FAO) in 1998 (Allen et al., 1998; Huang et al., 2014a, 2014b), and in order to simplify, the potential evaporation is called as evaporation in the article. The location of the stations in the basin is shown in Fig. 2, and their latitudes and longitudes are listed in Table 1 The monthly runoff (1960–2010) of the 5 hydrological stations in the Wei River was provided by the Bureau of Hydrology of the Yellow River Water Resources Commission. The hydrological gauge stations include the Linjiacun station, Xianyang station, Huaxian station, Zhangjiashan station and Zhuangtou station (Table 2).
3. Methods Four statistical methods were used in this study to analyze the spatial variations and temporal trends of the hydro-climatic series. (1) A simple linear regression method, which is a parametric test method, was utilized to test the long-term linear trend. (2) The Mann–Kendall test method, which was originally devised by Mann (1945) as a non-parametric test for detecting trends and the distribution of the test statistic derived by Kendall (1975), was employed to test the non-linear trend as well as the turning point. (3) The Kendall method, which is a non-parametric method for testing the correlation (Kendall, 1938; Sprent, 1990), was applied to test the correlation between runoff, precipitation and temperature. (4) The regression analysis was used to identify the main driving factors of runoff decline in the study area.
Fig. 1. The location of the Wei River Basin in the Yellow River basin.
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
3
Fig. 2. The location of the hydrological and meteorological stations in the Wei River Basin.
Table 1 The information of meteorological stations of Weihe River basin. No.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Stations
Lintao Huajialing Wuqi Guyuan Huanxian Yanan Xiji Pingliang Xifeng Changwu Luochuan Tongchuan Minxian Tianshui Baoji Wugong Xi’an Huashan Foping Zhenan Shangzhou
Location
Elevation (m)
Latitude (°N)
Longitude (°E)
35/21 35/23 36/55 36/00 36/35 36/36 35/58 35/33 35/44 35/12 35/49 35/05 34/26 34/35 34/21 34/15 34/18 34/29 33/31 33/26 33/52
103/51 105/00 108/16 106/16 107/18 109/30 105/43 106/40 107/38 107/48 109/30 109/04 104/01 105/45 107/08 108/13 108/56 110/05 107/59 109/09 109/58
1893 2450 1331 1753 1255 958 1916 1346 1421 1206 1159 978 2315 1141 612 447 397 2064 827 693 742
3.1. The Mann–Kendall (MK) trend test method The Mann–Kendall trend test, commonly known as Kendall tau statistic, is a non-parametric assessment of the significance of monotonic trends of hydro-meteorological variables (Zhu and Day, 2005; Novotny and Stefan, 2007). In the Mann–Kendall test, the null hypothesis H0 states that X1, ⋯Xn are samples of n independent and
identically distributed random variables with no seasonal changes. The alternative hypothesis H1 for a two-sided test defines the distributions of Xk and X j as non-identical for all k , j ≤ n ; with k ≠ j . The test statisticS , with zero mean and computed variance (as in Eq. (3)) is asymptotically normal and can be calculated using Eqs. (1) and (2) n− i
S=
n
∑ ∑
Sgn (x k − x j ) (1)
j=1 k=j+1
where
⎧ 1 xk > x j ⎪ ⎪ S gn(x k − x j ) = ⎨ 0 x k = x j ⎪ ⎪ ⎩−1 x k < x j Var (s) = ⎡⎣n (n − 1)(2n + 5) −
(2)
∑t
t (t − 1)(2t + 5)/18⎤⎦
(3)
The notation t is the extent of the given time and ∑ denotes the summation over the time span. In cases where the sample size n > 10, the standard normal variate Z is computed following Douglas et al. (2000) using:
⎧ ⎪ ⎪ ⎪ Z=⎨ ⎪ ⎪ ⎪ ⎩
S−1 Var (S) 0 S+1 Var (S)
⎫ S > 0⎪ ⎪ ⎪ 0 ⎬ ⎪ S < 0⎪ ⎪ ⎭
(4)
Thus in a two-sided trend test, H0 is accepted if Z ≤ Z α /2 at α level of significance. Positive S value indicates an ‘upward-trend’ while negative S value indicates a ‘downward-trend’.
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
4
Table 2 The information of hydrological stations of Wei River basin. Stations
Linjiacun Xianyang Huaxian Zhangjiashan Zhuangtou
Drainage area (km2)
30661 46827 106498 43216 25645
Time of establishment
1934.1.1 1931.6.10 1935.3.1 1932.1.1 1932.1.1
Location
Period
Longitude(°E)
Latitude (°N)
107/00 108/42 109/42 108/35 108/35
34/21 34/17 34/44 34/35 34/35
1956–2010 1956–2010 1956–2010 1956–2010 1956–2010
Fig. 3. The annual precipitation (a) Evaporation (b) and temperature (c) variations and their mean differences of decadal values to the long-term mean in the Wei River basin.
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎ Table 3 Seasonal variations of the precipitation (mm), evaporation (mm) and temperature (°C) in Wei River basin. Sub-catchment
Season
P(mm)
E(mm)
T(°C)
Upstream of Linjiacun
Spring Summer Autumn Winter
53.1 92.7 20.7 9.4
80.7 110.2 46.4 25.8
14.7 18.4 3.2 0.3
Spring Summer Autumn Winter
58.5 107.7 30.5 14.3
82.5 120.6 49.9 28.8
19.0 22.9 7.1 3.6
Spring Summer Autumn Winter
58.7 106.0 31.9 15.5
110.7 104.4 39.6 42.6
17.4 21.3 5.7 1.9
Upstream of Zhangjiashan
Spring Summer Autumn Winter
45.8 95.8 20.1 9.3
89.8 120.9 50.3 29.1
15.6 19.3 2.8 0.8
Upstream of Zhuangtou
Spring Summer Autumn Winter
49.0 105.5 22.1 11.0
93.6 98.3 49.5 51.2
15.6 19.3 2.7 1.1
Linjiacun to Xianyang
Xianyang to Huaxian
The sequential version of Mann–Kendall test (Sneyers, 1975) was used to test assumptions about the start of a trend within the sample X1, ⋯Xn from set of random variable X based on rank series of progressive and retrograde rows of the sample. The magnitudes of X j annual mean time-series and j = 1, ⋯ , n are compared by Xk ,
5
where k = 1, ⋯ , j − 1. For each comparison, the number of cases Xk > X j is counted and denoted by n j . The test statistic is normally distributed with mean given as
tj =
j
∑1
E (t) =
n j (j = 2, 3, ⋯n)
n (n − 1) 4
(5)
(6)
and variance given as
Var (t j ) =
[j (j − 1)(2j + 5)] 72
(7)
The sequential values of the statistic U (t) are calculated as
U (t) =
t j − E (t) Var (t j )
(8)
which is the forward sequence, and the backward sequence U ′ (t) is calculated using the same equation but in the reverse data series. In two-sided trend test, a null hypothesis is accepted at a significance level if U (t) ≤ U (t)1 − α /2; where U (t)1 − α /2 is the critical value of the standard normal distribution with a probability exceeding α /2. A positive U (t) denotes an upward trend while the reverse denotes a downward trend (i.e., U ′ (t) is similar to U (t)). In this study, α is set at 0.05 significant levels. The sequential Mann–Kendall test enables detection of the approximate time of occurrence of a trend from the intersection point of the forward and backward curves of the test statistic. If the intersection point is significant at α = 0.05, then the critical point of change is at that period (Moraes et al., 1998; Gerstengarbe and Werner, 1999). Hence the sequential Mann–Kendall test is an efficient way by which the starting time of a trend is pinpointed.
Fig. 4. Trends of three variables at 21weather stations in the Wei River Basin 1960–2010. Green regular triangle indicates an increasing trend at the 5% significance level. Red inverse triangle indicates a decreasing trend at the 5% significance level. Blue dot represents station with trends of significance below 10%. P: Precipitation; T: Temperature; E: Evaporation. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
6
Fig. 5. The annual change of runoff in Wei River basin.
3.2. The regression analysis Regression analysis was used to identify the main driving factors of runoff change. Based on the result of the sequential Mann–Kendall test, runoff data for each basin were divided into two periods ‘predevelopment period’ and ‘post-development period’. Data for the predevelopment period (which is the period before abrupt change) of annual runoff were used as baseline for the comparison with the postdevelopment period (i.e. the period after abrupt runoff change). The relationship between precipitation and runoff in the two periods was then analyzed and compared. Comparisons of correlations between runoff and precipitation in the two periods provide a good measure for the effect of human activities on runoff.
4. Results and discussion 4.1. Temporal variation of precipitation, temperature and evaporation in the Wei River basin In order to study the decadal variation of historical annual and seasonal precipitation, temperature and evaporation, the value of the three meteorological elements for the whole study period
(1960–2010) and for each decade (1960–1969,1970–1979,1980– 1989,1990–1999 and 2000–2010) were calculated and compared. The results were illustrated in Fig. 3. The results suggest that: (1) In the upper stream of Linjiacun a wet period (1960–1969) and a dry (1990–1999) period are clearly seen, in which the mean annual precipitation is 49.59 mm (42.30 mm) higher (lower) than the long-term average, respectively. The periods (1990–1999, 2000–2010) are warmer than the long-term average, while the period (1960–1969, 1970–1979, and 1980–1989) is cooler than the long-term average. The maximum and minimum evaporation appeared in 2000–2010(27.88 mm) and 1980–1989( 31.93 mm) compared with the long-term average, respectively. (2) Between Linjiacun and Xianyang, the biggest difference is found in the period (1980–1989), which has a mean precipitation of 64.83 mm higher than the long-term average, and the period (1990–1999) which has a mean precipitation of 69.78 mm lower than the long-term average. The period (2000– 2010) is 0.79 °C warmer than the long-term average, and before 1989 the decadal variations of temperature are lower than the long-term average. The variation of evaporation in this section of the river is larger than other sections. The evaporation of the period (1960–1969) is 43.56 mm higher than the long-term average, and after 1989 the decadal variations of evaporation are very
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
7
Fig. 6. Trends of annual runoff of the 5 sections in the Wei River Basin during the 51 years. Upward (downward) triangles indicate positive (negative) trends from M–K test. Blue dots show sections with no trends or trends below 10% significance level. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
close to the long-term average. (3) For the section of Xianyang to Huaxian, a dry period (1990–1999) and a wet period (1980–1989) are found which have a mean annual value of 55.76 mm (54.90 mm) lower (higher) than the long-term average, respectively. The period (2000–2010) is warmer, while the period (1960– 1989) is cooler than the long-term average. (4) For the upstream of Zhangjiashan the mean decadal changes of precipitation are not remarkable, the biggest difference is found for the period (1960– 1969) which has a mean precipitation of 47.16 mm higher than the long-term average and the precipitation of other periods is the closest to the long-term average. The period (2000–2010) is 0.84 °C warmer than the long-term average, and before 1989 the decadal variations of temperature are lower than the long-term average. The maximum and minimum evaporation appeared in 1970–1979 (38.36 mm) and 1980–1989 ( 53.9 mm) compared with the long-term average, respectively. (5) For the upstream of zhuangtou, a dry period (1990–1999) and a wet period (1960– 1969) are found which have a mean annual value of 44.30 mm (57.40 mm) lower (higher) than the long-term average, while the rest periods is more closer to the long-term average, respectively. The periods (1990–1999 and 2000–2010) are warmer, while the rest of the periods are cooler than the long-term average. For the period (1980–1989), evaporation is 47.80 mm higher than the long-term average (Fig. 3). (6) In terms of seasonal variations, precipitation, temperature and evaporation in summer has larger difference than that of precipitation, temperature and evaporation in other seasons. The maximum value are appeared in the section of Linjiacun to Xianyang in summer (Table 3). The above analysis shows that, there is a very dry period in the Wei River when it entered the 1990s. If this situation does not alter in the 21st century, it will have a serious effect on agriculture, industry and drinking water supply in the Wei River Basin.
4.2. Spatial distribution of precipitation, temperature and evaporation Applying the Mann–Kendall statistics method at each station in the Wei River Basin, temporal trends of annual precipitation, temperature and evaporation are tested at the significance levels of α = 0.05 and α = 0.10. These temporal trends are then interpolated to show their spatial distributions. Fig.4 displays the location of the 21 weather stations and the direction of trends based on the M–K test of three variables of each station. Precipitation decreased at 6 stations around the boundary of the Wei Basin at the α = 0.05 significance level, while precipitation at the other 15 stations in the basin shows the trend of significance below 10%. Temperature increased statistically significantly at all stations of the Wei River Basin at the significance level of α = 0.05. Potential evaporation statistically significantly increased at 50% of the stations around the central basin only, with one exception in the South of the basin. The above analysis indicates that temperature has increased dramatically in the Wei River Basin. The significant increase in air temperature has resulted in a dramatic decrease in runoff caused by increased evaporation (refer to Fig. 5). The Wei River is the economic center of Shaanxi Province and rapid economic development experienced during last decades has increased air pollution in the region, which, in turn, has partly caused temperature increase and runoff decrease. The increasing trend in air temperature over the past decades in the region is consistent with global warming as reported across most of the globe. 4.3. Temporal variation of runoff The 51 years variation of runoff in the Wei River exhibits alternation from abundance to paucity (Fig. 5). The runoff at the five sections of the Wei River Basin tends to decrease over the past 51
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
8
Fig. 7. Sequential Mann–Kendall tests for annual runoff. Dotted horizontal lines represent critical values corresponding to the 95% confidence interval. Table 4 Starting time of runoff trend, abrupt change time, and significant runoff change time in the 5 sections determined by sequential Mann–Kendall test. Stations
Start of trend
Abrupt change
Significant trend
Upstream of Linjiacun Linjiacun to Xianyang Xianyang to Huaxian Upstream of Zhangjiashan Upstream of Zhuangtou
1970 1968 1964 1970 1970
1985 1989 1966 1996 1996
1986 1996 No 2004 2004
years. The annual runoff has been declining at the rate of 0.54, 0.33, 0.07, 0.16 and 0.08 108 m3/year, respectively, from the 1960s to the 2000s. From the 1960s to the 2000s, runoff declined by 86.67%, 53.44%, 16%, 50.24% and 43.62% respectively. The decrease of runoff will have a direct impact on the economy and society of the Wei River Basin. 4.4. Spatial distribution of runoff trends Annual trends of runoff were analyzed with the Mann–Kendall test. The trends of annual runoff at each section of the Wei River
Basin are shown in Fig. 6. The stations with significant negative trends of annual runoff are mainly distributed in the West of the Basin. To some extents, the area is consistent with the spatial distribution of the station with significant negative trends in precipitation (Fig. 4). At the section of Xianyang to Huaxian, annual runoff decline is not obvious. 4.5. Abrupt runoff decline The sequential Mann–Kendall test was used to graphically illustrate the forward and backward trends of runoff of the five catchments in the period of 1960–2010 (Fig. 7). Overall, for the whole available period, the Mann–Kendall test results showed a significantly decreasing trend for annual runoff in the five sections of the Wei River, which is statistically significant at the significance of α = 0.05 level. The break points of the annual runoff were shown in Table 4. In general, 1970–2010 is the period with abrupt decline in runoff in the study area. Linjiacun to Xianyang is the catchment with the earliest runoff decline. Runoff decline in the catchment of upstream of Linjiacun, upstream of Zhangjiashan and upstream of Zhuangtou started as early as 1970, whilst drastically accelerated in 1985, 1996 and 1996, respectively. The trend became significant
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
9
Fig. 8. Correlations between precipitation and runoff in the two different periods (pre-development period and post-development period).
after 1986, 2004 and 2004, respectively. Runoff decline in Linjiacun to Xianyang started as early as 1968. It had accelerated since 1989 and the trend became significant after 1996. In the section from Xianyang to Huaxian, runoff decline started in 1964, much earlier than the other 4 sub-catchments mentioned above. However, the decline accelerated in 1966 and it did not have a significant runoff change. 4.6. Relationship between runoff and precipitation Based on the results of the sequential Mann–Kendall test analysis, runoff data were divided into pre-development period and post-development period. Fig. 8 illustrates the correlation between precipitation and runoff in the two periods. In most of the sub-catchments, the correlation between precipitation and runoff in the pre-development period is stronger than that in the postdevelopment period. R2 for pre-development period is always higher than R2 of the post-development. This suggests that in the post-development period, the changes in runoff is less influenced by precipitation, and therefore could mainly be driven by human activities. The regression lines for the pre-development period lie below that for the post-development period. This again implies that, under similar annual precipitation, runoff in the post-development period is less than that in the pre-development period. Thus runoff should be driven by intensifying human activities in
the region. It could therefore be deduced that human activity is the main driving factor of declining runoff in the study area. Regression analysis suggests that, human activities are possibly the main cause of runoff decline. Furthermore, the sequential Mann–Kendall test clearly shows abrupt decline in runoff in most of the sub-catchments after 1970. Thus considerable human activities are considered the most possible drive for runoff decline in the catchments after 1970. The year of 1970 is the beginning of the development of the Wei River Basin. So the increasing economic activities then result in the increasing water consumption. Meanwhile, all kinds of the water conservancy project in the Wei River Basin began to put into operation in the late 1970s, and the water storage and the evaporation also resulted in the decrease of runoff in the Wei River Basin.
5. Conclusions In this study, temporal trends of precipitation, temperature, evaporation as well as their spatial distributions in the Wei River Basin were examined. The temporal and spatial trends of runoff and the relationship between runoff and precipitation in the Wei River Basin were also analyzed. The following main conclusions are drawn:
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
10
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
(1) The temporal variation shows that, there is a very dry period in the Wei River when it entered the 1990s. For all the subcatchments, the period of 1990–2010 has a mean precipitation which is lower than the long-term average. (2) In the period of 1980–1989, the evaporation is the highest compared with the long-term in the five sub-catchments. (3) The temperature shows an increasing trend in the entire period. After the year of 1990, temperature is higher than the long-term average in the five sub-catchments. (4) The runoff at the five sections of the Wei River tends to decrease over the past 51 years. The runoff in the upstream of Linjiacun exhibits the largest declining rate. The decrease in runoff will have a direct impact on the economy and society of the Wei River Basin. (5) Based on the sequential Mann–Kendall test, the precipitation decreased at 6 stations of the Wei River Basin at the α = 0.05 significance level. Temperature increased significantly at all stations at the α = 0.05 significance level. Potential evaporation increased significantly at 50% of the stations in the basin only. The significant increase in air temperature has resulted in a remarkable decrease in runoff through increased evaporation. (6) The Mann–Kendall test analysis shows significant decline in runoff in the sub-catchments of the Wei River Basin in the period of 1960–2010. Amongst the 5 sub-catchments, a runoff slope break (after the year of 1970) is detected for all the subcatchments except the section of Xianyang to Huaxian (1966). (7) Based on the results of the sequential Mann-–Kendall test analysis, runoff data were divided into pre-development period and post-development period. The correlation between precipitation and runoff in the pre-development period is stronger than that in the post-development period, implying that in the post-development period, runoff is less influenced by precipitation, and could therefore be driven by human activities. Regression analysis indicates that human activities are possibly the main cause of runoff decline after 1970. The year of 1970 is the beginning period of the development of the Wei River Basin. The increasing economic activity then results in the increasing water consumption and all kinds of the water conservancy project in the Wei River Basin began to put into operation in the late 1970s, the water storage and the evaporation also led to the decrease of runoff in the Wei River Basin.
Acknowledgements This research was supported by the Special Projects of Ministry of Water resources of China (Grant no. 201101043), Natural Science Foundation of China (51190093, 51179149), and the New Century Talents Program of Ministry of Education of China (NCET-10-0933).
References
FAO, Rome, p. 56. Dai, 2013. Increasing drought under global warming in observations and models. Nat. Déry, S.J., Hernández-Henríquez, M.A., Burford, J.E., Wood, E.F., 2009. Observational evidence of an intensifying hydrological cycle in northern Canada. Geophys. Res. Lett. 36, L13402. http://dx.doi.org/10.1029/2009GL03885. Douglas, E.M., Vogel, R.M., Kroll, C.N., 2000. Trends in floods and low flows in the United States: impact of spatial correlation. J. Hydrol. 240 (1–2), 90–105. Fraiture, Cd, 2007. Integrated water and food analysis at the global and basin level. An application of WATERSIM. Water Resour. Manag. 21, 185–198. Fu, G., Charles, S.P., Viney, N.R., Chen, S., Wu, J.Q., 2007. Impacts of climate variability on streamflow in the Yellow River. Hydrol. Process. 21, 3431–3439. http: //dx.doi.org/10.1002/hyp.6574. Gerstengarbe, F.W., Werner, P.C., 1999. Estimation of the beginning and end of recurrent events within a climate regime. Clim. Res. 11, 97–107. Hao, X.M., Chen, Y., Xu, C., Li, W., 2008. Impacts of climate change and human activities on the surface runoff in the tarim river basin over the last fifty years. Water Resour. Manag. 22, 1159–1171. Huang, S.Z., Chang, J.X., Huang, Q., et al., 2014. Spatio-temporal changes and frequency analysis of drought in the Wei River Basin, China. Water Resour. Manag. 28 (10), 3095–3110. Huang, S.Z., Chang, J.X., Huang, Q., et al., 2014. Spatio-temporal changes in potential evaporation based on entropy across the Wei River Basin. Water Resour. Manag. 28 (13), 4599–4613. Huntington, T.G., 2006. Evidence for intensification of the global water cycle: review and synthesis. J. Hydrol. 319, 83–95. IPCC, 2001. Climate Change 2001: The Scientific Basin. Cambridge University Press, Cambridge, p. 785. IPCC, 2007. Climate change 2007: the physical science basis. In: Solomon, S., Qin, D.,. Kang, B., Lee, S.-J., Kang, D.-H., Kim, Y.-O., 2007. A flood risk projection for Yongdam dam against future climate change. J. Hydro-Environ. Res. 1, 118e125. Kendall, M.G., 1938. A new measure of rank correlation. Biometrika30, 81–93. Kundzewicz, Z.W., Nohara, D., Jiang, T., Oki, T., Su, B.D., Takeuchi, K., 2009. Discharge of large Asian rivers e observations and projections. Quat. Int. 208, 4e10. Labat, D., Goddéris, Y., Probst, J.L., Guyot, J.L., 2004. Evidence for global runoff increase related to climate warming. Adv. Water Resour. 27 (6), 631–642. Lioubimtseva, E., Henebry, G., 2009. Climate and environmental change in arid Central Asia: impacts, vulnerability, and adaptations. J. Arid. Environ. 73 (11), 963–977. Liu, C.M., Xia, J., 2004. Global Water System Project (GWSP), China. o http:www. gwsp.org/chinese_comm.html 4. Luo, X., Guo, X., 2004. Man’s disturbance and contabescence of water environment in Haihe valley China Population. Resour. Environ. 14 (5), 12–16 (in Chinese). Milliman, J.D., Farnsworth, K.L., Jones, P.D., Xu, K.H., Smith, L.C., 2008. Climatic and anthropogenic factors affecting river discharge to the global ocean, 1951–2000. Global Planet. Change 62, 187–194. Moraes, J.M., Pellegrino, H.Q., Ballester, M.V., et al., 1998. Trends in hydrological parameters of southern Brazilian watershed and its relation to human induced changes. Water Resour. Manage. 12, 295–311. Novotny, E.V., Stefan, H.G., 2007. Stream flow in Minnesota: indicator of climate change. J. Hydrol. 334, 319–333. Piao, S., Ito, A., Li, S., Huang, Y., Ciais, P., Wang, X., Peng, S., Andres, R.J., Fang, J., Jeong, S., Mao, J., Mohammat, A., Muraoka, H., Nan, H., Peng, C., Peylin, P., Shi, X., Sitch, S., Tao, S., Tian, H., Xu, M., Yu, G., Zeng, N., Zhu, B., 2012. The carbon budget of terrestrial ecosystems in East Asia over the last two decades. Biogeosci. Discuss. 9 (3), 4025. Sankarasubramanian, A., Vogel, R.M., 2003. Hydroclimatology of the continental United States. Geophys. Res. Lett. 30 (7), 1363. Shrestha, U.B., Gautam, S., Bawa, K.S., 2012. Widespread climate change in the Himalayas and associated changes in local ecosystems. PLoS One 7 (5), e36741. Sneyers, R., 1975. Sur l’ analyse statistique des séries d’ observations. WMO Tech Note. Sprent, P., 1990. Applied Nonparametric Statistical Methods. Chapman and Hall, London. Vorosmarty, C.J., Green, P., Salisbury, J., Lammers, R.B., 2000. Global water resources: vulnerability from climate change and population growth. Science 289, 284–288. Yao, Q., Lei, Y.P., Li, H.J., Wang, X.L., Kang, L.Y., 2008. Study on hydrologic response to climate and land cover change in Bayi Reservoir Basin of Taihan Mountain area. Water Sav. Irrig. 3, 19–23 (in Chinese). Zhu, Y., Day, R.L., 2005. Analysis of streamflow trends and the effects of climate in Pennsylvania, 1971 to 2001. J. Am. Water Resour. Assoc. 41 (6), 1393–1405.
Allen, R.G., Pereira, L.S., Raes, D., et al., 1998. Crop Evapotranspiration Guidelines for Computing Crop Water Requirements, FAO Irrigation and Drainage Paper No.
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
Contents lists available at ScienceDirect
Environmental Research journal homepage: www.elsevier.com/locate/envres
Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin Jing Zhao n, Qiang Huang, Jianxia Chang, Dengfeng Liu, Shengzhi Huang, Xiaoyu Shi State Key Laboratory Base of Eco-Hydraulic Engineering in Arid Area, Xi’an University of Technology, Xi’an 710048, China
art ic l e i nf o
a b s t r a c t
Article history: Received 26 September 2014 Received in revised form 19 December 2014 Accepted 31 December 2014
The Wei River is the largest tributary of the Yellow River in China. The relationship between runoff and precipitation in the Wei River Basin has been changed due to the changing climate and increasingly intensified human activities. In this paper, we determine abrupt changes in hydro-climatic variables and identify the main driving factors for the changes in the Wei River Basin. The nature of the changes is analysed based on data collected at twenty-one weather stations and five hydrological stations in the period of 1960–2010. The sequential Mann–Kendall test analysis is used to capture temporal trends and abrupt changes in the five sub-catchments of the Wei River Basin. A non-parametric trend test at the basin scale for annual data shows a decreasing trend of precipitation and runoff over the past fifty-one years. The temperature exhibits an increase trend in the entire period. The potential evaporation was calculated based on the Penman-Monteith equation, presenting an increasing trend of evaporation since 1990. The stations with a significant decreasing trend in annual runoff mainly are located in the west of the Wei River primarily interfered by human activities. Regression analysis indicates that human activity was possibly the main cause of the decline of runoff after 1970. & 2015 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Keywords: Precipitation The Mann–Kendall test method Change points Trends The Wei River Basin
1. Introduction Climate change is an important factor affecting hydrological cycle, and the hydrological regime could be severely modified in response to the anticipated changes in temperature and precipitation during the present century. The increasing average surface air temperature intensified global hydrological cycles during the 20th century is a general consensus (Huntington, 2006; Milliman et al., 2008; Déry et al., 2009). Air temperature near the surface of the Earth increased by 0.74 °C (IPCC, 2007) over the last century, and the observed air temperature in China has increased by 1.1 °C since the 1980s at an increasing rate of 0.22 °C/decade, which is higher than the global average (Kundzewicz et al., 2009; Piao et al., 2012). Rising global temperatures have been accompanied by changes in weather and climate. Many regions have seen significant changes in precipitation which have led to more floods, droughts, as well as more frequent and severe heat waves (IPCC, 2007; Shrestha et al., 2012; Kang et al., 2007; Dai, 2013). Streamflow as an important hydrological variable is a combination of what happens to precipitation, temperature and other components of hydrological cycle within a large basin scale (Z 100,000 km2); as such, streamflow can be used as an indicator of n
Corresponding author. Fax: þ 86 29 82312797. E-mail address: [email protected] (J. Zhao).
hydrological responses to climate change (Hao et al., 2008; Fu et al., 2007; Sankarasubramanian and Vogel, 2003; Lioubimtseva and Henebry 2009). The increasing shortage of water in most river basins is mainly due to human activities (Vorosmarty et al., 2000). Runoff, an important source of replenishment for surface and groundwater storage, is also declining in most river basins. Therefore, in recent decades, analyzing the change trend and identifying the driving factors on runoff has been the focus of hydrological research (Labat et al., 2004; Liu and Xia, 2004; Fraiture, 2007). The traditional regression method and the hydrological models are usually used to identify the magnitude of impacts of climate change on runoff (e.g., Yao et al., 2008). Additionally, the methods such as the Crammer's method, moving Ttest (MTT), Yamamoto method, Mann–Kendall method and Pettitttest method also can be used to detect abrupt changes in hydrometeorological phenomena. Wei River Basin, the largest tributary of the Yellow River, is an important foodstuff base in China. Natural streamflow as the main source of surface water plays a key role in agricultural irrigation, socio-economic development and local eco-environmental conservation in the basin. Precipitation and temperature are primary atmospheric factors affecting the streamflow. Annual and seasonal precipitation in the Wei river show decreasing trends, especially in summer precipitation, and the temperature shows an increasing trend of 0.3 °C per decade in recent 50 years, obviously exceeding the warming amplitude of global average temperature. Studies
http://dx.doi.org/10.1016/j.envres.2014.12.028 0013-9351/& 2015 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
2
have shown that the runoff of the Wei River has decreased significantly in recent decades. The shortage of water resources in the Wei River Basin is becoming more and more serious. The annual per-capita water availability is far below the recognized standard (IPCC, 2001). The descent rate of precipitation and runoff are extremely inconsistent. This does not only limit socio-economic development in the region, but also poses considerable threat to the health of the ecosystem and the environment at the downstream of the catchment (Luo and Guo, 2004). The variability of streamflow in the Wei River Basin has attracted attentions and interests from academic circles and local government. The objective of this paper is to comprehensively investigate the variability of long-term historical records of climate and hydrological data for the longest time-series and highest number of stations available in the Wei River Basin, to determine runoff slope break by means of the sequential Mann-Kendall test, and then to analyze the main driving factors in related period of abrupt runoff decline.
2. Study area and data 2.1. Study area The Wei River Basin, which originates from the north side of Niaoshu Mountain, is the largest tributary of the Yellow River. With its length of 818 km and the basin area of 13.5 10 km2, the Wei River flows from east to west through the Loess Plateau in East Gansu, the Tianshui Basin, the Baoji Vally, the Guanzhong Basin, and makes the confluence with Yellow River at Tongguan County in Shaanxi Province (Fig 1). The length of the channel is approximately 502 km and the basin area is approximately 6.24 104 km2 in Shaanxi Province before the confluence. As the largest branch of the Yellow River, the average annual runoff of Wei River is 8.09 billion m3 at Huaxian Station. The drainage basin lies between 103.5°E–110.5°E and 33.5°N–37.5°N. Located in the continental monsoon climate zone, the Wei River Basin is characterized by abundant precipitation and high temperature in summer and by rare precipitation and very low temperature in winter. The annual precipitation of the basin is approximately 559 mm (Huang et al., 2014a, 2014b). Topographically, the altitude decreases from the highest northwest mountainous areas to the lowest Guanzhong Plain in the southeast and south portion of the basin. It is worth mentioning that the Guanzhong Plain is designated as a state key economic development zone, acting as a great stimulus to the economic development of surrounding area. Therefore, the Guanzhong
Plain's economic development will directly affect the sustainable development of Shaanxi province's economic society. In recent years, with the development of economy and population growth, the Guanzhong region water demand increases greatly. At present, it is difficult for the local water of Guanzhong region to satisfy the requirement of economic and social development. Given the significance of water security in the Guanzhong plain, a deep understanding of the change in climate and runoff is of great significance in the Wei River Basin. 2.2. Data The following hydro-meteorological data were used in this study. Daily precipitation and temperature data (1960–2010) of 21 National Meteorological Observatory (NMO) stations were provided by the National Climatic Centre of China. The potential evaporation data were estimated by the Penman–Menteith method, which was recommended by the World's Food and Agriculture Organization (FAO) in 1998 (Allen et al., 1998; Huang et al., 2014a, 2014b), and in order to simplify, the potential evaporation is called as evaporation in the article. The location of the stations in the basin is shown in Fig. 2, and their latitudes and longitudes are listed in Table 1 The monthly runoff (1960–2010) of the 5 hydrological stations in the Wei River was provided by the Bureau of Hydrology of the Yellow River Water Resources Commission. The hydrological gauge stations include the Linjiacun station, Xianyang station, Huaxian station, Zhangjiashan station and Zhuangtou station (Table 2).
3. Methods Four statistical methods were used in this study to analyze the spatial variations and temporal trends of the hydro-climatic series. (1) A simple linear regression method, which is a parametric test method, was utilized to test the long-term linear trend. (2) The Mann–Kendall test method, which was originally devised by Mann (1945) as a non-parametric test for detecting trends and the distribution of the test statistic derived by Kendall (1975), was employed to test the non-linear trend as well as the turning point. (3) The Kendall method, which is a non-parametric method for testing the correlation (Kendall, 1938; Sprent, 1990), was applied to test the correlation between runoff, precipitation and temperature. (4) The regression analysis was used to identify the main driving factors of runoff decline in the study area.
Fig. 1. The location of the Wei River Basin in the Yellow River basin.
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
3
Fig. 2. The location of the hydrological and meteorological stations in the Wei River Basin.
Table 1 The information of meteorological stations of Weihe River basin. No.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Stations
Lintao Huajialing Wuqi Guyuan Huanxian Yanan Xiji Pingliang Xifeng Changwu Luochuan Tongchuan Minxian Tianshui Baoji Wugong Xi’an Huashan Foping Zhenan Shangzhou
Location
Elevation (m)
Latitude (°N)
Longitude (°E)
35/21 35/23 36/55 36/00 36/35 36/36 35/58 35/33 35/44 35/12 35/49 35/05 34/26 34/35 34/21 34/15 34/18 34/29 33/31 33/26 33/52
103/51 105/00 108/16 106/16 107/18 109/30 105/43 106/40 107/38 107/48 109/30 109/04 104/01 105/45 107/08 108/13 108/56 110/05 107/59 109/09 109/58
1893 2450 1331 1753 1255 958 1916 1346 1421 1206 1159 978 2315 1141 612 447 397 2064 827 693 742
3.1. The Mann–Kendall (MK) trend test method The Mann–Kendall trend test, commonly known as Kendall tau statistic, is a non-parametric assessment of the significance of monotonic trends of hydro-meteorological variables (Zhu and Day, 2005; Novotny and Stefan, 2007). In the Mann–Kendall test, the null hypothesis H0 states that X1, ⋯Xn are samples of n independent and
identically distributed random variables with no seasonal changes. The alternative hypothesis H1 for a two-sided test defines the distributions of Xk and X j as non-identical for all k , j ≤ n ; with k ≠ j . The test statisticS , with zero mean and computed variance (as in Eq. (3)) is asymptotically normal and can be calculated using Eqs. (1) and (2) n− i
S=
n
∑ ∑
Sgn (x k − x j ) (1)
j=1 k=j+1
where
⎧ 1 xk > x j ⎪ ⎪ S gn(x k − x j ) = ⎨ 0 x k = x j ⎪ ⎪ ⎩−1 x k < x j Var (s) = ⎡⎣n (n − 1)(2n + 5) −
(2)
∑t
t (t − 1)(2t + 5)/18⎤⎦
(3)
The notation t is the extent of the given time and ∑ denotes the summation over the time span. In cases where the sample size n > 10, the standard normal variate Z is computed following Douglas et al. (2000) using:
⎧ ⎪ ⎪ ⎪ Z=⎨ ⎪ ⎪ ⎪ ⎩
S−1 Var (S) 0 S+1 Var (S)
⎫ S > 0⎪ ⎪ ⎪ 0 ⎬ ⎪ S < 0⎪ ⎪ ⎭
(4)
Thus in a two-sided trend test, H0 is accepted if Z ≤ Z α /2 at α level of significance. Positive S value indicates an ‘upward-trend’ while negative S value indicates a ‘downward-trend’.
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
4
Table 2 The information of hydrological stations of Wei River basin. Stations
Linjiacun Xianyang Huaxian Zhangjiashan Zhuangtou
Drainage area (km2)
30661 46827 106498 43216 25645
Time of establishment
1934.1.1 1931.6.10 1935.3.1 1932.1.1 1932.1.1
Location
Period
Longitude(°E)
Latitude (°N)
107/00 108/42 109/42 108/35 108/35
34/21 34/17 34/44 34/35 34/35
1956–2010 1956–2010 1956–2010 1956–2010 1956–2010
Fig. 3. The annual precipitation (a) Evaporation (b) and temperature (c) variations and their mean differences of decadal values to the long-term mean in the Wei River basin.
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎ Table 3 Seasonal variations of the precipitation (mm), evaporation (mm) and temperature (°C) in Wei River basin. Sub-catchment
Season
P(mm)
E(mm)
T(°C)
Upstream of Linjiacun
Spring Summer Autumn Winter
53.1 92.7 20.7 9.4
80.7 110.2 46.4 25.8
14.7 18.4 3.2 0.3
Spring Summer Autumn Winter
58.5 107.7 30.5 14.3
82.5 120.6 49.9 28.8
19.0 22.9 7.1 3.6
Spring Summer Autumn Winter
58.7 106.0 31.9 15.5
110.7 104.4 39.6 42.6
17.4 21.3 5.7 1.9
Upstream of Zhangjiashan
Spring Summer Autumn Winter
45.8 95.8 20.1 9.3
89.8 120.9 50.3 29.1
15.6 19.3 2.8 0.8
Upstream of Zhuangtou
Spring Summer Autumn Winter
49.0 105.5 22.1 11.0
93.6 98.3 49.5 51.2
15.6 19.3 2.7 1.1
Linjiacun to Xianyang
Xianyang to Huaxian
The sequential version of Mann–Kendall test (Sneyers, 1975) was used to test assumptions about the start of a trend within the sample X1, ⋯Xn from set of random variable X based on rank series of progressive and retrograde rows of the sample. The magnitudes of X j annual mean time-series and j = 1, ⋯ , n are compared by Xk ,
5
where k = 1, ⋯ , j − 1. For each comparison, the number of cases Xk > X j is counted and denoted by n j . The test statistic is normally distributed with mean given as
tj =
j
∑1
E (t) =
n j (j = 2, 3, ⋯n)
n (n − 1) 4
(5)
(6)
and variance given as
Var (t j ) =
[j (j − 1)(2j + 5)] 72
(7)
The sequential values of the statistic U (t) are calculated as
U (t) =
t j − E (t) Var (t j )
(8)
which is the forward sequence, and the backward sequence U ′ (t) is calculated using the same equation but in the reverse data series. In two-sided trend test, a null hypothesis is accepted at a significance level if U (t) ≤ U (t)1 − α /2; where U (t)1 − α /2 is the critical value of the standard normal distribution with a probability exceeding α /2. A positive U (t) denotes an upward trend while the reverse denotes a downward trend (i.e., U ′ (t) is similar to U (t)). In this study, α is set at 0.05 significant levels. The sequential Mann–Kendall test enables detection of the approximate time of occurrence of a trend from the intersection point of the forward and backward curves of the test statistic. If the intersection point is significant at α = 0.05, then the critical point of change is at that period (Moraes et al., 1998; Gerstengarbe and Werner, 1999). Hence the sequential Mann–Kendall test is an efficient way by which the starting time of a trend is pinpointed.
Fig. 4. Trends of three variables at 21weather stations in the Wei River Basin 1960–2010. Green regular triangle indicates an increasing trend at the 5% significance level. Red inverse triangle indicates a decreasing trend at the 5% significance level. Blue dot represents station with trends of significance below 10%. P: Precipitation; T: Temperature; E: Evaporation. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
6
Fig. 5. The annual change of runoff in Wei River basin.
3.2. The regression analysis Regression analysis was used to identify the main driving factors of runoff change. Based on the result of the sequential Mann–Kendall test, runoff data for each basin were divided into two periods ‘predevelopment period’ and ‘post-development period’. Data for the predevelopment period (which is the period before abrupt change) of annual runoff were used as baseline for the comparison with the postdevelopment period (i.e. the period after abrupt runoff change). The relationship between precipitation and runoff in the two periods was then analyzed and compared. Comparisons of correlations between runoff and precipitation in the two periods provide a good measure for the effect of human activities on runoff.
4. Results and discussion 4.1. Temporal variation of precipitation, temperature and evaporation in the Wei River basin In order to study the decadal variation of historical annual and seasonal precipitation, temperature and evaporation, the value of the three meteorological elements for the whole study period
(1960–2010) and for each decade (1960–1969,1970–1979,1980– 1989,1990–1999 and 2000–2010) were calculated and compared. The results were illustrated in Fig. 3. The results suggest that: (1) In the upper stream of Linjiacun a wet period (1960–1969) and a dry (1990–1999) period are clearly seen, in which the mean annual precipitation is 49.59 mm (42.30 mm) higher (lower) than the long-term average, respectively. The periods (1990–1999, 2000–2010) are warmer than the long-term average, while the period (1960–1969, 1970–1979, and 1980–1989) is cooler than the long-term average. The maximum and minimum evaporation appeared in 2000–2010(27.88 mm) and 1980–1989( 31.93 mm) compared with the long-term average, respectively. (2) Between Linjiacun and Xianyang, the biggest difference is found in the period (1980–1989), which has a mean precipitation of 64.83 mm higher than the long-term average, and the period (1990–1999) which has a mean precipitation of 69.78 mm lower than the long-term average. The period (2000– 2010) is 0.79 °C warmer than the long-term average, and before 1989 the decadal variations of temperature are lower than the long-term average. The variation of evaporation in this section of the river is larger than other sections. The evaporation of the period (1960–1969) is 43.56 mm higher than the long-term average, and after 1989 the decadal variations of evaporation are very
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
7
Fig. 6. Trends of annual runoff of the 5 sections in the Wei River Basin during the 51 years. Upward (downward) triangles indicate positive (negative) trends from M–K test. Blue dots show sections with no trends or trends below 10% significance level. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
close to the long-term average. (3) For the section of Xianyang to Huaxian, a dry period (1990–1999) and a wet period (1980–1989) are found which have a mean annual value of 55.76 mm (54.90 mm) lower (higher) than the long-term average, respectively. The period (2000–2010) is warmer, while the period (1960– 1989) is cooler than the long-term average. (4) For the upstream of Zhangjiashan the mean decadal changes of precipitation are not remarkable, the biggest difference is found for the period (1960– 1969) which has a mean precipitation of 47.16 mm higher than the long-term average and the precipitation of other periods is the closest to the long-term average. The period (2000–2010) is 0.84 °C warmer than the long-term average, and before 1989 the decadal variations of temperature are lower than the long-term average. The maximum and minimum evaporation appeared in 1970–1979 (38.36 mm) and 1980–1989 ( 53.9 mm) compared with the long-term average, respectively. (5) For the upstream of zhuangtou, a dry period (1990–1999) and a wet period (1960– 1969) are found which have a mean annual value of 44.30 mm (57.40 mm) lower (higher) than the long-term average, while the rest periods is more closer to the long-term average, respectively. The periods (1990–1999 and 2000–2010) are warmer, while the rest of the periods are cooler than the long-term average. For the period (1980–1989), evaporation is 47.80 mm higher than the long-term average (Fig. 3). (6) In terms of seasonal variations, precipitation, temperature and evaporation in summer has larger difference than that of precipitation, temperature and evaporation in other seasons. The maximum value are appeared in the section of Linjiacun to Xianyang in summer (Table 3). The above analysis shows that, there is a very dry period in the Wei River when it entered the 1990s. If this situation does not alter in the 21st century, it will have a serious effect on agriculture, industry and drinking water supply in the Wei River Basin.
4.2. Spatial distribution of precipitation, temperature and evaporation Applying the Mann–Kendall statistics method at each station in the Wei River Basin, temporal trends of annual precipitation, temperature and evaporation are tested at the significance levels of α = 0.05 and α = 0.10. These temporal trends are then interpolated to show their spatial distributions. Fig.4 displays the location of the 21 weather stations and the direction of trends based on the M–K test of three variables of each station. Precipitation decreased at 6 stations around the boundary of the Wei Basin at the α = 0.05 significance level, while precipitation at the other 15 stations in the basin shows the trend of significance below 10%. Temperature increased statistically significantly at all stations of the Wei River Basin at the significance level of α = 0.05. Potential evaporation statistically significantly increased at 50% of the stations around the central basin only, with one exception in the South of the basin. The above analysis indicates that temperature has increased dramatically in the Wei River Basin. The significant increase in air temperature has resulted in a dramatic decrease in runoff caused by increased evaporation (refer to Fig. 5). The Wei River is the economic center of Shaanxi Province and rapid economic development experienced during last decades has increased air pollution in the region, which, in turn, has partly caused temperature increase and runoff decrease. The increasing trend in air temperature over the past decades in the region is consistent with global warming as reported across most of the globe. 4.3. Temporal variation of runoff The 51 years variation of runoff in the Wei River exhibits alternation from abundance to paucity (Fig. 5). The runoff at the five sections of the Wei River Basin tends to decrease over the past 51
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
8
Fig. 7. Sequential Mann–Kendall tests for annual runoff. Dotted horizontal lines represent critical values corresponding to the 95% confidence interval. Table 4 Starting time of runoff trend, abrupt change time, and significant runoff change time in the 5 sections determined by sequential Mann–Kendall test. Stations
Start of trend
Abrupt change
Significant trend
Upstream of Linjiacun Linjiacun to Xianyang Xianyang to Huaxian Upstream of Zhangjiashan Upstream of Zhuangtou
1970 1968 1964 1970 1970
1985 1989 1966 1996 1996
1986 1996 No 2004 2004
years. The annual runoff has been declining at the rate of 0.54, 0.33, 0.07, 0.16 and 0.08 108 m3/year, respectively, from the 1960s to the 2000s. From the 1960s to the 2000s, runoff declined by 86.67%, 53.44%, 16%, 50.24% and 43.62% respectively. The decrease of runoff will have a direct impact on the economy and society of the Wei River Basin. 4.4. Spatial distribution of runoff trends Annual trends of runoff were analyzed with the Mann–Kendall test. The trends of annual runoff at each section of the Wei River
Basin are shown in Fig. 6. The stations with significant negative trends of annual runoff are mainly distributed in the West of the Basin. To some extents, the area is consistent with the spatial distribution of the station with significant negative trends in precipitation (Fig. 4). At the section of Xianyang to Huaxian, annual runoff decline is not obvious. 4.5. Abrupt runoff decline The sequential Mann–Kendall test was used to graphically illustrate the forward and backward trends of runoff of the five catchments in the period of 1960–2010 (Fig. 7). Overall, for the whole available period, the Mann–Kendall test results showed a significantly decreasing trend for annual runoff in the five sections of the Wei River, which is statistically significant at the significance of α = 0.05 level. The break points of the annual runoff were shown in Table 4. In general, 1970–2010 is the period with abrupt decline in runoff in the study area. Linjiacun to Xianyang is the catchment with the earliest runoff decline. Runoff decline in the catchment of upstream of Linjiacun, upstream of Zhangjiashan and upstream of Zhuangtou started as early as 1970, whilst drastically accelerated in 1985, 1996 and 1996, respectively. The trend became significant
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
9
Fig. 8. Correlations between precipitation and runoff in the two different periods (pre-development period and post-development period).
after 1986, 2004 and 2004, respectively. Runoff decline in Linjiacun to Xianyang started as early as 1968. It had accelerated since 1989 and the trend became significant after 1996. In the section from Xianyang to Huaxian, runoff decline started in 1964, much earlier than the other 4 sub-catchments mentioned above. However, the decline accelerated in 1966 and it did not have a significant runoff change. 4.6. Relationship between runoff and precipitation Based on the results of the sequential Mann–Kendall test analysis, runoff data were divided into pre-development period and post-development period. Fig. 8 illustrates the correlation between precipitation and runoff in the two periods. In most of the sub-catchments, the correlation between precipitation and runoff in the pre-development period is stronger than that in the postdevelopment period. R2 for pre-development period is always higher than R2 of the post-development. This suggests that in the post-development period, the changes in runoff is less influenced by precipitation, and therefore could mainly be driven by human activities. The regression lines for the pre-development period lie below that for the post-development period. This again implies that, under similar annual precipitation, runoff in the post-development period is less than that in the pre-development period. Thus runoff should be driven by intensifying human activities in
the region. It could therefore be deduced that human activity is the main driving factor of declining runoff in the study area. Regression analysis suggests that, human activities are possibly the main cause of runoff decline. Furthermore, the sequential Mann–Kendall test clearly shows abrupt decline in runoff in most of the sub-catchments after 1970. Thus considerable human activities are considered the most possible drive for runoff decline in the catchments after 1970. The year of 1970 is the beginning of the development of the Wei River Basin. So the increasing economic activities then result in the increasing water consumption. Meanwhile, all kinds of the water conservancy project in the Wei River Basin began to put into operation in the late 1970s, and the water storage and the evaporation also resulted in the decrease of runoff in the Wei River Basin.
5. Conclusions In this study, temporal trends of precipitation, temperature, evaporation as well as their spatial distributions in the Wei River Basin were examined. The temporal and spatial trends of runoff and the relationship between runoff and precipitation in the Wei River Basin were also analyzed. The following main conclusions are drawn:
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
10
J. Zhao et al. / Environmental Research ∎ (∎∎∎∎) ∎∎∎–∎∎∎
(1) The temporal variation shows that, there is a very dry period in the Wei River when it entered the 1990s. For all the subcatchments, the period of 1990–2010 has a mean precipitation which is lower than the long-term average. (2) In the period of 1980–1989, the evaporation is the highest compared with the long-term in the five sub-catchments. (3) The temperature shows an increasing trend in the entire period. After the year of 1990, temperature is higher than the long-term average in the five sub-catchments. (4) The runoff at the five sections of the Wei River tends to decrease over the past 51 years. The runoff in the upstream of Linjiacun exhibits the largest declining rate. The decrease in runoff will have a direct impact on the economy and society of the Wei River Basin. (5) Based on the sequential Mann–Kendall test, the precipitation decreased at 6 stations of the Wei River Basin at the α = 0.05 significance level. Temperature increased significantly at all stations at the α = 0.05 significance level. Potential evaporation increased significantly at 50% of the stations in the basin only. The significant increase in air temperature has resulted in a remarkable decrease in runoff through increased evaporation. (6) The Mann–Kendall test analysis shows significant decline in runoff in the sub-catchments of the Wei River Basin in the period of 1960–2010. Amongst the 5 sub-catchments, a runoff slope break (after the year of 1970) is detected for all the subcatchments except the section of Xianyang to Huaxian (1966). (7) Based on the results of the sequential Mann-–Kendall test analysis, runoff data were divided into pre-development period and post-development period. The correlation between precipitation and runoff in the pre-development period is stronger than that in the post-development period, implying that in the post-development period, runoff is less influenced by precipitation, and could therefore be driven by human activities. Regression analysis indicates that human activities are possibly the main cause of runoff decline after 1970. The year of 1970 is the beginning period of the development of the Wei River Basin. The increasing economic activity then results in the increasing water consumption and all kinds of the water conservancy project in the Wei River Basin began to put into operation in the late 1970s, the water storage and the evaporation also led to the decrease of runoff in the Wei River Basin.
Acknowledgements This research was supported by the Special Projects of Ministry of Water resources of China (Grant no. 201101043), Natural Science Foundation of China (51190093, 51179149), and the New Century Talents Program of Ministry of Education of China (NCET-10-0933).
References
FAO, Rome, p. 56. Dai, 2013. Increasing drought under global warming in observations and models. Nat. Déry, S.J., Hernández-Henríquez, M.A., Burford, J.E., Wood, E.F., 2009. Observational evidence of an intensifying hydrological cycle in northern Canada. Geophys. Res. Lett. 36, L13402. http://dx.doi.org/10.1029/2009GL03885. Douglas, E.M., Vogel, R.M., Kroll, C.N., 2000. Trends in floods and low flows in the United States: impact of spatial correlation. J. Hydrol. 240 (1–2), 90–105. Fraiture, Cd, 2007. Integrated water and food analysis at the global and basin level. An application of WATERSIM. Water Resour. Manag. 21, 185–198. Fu, G., Charles, S.P., Viney, N.R., Chen, S., Wu, J.Q., 2007. Impacts of climate variability on streamflow in the Yellow River. Hydrol. Process. 21, 3431–3439. http: //dx.doi.org/10.1002/hyp.6574. Gerstengarbe, F.W., Werner, P.C., 1999. Estimation of the beginning and end of recurrent events within a climate regime. Clim. Res. 11, 97–107. Hao, X.M., Chen, Y., Xu, C., Li, W., 2008. Impacts of climate change and human activities on the surface runoff in the tarim river basin over the last fifty years. Water Resour. Manag. 22, 1159–1171. Huang, S.Z., Chang, J.X., Huang, Q., et al., 2014. Spatio-temporal changes and frequency analysis of drought in the Wei River Basin, China. Water Resour. Manag. 28 (10), 3095–3110. Huang, S.Z., Chang, J.X., Huang, Q., et al., 2014. Spatio-temporal changes in potential evaporation based on entropy across the Wei River Basin. Water Resour. Manag. 28 (13), 4599–4613. Huntington, T.G., 2006. Evidence for intensification of the global water cycle: review and synthesis. J. Hydrol. 319, 83–95. IPCC, 2001. Climate Change 2001: The Scientific Basin. Cambridge University Press, Cambridge, p. 785. IPCC, 2007. Climate change 2007: the physical science basis. In: Solomon, S., Qin, D.,. Kang, B., Lee, S.-J., Kang, D.-H., Kim, Y.-O., 2007. A flood risk projection for Yongdam dam against future climate change. J. Hydro-Environ. Res. 1, 118e125. Kendall, M.G., 1938. A new measure of rank correlation. Biometrika30, 81–93. Kundzewicz, Z.W., Nohara, D., Jiang, T., Oki, T., Su, B.D., Takeuchi, K., 2009. Discharge of large Asian rivers e observations and projections. Quat. Int. 208, 4e10. Labat, D., Goddéris, Y., Probst, J.L., Guyot, J.L., 2004. Evidence for global runoff increase related to climate warming. Adv. Water Resour. 27 (6), 631–642. Lioubimtseva, E., Henebry, G., 2009. Climate and environmental change in arid Central Asia: impacts, vulnerability, and adaptations. J. Arid. Environ. 73 (11), 963–977. Liu, C.M., Xia, J., 2004. Global Water System Project (GWSP), China. o http:www. gwsp.org/chinese_comm.html 4. Luo, X., Guo, X., 2004. Man’s disturbance and contabescence of water environment in Haihe valley China Population. Resour. Environ. 14 (5), 12–16 (in Chinese). Milliman, J.D., Farnsworth, K.L., Jones, P.D., Xu, K.H., Smith, L.C., 2008. Climatic and anthropogenic factors affecting river discharge to the global ocean, 1951–2000. Global Planet. Change 62, 187–194. Moraes, J.M., Pellegrino, H.Q., Ballester, M.V., et al., 1998. Trends in hydrological parameters of southern Brazilian watershed and its relation to human induced changes. Water Resour. Manage. 12, 295–311. Novotny, E.V., Stefan, H.G., 2007. Stream flow in Minnesota: indicator of climate change. J. Hydrol. 334, 319–333. Piao, S., Ito, A., Li, S., Huang, Y., Ciais, P., Wang, X., Peng, S., Andres, R.J., Fang, J., Jeong, S., Mao, J., Mohammat, A., Muraoka, H., Nan, H., Peng, C., Peylin, P., Shi, X., Sitch, S., Tao, S., Tian, H., Xu, M., Yu, G., Zeng, N., Zhu, B., 2012. The carbon budget of terrestrial ecosystems in East Asia over the last two decades. Biogeosci. Discuss. 9 (3), 4025. Sankarasubramanian, A., Vogel, R.M., 2003. Hydroclimatology of the continental United States. Geophys. Res. Lett. 30 (7), 1363. Shrestha, U.B., Gautam, S., Bawa, K.S., 2012. Widespread climate change in the Himalayas and associated changes in local ecosystems. PLoS One 7 (5), e36741. Sneyers, R., 1975. Sur l’ analyse statistique des séries d’ observations. WMO Tech Note. Sprent, P., 1990. Applied Nonparametric Statistical Methods. Chapman and Hall, London. Vorosmarty, C.J., Green, P., Salisbury, J., Lammers, R.B., 2000. Global water resources: vulnerability from climate change and population growth. Science 289, 284–288. Yao, Q., Lei, Y.P., Li, H.J., Wang, X.L., Kang, L.Y., 2008. Study on hydrologic response to climate and land cover change in Bayi Reservoir Basin of Taihan Mountain area. Water Sav. Irrig. 3, 19–23 (in Chinese). Zhu, Y., Day, R.L., 2005. Analysis of streamflow trends and the effects of climate in Pennsylvania, 1971 to 2001. J. Am. Water Resour. Assoc. 41 (6), 1393–1405.
Allen, R.G., Pereira, L.S., Raes, D., et al., 1998. Crop Evapotranspiration Guidelines for Computing Crop Water Requirements, FAO Irrigation and Drainage Paper No.
Please cite this article as: Zhao, J., et al., Analysis of temporal and spatial trends of hydro-climatic variables in the Wei River Basin. Environ. Res. (2015), http://dx.doi.org/10.1016/j.envres.2014.12.028i
