Reference evapotranspiration changes in China: natural processes or ...

Theor Appl Climatol (2011) 103:479–488 DOI 10.1007/s00704-010-0315-6

ORIGINAL PAPER

Reference evapotranspiration changes in China: natural processes or human influences? Qiang Zhang & Chong-Yu Xu & Xiaohong Chen

Received: 20 January 2010 / Accepted: 28 July 2010 / Published online: 9 August 2010 # Springer-Verlag 2010

Abstract In this study, we systematically analyze the changing properties of reference evapotranspiration (ETref) across China using Penman–Monteith (P-M) method, exploring the major sensitive meteorological variables for ETref, and investigating influences of human activities, mainly urbanization in this study, on ETref changes in both space and time. We obtain some important conclusions: (1) decreasing annual and seasonal ETref is observed in the east, south and northwest China. However, a long strip lying between these regions is identified to be characterized by increasing ETref; (2) in the regions east to 100°E, the net total solar radiation is the main cause behind the decreasing ETref. In northwest China, however, relative humidity is recognized as the most sensitive variable for the ETref; (3) in the east and south China, urbanization greatly influences the ETref by directly decreasing net solar radiation. The increased air pollution and aerosols in the highly urbanized regions are the main driving factors causing decreasing net radiation; and (4) this study reveals accelerating hydrological cycle from south to north China. Besides, increasing ETref in the source regions of large rivers in China may pose new challenges for the basin-scale water resource management. The results of this Q. Zhang (*) : X. Chen Department of Water Resources and Environment, Sun Yat-sen University, Guangzhou 510275, China e-mail: [email protected] Q. Zhang : X. Chen Key Laboratory of Water Cycle and Water Security in Southern China of Guangdong High Education Institute, Sun Yat-sen University, Guangzhou 510275, China C.-Y. Xu Department of Geosciences, University of Oslo, P O Box 1047, Blindern, 0316 Oslo, Norway

study highlight the integrated effects of climate changes and human activities on ETref changes in different regions of China, which will be of great scientific and practical merits in in-depth understanding of hydrological cycle alterations under the changing environment in China.

1 Introduction Reference evapotranspiration (ETref) is one of the most important hydrological components for scheduling irrigation systems, preparing input data to hydrological water-balance models, assessing hydrological impacts of climate changes (Blaney and Criddle 1950; Xu and Li 2003; Xu and Singh 2005; Gong et al. 2006). Besides, what is important is that estimation of actual evapotranspiration rate for a specific crop requires first calculating potential or reference evapotranspiration (ETp or ETref) and then applying the proper crop coefficients (Kc) to estimate actual crop evapotranspiration (Xu et al. 2006). ETref is a kind of measure of the evaporative demand of the atmosphere independent of crop type, crop development and management practices and is affected only by climatic factors. Therefore, ETref is a climatic parameter and can be computed directly from meteorological data (Allen et al. 1998). There are some methods available for estimation of ETref (e.g., Xu and Singh 2002). Wherein, the Penman– Monteith (P-M) approach was recommended by FAO (e.g., Allen et al. 1998) as a standard tool to calculate ETref. The P-M approach is a physically based technique and can be used globally without any need for additional adjustments of parameters. Xu et al. (2006) studied the changing properties of Penman–Monteith ETref in the Yangtze River basin, showing that the ETref changes are in good line with those of pan evaporation in both space and time. Gong et al.

480

(2006) analyzed sensitivity of the P-M ETref to key climatic variables in the Yangtze River basin, indicating that relative humidity was the most sensitive variable, followed by shortwave radiation, air temperature, and wind speed. Evapotranspiration is the bridge connecting energy balance and water balance (Xu et al. 2005), being one of the most active hydrological components and is heavily influenced by land use changes and climate variations regionally and globally. The currently well-evidenced global warming characterized by increasing temperature has the potential to alter the hydrological cycle and therefore causes uneven distribution of water resources. Hydrologists and meteorologists suggested that an increase in surface temperature can result in higher evaporation rates and enables the atmosphere to transport higher amounts of water vapor, which, in turn, leads to accelerated hydrological cycle (e.g., Menzel and Bürger 2002). Warmer temperature increases the holding capacity of water vapor of atmosphere and which can cause higher probability of rainstorm or high-intensity precipitation and triggers occurrence of flood and drought hazards. Climatic changes because of global warming might result in increase and intensification of extreme events (WMO 2003). Groisman et al. (1999) demonstrated that the probability of daily precipitation exceeding 50.8 mm in mid-latitude countries (the USA, Mexico, China, and Australia) increased by about 20% in the later twentieth century. Suppiah and Hennessy (1998) pointed out that the heavy precipitation events in most parts of Australia have increased. In China, Zhai et al. (1999) indicated that the intensive precipitation events have increased in the west China since 1950. Zhang et al. (2008) found increasing precipitation intensity in the middle and lower Yangtze River basin. Increasing precipitation concentration was also observed in the lower Pearl River basin and which was attributed to increasing air surface temperature (Zhang et al. 2009a). Therefore, we can say that the increasing temperature has exerted tremendous influences on hydrological cycle. In-depth study of evapotranspiration changes in both time and space can shed light on the way and degree to which the climate changes impact the hydrological cycle. Generally, it can be easily and readily accepted that increasing temperature can cause increasing evaporation (Robock et al. 2000; Taikan and Shinjiro 2005). However, observations indicate decreasing pan evaporation and ETref (Peterson et al. 1995; Chattopadhyay and Hulme 1997; Brutsaert and Parlange 1998; Roderick and Farquhar 2002, 2004; Michael et al. 2004; Xu et al. 2006). This is usually known as pan evaporation paradox (e.g., Brutsaert and Parlange 1998). The evaporation paradox is a very important scientific problem and was warmly discussed by many scholars. Peterson et al. (1995) suggested that the downward trend in pan evaporation over most of the USA and former Soviet Union implies decreasing terrestrial

Q. Zhang et al.

evaporation component of the hydrological cycle. Roderick and Farquhar (2002) advocated that the decrease in evaporation is consistent with the observed widespread decreases in sunlight resulting from increasing cloud coverage and aerosol concentration. However, Brutsaert and Parlange (1998) held different viewpoints, pointing out that “in non-humid environments, measured pan evaporation is not a good measure of potential evaporation; moreover, in many situations, decreasing pan evaporation actually provides a strong indication of increasing terrestrial evaporation.” The foregoing discussions imply that, due to the complex nature of evaporation, our knowledge of evaporation and related causes are still terribly limited. Actually, the evaporation changes are the integrated consequences of more than one influencing factors. Biased conclusions can be obtained if we focus on one or two influencing factors only. Besides, the human-induced impacts such as influences from urbanization, aerosol changes on evaporation changes are seldom discussed. Zhang et al. (2004) indicated that the decreasing net total radiation is mainly due to increased air pollution, implying evident human influences on net solar radiation. Xu et al. (2006) pointed out that the most important predictor for the decreasing trend in the ETref and pan evaporation in the Yangtze River basin is the net total radiation followed by wind speed, but they did not discuss the reasons that the wind speed is significantly decreasing. Therefore, they pointed out the direction for the further study. This is the major motivation of this current study. In this study, we analyze seasonal trends of ETref and major meteorological variables, e.g. wind speed, temperature and so on. Comparison between the spatial distribution of large cities with population density of >100 people/km2 and changes of meteorological variables is also performed with aim to investigate influences of urbanization on changes of meteorological variables and ETref. This study will shed light on possible human-induced influences on spatio-temporal variations of meteorological components including ETref, clarifying how climate changes and human activities work together to impact the ETref changes in both time and space. This study also helps to understand ETref variations under the changing environment and their implications for hydrological cycle under the background of global warming. In this case, the objectives of this study are: (1) to explore annual and seasonal changes of ETref over China; (2) to clarify the factors exerting greater impacts on ETref changes in both time and space; (3) to investigate role of human activities in ETref changes and other major meteorological variables by comparing spatial patterns of large cities in China and those of climate variables; and (4) to highlight the possible implications of ETref changes for water resource management under the changing environment in China.

Reference evapotranspiration changes in China

2 Data In this study, we analyze daily meteorological data at 590 stations. Locations of the meteorological stations are shown in Fig. 1. The data include maximum and minimum temperature and daily mean air temperature at 2 m height above the ground, wind speed, relative humidity, sunshine hours, vapor pressure covering 1960–2005. They have been provided by the National Climate Center of the China Meteorological Administration. The missing data are filled up by building regression relations between neighboring stations. The correlation coefficient, R2, is >0.85 and is even as high as >0.98. Therefore, the filled series can satisfy the quality requirements of this study.

3 Methods 3.1 The P-M method The P-M method has been recommended as the sole standard method for computation of ETref by FAO (Allen et al. 1998) and was introduced with good details by Xu et al. (2006) and Gong et al. (2006). For the sake of completeness of this paper, we briefly introduce this method in this section. The reason this method was widely used and was chosen in this study is that this method is physically based and explicitly incorporates both physiological and aerodynamic parameters. The ETref can be computed as: ETref ¼

0:408ΔðRn GÞ þ g Ta900 þ273 u2 ðes ea Þ Δ þ gð1 þ 0:34u2 Þ

Fig. 1 Locations of the meteorological stations considered in this study and the ten drainage basins. The solid dots denote the rain gauging stations. The gray solid dots denote large cities with population density of >100 people/km2. Numbers denote the ten drainage basins: 1 SongHuajiang River, 2: Liaohe River, 3 Haihe River, 4 Yellow River, 5 Huaihe River, 6 Yangtze River, 7 SE rivers (rivers in the southeast China), 8 Pearl River, 9 SW rivers (rivers in the southwest China);,10 NW rivers (rivers in the northwest China)

481

where ETref is the reference evapotranspiration (mmday−1), Rn the net radiation at the crop surface (MJm−2 day−1), G is the soil heat flux density (MJm−2 day−1), T the mean daily air temperature at 2 m height (°C), u2 the wind speed at 2 m height (ms−1), es the saturation vapor pressure (kPa), ea the actual vapor pressure (kPa), es −ea the saturation vapor pressure deficit (kPa), Δ the slope of the vapor pressure (kPa °C−1), γ the psychrometric constant (kPa °C−1). The unit conversion between MJ and watt is: 1,000 Wattsh= 3.6 MJ. The computation procedure follows that given in Chapter 3 of the FAO paper 56 (Allen et al. 1998). 3.2 Trend test Significance of the trends of the meteorological series is evaluated by the Mann–Kendall trend test technique (MK test). The rank-based nonparametric Mann–Kendall test (Mann 1945; Kendall 1975) can test trends of a time series without requiring normality or linearity (Wang et al. 2008), and was therefore highly recommended for general use by the World Meteorological Organization (Mitchell et al. 1966). It was widely used in detection of trends in hydrological (e.g. Zhang et al. 2006) and meteorological series (Zhang et al. 2009a). This paper also uses the Mann– Kendall (MK) test method to analyze trends within the meteorological series. The confidence level used in this study is 95%.

4 Results and discussions 4.1 Trends of reference evapotranspiration Figure 2 illustrates spatial patterns of temporal changes of ETref. Discernable spatial patterns of annual ETref are identified from Fig. 2a. Haihe River basin, Huaihe River basin, the middle and lower Yangtze River basin, the SE rivers and the Pearl River basin are dominated by significantly decreasing ETref. The NW rivers are also characterized by significantly decreasing ETref except eight stations showing significantly increasing ETref. Generally, three regions could be identified with different changing properties of ETref (Fig. 2a): (1) the east China, the middle and the south China. These regions are dominated by significantly decreasing ETref except several stations located along the coastal regions of the SE China which are characterized by not significant ETref changes; (2) the northwest China. This place is again featured by significantly decreasing ETref; and (3) a strip lying between these two regions in the SW-NE direction. The stations along this strip are characterized by not significant ETref changes. Stations with significantly increasing ETref mainly concentrate in the upper Yellow River basin, the upper Yangtze River basin and in the SW rivers. The locations of

482

these rivers can be referred to Fig. 1. Besides, some stations characterized by significantly increasing ETref are also found in the SongHuajiang River and the northeast China. It can be observed from Fig. 2b that, in summer, similar spatial patterns of ETref could be identified when compared to annual changes (Fig. 2a). The difference is that fewer (more) stations are characterized by significantly increasing (decreasing) ETref in the strip between the southeast and the northwest China in comparison with annual changes (Fig. 2a). Only seven stations show significantly increasing

Q. Zhang et al.

ETref. This result suggests decreasing evaporation capacity in summer. Figure 2c illustrates distinctly different spatial patterns of ETref changes over China when compared to annual changes of ETref and ETref variations in summer. One remarkable difference is that the NW and the SE China are characterized by not significant ETref changes. Besides, some stations in these two regions are featured by significantly increasing ETref. Therefore, comparatively, the ETref in these two regions, i.e., the NW and the SE China, increases in winter when compared to annual changes and ETref changes in summer. As far as ETref changes at stations in the strip are concerned, more stations are found to be characterized by significantly increasing ETref, and so do the ETref changes in the Songhuajiang River. Thus, based on what aforementioned, the evaporation capacity of winter increased over the entire China. ETref changes are the results of many influencing factors, climate changes or human influences. In the following sections, we will discuss the effects of various influencing factors on ETref variations in both space and time. 4.2 Sensitivity of ETref changes to meteorological variables Xu et al. (2006) evaluated the sensitivity of ETref to major meteorological variables by scenario analysis, i.e., they generated seven scenarios for each meteorological variable using the following equation: X ðtÞ ¼ X ðtÞ þ ΔX ; ΔX ¼ 0; 10%; 20%; 30% of X ðtÞ

Fig. 2 Spatio-temporal patterns of reference evapotranspiration changes over China. a annual changes of reference evapotranspiration; b changes of reference evapotranspiration in summer; and c changes of reference evapotranspiration in winter. Thick solid contours indicate significant increasing trend; thin solid contours show changes not significant at >95% confidence level; and dashed contours show significant decreasing trends

where X is the meteorological variable, and t is the time in day. The results of this method should be discussed based on the actual variations of meteorological variables and ETref changes. Xu et al. (2006) showed higher sensitivity of ETref to relative humidity followed by net radiation, air temperature and wind speed. However, the contribution of relative humidity to the decreasing trend in the ETref in the Yangtze River basin is much smaller than that of the net radiation in that the trends of the relative humidity over the whole Yangtze River basin are not significant (Xu et al. 2006). They suggested that the net total radiation be the main cause behind the decreasing trend of the ETref because it is not only one of the most sensitive variables but also a variable with significantly decreasing trend. In this study, we evaluate sensitivity of ETref to meteorological variables by using the correlation coefficients between their MK trends. The main idea behind this method is that the higher the similarity between the changing trends of the independent and dependent variables, the higher the sensitivity of the dependent variable to the independent variable. Our analysis results show larger correlation coefficient (r=0.55)


483

between trends of ETref and net radiation on the annual basis (Table 1 and the upper panel of Fig. 3) than the rest three meteorological variables: −0.49 for the relative humidity (the absolute value is 0.49); 0.43 for wind speed; and 0.33 for temperature. Therefore, on annual basis, the net radiation (r=0.55) is the most sensitive meteorological variable for the ETref followed by relative humidity (r=−0.49). This conclusion is in good line with that of Xu et al. (2006). Therefore, the method used in this study has the similar performance as that used by Xu et al. (2006). The difference is that computation procedure of the method in this current study is relatively simpler than that of Xu et al. (2006). As for the meteorological variables the ETref is sensitive to at the stations located along the strip lying between the SE and the NW China, Fig. 4 shows that the wind speed could be the meteorological variables exerting greater influences on the ETref changes, and is followed by the net radiation (Table 1). Table 1 shows that correlations between ETref, wind speed and net radiation are statistically significant at >99% confidence level. 4.3 Meteorological variables influencing ETref changes and possible impacts from human activities Figure 5 illustrates spatial distribution of wind speed changes over China. Generally, China is dominated by significantly decreasing wind speed. Tens of stations showing significantly increasing wind speed distribute

sporadically between 100°E and 120°E. No fixed and discernable spatial patterns can be identified. Relations between ETref and wind speed in space are also ambiguous. Therefore, we do not intend to attribute ETref changes to wind speed variations based on the foregoing analysis results. A closer look at Fig. 5 indicates that the stations characterized by significantly increasing wind speed are mostly outside of regions covered by large cities denoted by large gray dots. This result may imply possible influences of urbanization on measured ground surface wind speed; at least it is true in the regions south to 40°N and east to 100°E. Figure 6 illustrates spatial patterns of net radiation. On the annual basis, regions of China east to 100°E are dominated by significantly decreasing net radiation. Significantly increasing net radiation can be found in the upper Yangtze River basin, the upper Yellow River basin, the west parts of the NW rivers, the west parts of the NW China and also in the north corner of the NW China (Fig. 6a). Figure 2a demonstrates that the places dominated by significantly increasing net radiation are roughly also those characterized by significantly increasing annual ETref. Regions in the east and the south China which are dominated by significantly decreasing net radiation (Fig. 6a) correspond well to those featured by significantly decreasing annual ETref (Fig. 2a). Figure 6b shows that significant decrease of net radiation prevails across major parts of China, only a couple of stations in the upper Yangtze River basin and the west corner of China show significantly increasing net radiation. Similarly, ETref in summer shows

Table 1 Sensitivity of ETref to meteorological variables evaluated by correlation coefficients between MK trends of ETref and the four meteorological variables, i.e. net solar radiation, relative humidity, wind speed and temperature

Annual changes of ETref Net solar radiation Relative humidity Wind speed Temperature ETref in summer Net solar radiation Relative humidity Wind speed Temperature ETref in winter Net solar radiation Relative humidity Wind speed Temperature

At stations in the SE China

At stations located along the strip

At stations in the NW China

Correlation coefficient

p values


p values


p values

0.55 −0.49 0.43 0.33

0.00 0.00 0.00 0.00

0.42 −0.30 0.59 0.11

0.00 0.00 0.00 0.22

0.25 −0.53 0.79 0.16

0.06 0.00 0.00 0.28

0.76 −0.45 0.43 0.40

0.00 0.00 0.00 0.00

0.43 −0.43 0.63 0.30

0.00 0.00 0.00 0.00

0.19 −0.53 0.86 0.27

0.19 0.00 0.00 0.06

0.27 −0.54 0.64 0.07

0.00 0.00 0.00 0.42

−0.32 −0.13 0.56 −0.06

0.00 0.13 0.00 0.52

−0.12 −0.21 0.60 −0.03

0.40 0.15 0.00 0.85

The p values show the significant level of the correlation coefficients. The p value of 95% confidence level; and dashed contours show significant decreasing trends

1 0

0.63

−1 −2 2 1 0

0.56

−1 −2

NR

RH

WS

TP

Climate variables Fig. 4 Correlations between reference evapotranspiration and four major meteorological variables, i.e., NR net solar radiation (MJm−2 day−1); RH relative humidity (%), WS wind speed (m/s), and TP temperature in the belt lying between southeastern China and northwestern China. The upper panel is annual, middle summer and bottom winter

the net radiation in winter when compared to those in summer (Fig. 6b) and annual changes of the net radiation (Fig. 6c). The major difference is that a majority of stations located along the strip between the SE and the NW China are characterized by significantly increasing net radiation. However, changes of radiation in the SE China are not significant. Besides, spatial distribution of significantly decreasing net radiation in the east and the south China match well that of large cities. Correspondingly, the regions in the strip are dominated by significantly increasing net radiation and ETref. Changes of net radiation and ETref in the east and south China are not significant (Figs. 2c and 6c).


485 2 1 0

0.79


−1 −2 2 1 0

0.86

−1 −2 2 1 0

0.60

−1 −2

NR

RH

WS

TP

Time variables Fig. 7 Correlations between reference evapotranspiration and four major meteorological variables in the northwest China. NR net solar radiation (MJm−2 day−1); RH relative humidity (%); WS wind speed (m/s); and TP temperature. The upper panel is annual, middle summer and bottom winter

Fig. 6 Spatio-temporal patterns of net radiation variations over China. a Annual changes of net radiation; b changes of net radiation in summer; and c changes of net radiation in winter. Thick solid contours indicate significant increasing trend; thin solid contours denote changes not significant at >95% confidence level; and dashed contours show significant decreasing trends

As for underlying causes behind the ETref in the NW China, Table 1 and Fig. 7 show two sensitive meteorological variables for ETref, i.e., the wind speed and the relative humidity. In winter, only the winter speed could be accepted as the sensitive meteorological variables for ETref in the NW China. We demonstrate in the aforementioned sections that the wind speed in China is decreasing and no obvious relations could be identified between ETref and the wind speed. The results of this study also provided no evidence for good relations between ETref and wind speed. Therefore, we try to analyze spatio-temporal patterns of relative

humidity and ETref in NW China. Figure 8a shows annual variations of relative humidity in the NW China. It can be observed from Fig. 8a that majority of the NW China are characterized by significantly increasing relative humidity. Northeast parts of the NW China are featured by not significant relativity humidity changes. Similar changes of relative humidity are observed in the NW China. ETref in the NW China is in significantly decreasing trends. Table 1 also indicates significant negative correlations between ETref and the relative humidity in the NW China. In winter, not significant changes of the relative humidity are detected in the NW China except that several stations distribute sporadically in the west parts of the NW China showing significantly increasing relative humidity (Fig. 8c). Figure 2c shows not significant ETref changes in winter. All these results clearly indicate remarkable influences of relative humidity changes on ETref in the NW China. In the regions east to 110°E, ETref changes are heavily influenced by the net radiation variations, and the decreasing trend of the net total radiation is the main cause behind the decreasing ETref. Urbanization, to a certain degree, plays an important role in decrease of the net radiation due to the fact that the spatial distribution of decreasing net radiation matches well that of large cities with population density of >100 persons/km2. Studies by Zhang et al. (2004) and Liu et al. (2004) attributed the decreasing net total radiation to increased air pollution or aerosol in highly urbanized regions in the east China. Roderick and Farquhar (2002), Brutsaert and Parlange (1998) also advocated that the decrease in

486

Q. Zhang et al.

evaporation trend is not determined by temperature alone. Increasing temperature does not necessarily cause increasing evaporation due to the changes in other dominant variables. Increasing ETref in winter when compared to ETref in summer in this study is good evidence. Furthermore, our previous (Zhang et al. 2009b) studies indicate significantly increasing temperature in the NW China, particularly in winter. But ETref changes in the NW China are not significant. Zhang et al. (2009c) studied the changing properties of precipitation in both space and time over China and found increasing annual and summer precipitation. Therefore, increasing relative humidity in the NW China could be due to the increasing precipitation. NW China is characterized by arid climate. Increasing precipitation could lead to increasing relative humidity. Thus, decreasing ETref in the NW China could be accounted for based on the aforementioned observations.

5 Conclusions We analyze spatio-temporal changes of ETref across China by using P-M approach. Underlying causes behind ETref changes are also thoroughly investigated by using sensitivity analysis of ETref to major meteorological variables, i.e., temperature, relative humidity, net radiation, and wind speed. Besides, influences of human activities, mainly urbanization in this study, on ETref changes are also discussed based on the analysis results of this study and also based on the previous studies. Interesting and important conclusions are obtained as the follows.

Fig. 8 Spatio-temporal patterns of relative humidity variations over China. a Annual changes of relative humidity; b changes of relative humidity in summer; and c changes of relative humidity in winter. Thick solid contours indicate significant increasing trend; thin solid contours denote changes not significant at >95% confidence level; and dashed contours show significant decreasing trends

evaporation could be attributed to the observed large and widespread decreases in sunlight resulting from increasing cloud coverage and aerosol concentration. The observations in this study tend to further corroborate the aforementioned perspectives. Besides, our analysis also points out no significant correlations between temperature and ETref. Table 1 indicates that the influences of temperature changes on ETref could almost be ignored and it was particularly true for winter and for annual ETref changes. Therefore, the results of this study seem to be in agreement with those by Ohmura and Wild (2002) that the changing direction of the

1. Significantly decreasing trends of annual and summer ETref are found mainly in the Haihe River, Huaihe River, the middle and the lower Yangtze River, the SE rivers, the Pearl River, and the NW rivers. Significantly increasing ETref could be observed mainly at the stations located along a long strip lying between the NW China and the SE China. Besides, increasing ETref is detected in winter. Decreasing summer ETref is observed along a long strip lying between the NW China and the SE China. 2. Different meteorological variables to which the ETref is sensitive are identified in different parts of China. In the regions east to 100°E, changes of net radiation have the remarkable contribution to the ETref variations. However, in the NW China, relative humidity seems to have greater impacts on ETref changes than other meteorological variables. As for human influences, stations characterized by significantly increasing wind speed seem to be away from large cities, which might imply influences of urbanization on wind speed on regional scale. Besides, parts of China east to 100°E


covered by large cities are also characterized by significantly decreasing net radiation and vice versa, showing remarkable impacts of urbanization on net radiation. Study results (e.g. Zhang et al. 2004; Liu et al. 2004) attributed decreasing net radiation in highly urbanized regions to air pollution and aerosols. The strip lying between the east China and the NW China is characterized by increasing net radiation and ETref which further ascertains the influences of both the net radiation and urbanization-induced decreasing radiation on ETref. In NW China, relative humidity is the major meteorological variable having large contribution to ETref changes. Our previous studies (Zhang et al. 2009b, c) demonstrated increasing temperature and precipitation in the NW China, and which should be regarded as the main cause for increasing relative humidity and decreasing ETref. Considering increasing ETref along the strip between the east China and the NW China, we can say that the hydrological cycle comes to be accelerated from the south to north China. 3. This study confirms the remarkable influences of human activities, mainly the urbanization, on ETref by directly influencing changes of net radiation. Increasing ETref is mainly observed in the upper Yangtze River basin, the Yellow River basin and the southwest parts of China. These regions act as the major streamflow source of large rivers in China, though precipitation also plays important role in streamflow production in the middle and lower reaches of these rivers. Therefore, altered hydrological cycle as an integrated result of climate changes and human activities should be taken into account with great cautions when policy for river basin-scale water resource management is making. This study clarifies different sensitive meteorological variables for ETref and also ascertains the way and degree to which human activities impact ETref. Besides, ETref changes in both time and space in China are thoroughly investigated. All these important conclusions and study results are expected to shed light on hydrological cycle changes under the changing environment in China and also help to pave the way for the similar studies in other places of the world.

Acknowledgments This work was financially supported by the ‘985 Project’ (Grant No.: 37000-3171315), the Program for Outstanding Young Teachers of the Sun Yat-sen University (Grant No.: 200937000-1132381), Xinjiang Technology Innovative Program (Grant No.: 200931105), the State Key Laboratory of Hydrology—Water Resources and Hydraulic Engineering (Grant No.: 2009491511), the Postdoctoral Foundation of the Guangdong Province (Grant No.: 2009-37000-4203384), and by the 111 Project under Grant B08048, Ministry of Education and State Administration of Foreign Experts Affairs, P. R. China. Cordial gratitude should be extended to the editor, Prof. Dr. Hartmut Grassl, and anonymous reviewers for their

487 pertinent and professional comments and suggestions which greatly improved the quality of this manuscript.

References Allen RG, Pereira LS, Raes D, Smith M (1998) Crop evapotranspiration— guidelines for computing crop water requirements—FAO Irrigation and Drainage Paper 56. FAO, 1998. ISBN 92-5-104219-5 Blaney HF, Criddle WD (1950) Determining water requirements in irrigated area from climatological irrigation data. US Department of Agriculture, Soil Conservation Service Technical Paper No. 96, p 48 Brutsaert W, Parlange MB (1998) Hydrologic cycle explains the evaporation paradox. Nature 396:30 Chattopadhyay N, Hulme M (1997) Evaporation and potential evapotranspiration in India under conditions of recent and future climate change. Agric For Meteorol 87:55–73 Gong LB, Xu C-Y, Chen DL, Halldin S, Chen YD (2006) Sensitivity of the Penman–Monteith reference evapotranspiration to key climatic variables in the Changjiang (Yangtze River) basin. J Hydrol 329:620–629 Groisman PY, Karl TR, Easterling DR, Knight RW, Jameson PF, Hennessy KJ, Suppiah R, Page CM, Wibig J, Fortuniak K, Razuvaev V, Douglas A, Rorland E, Zhai PM (1999) Changes in probability of heavy precipitation: important indicators of climatic change. Clim Change 42:243–283 Kendall MG (1975) Rank correlation methods. Griffin, London Liu B, Xu M, Henderson M, Gong W (2004) A spatial analysis of pan evaporation trends in China, 1955–2000. J Geophys Res 109: D15102. doi:10.1029/2004JD004511 Mann HB (1945) Nonparametric tests against trend. Econometrica 13:245–259 Menzel L, Bürger G (2002) Climate change scenarios and runoff response in the Mulde catchment (Southern Elbe, Germany). J Hydrol 267:53–64 Michael T, Hobbins, Jorge AR (2004) Trends in pan evaporation and actual evapotranspiration across the conterminous U.S.: Paradoxical or complementary? Geophys Res Lett 31. doi:10.1029/2004GL019846 Mitchell JM, Dzerdzeevskii B, Flohn H, Hofmeyr WL, Lamb HH, Rao KN, Wallén CC (1966) Climate change, WMO Technical Note No. 79, World Meteorological Organization, p 79 Ohmura A, Wild M (2002) Is the hydrological cycle accelerating? Science 298:1345–1346 Peterson TC, Golubev VS, Groisman PY (1995) Evaporation losing its strength. Nature 337:687–688 Robock A, Konstantin VY, Srinivasan G (2000) The global soil moisture data bank. Bull Am Meteorol Soc 81:281–1299 Roderick ML, Farquhar GD (2002) The cause of decreased pan evaporation over the past 50 years. Science 298(15):1410–1411 Roderick ML, Farquhar GD (2004) Changes in Australian pan evaporation from 1970 to 2002. Int J Climatol 24(9):1077–1090 Suppiah R, Hennessy K (1998) Tends in seasonal rainfall, heavy raindays, and number of dry days in Australia 1910–1990. Int J Climatol 18:1141–1155 Taikan O, Shinjiro K (2005) Global hydrological cycles and world water resource. Science 313:1068–1072 Wang W, Chen X, Shi P, van Gelder PHAJM (2008) Detecting changes in extreme precipitation and extreme streamflow in the Dongjiang River Basin in southern China. Hydrol Earth Syst Sci 12:207–221 World Meteorological Organization (WMO) (2003) Statement on the status of global climate in 2003. WMO Publ. no. 966, WMO, Geneva, Switzerland

488 Xu ZX, Li JY (2003) A distributed approach for estimating basin evapotranspiration: comparison of the combination equation and the complementary relationship approaches. Hydrol Process 17:1509–1523 Xu C-Y, Singh VP (2002) Cross-comparison of empirical equations for calculating potential evapotranspiration with data from Switzerland. Water Resour Manage 16:197–219 Xu C-Y, Singh VP (2005) Evaluation of three complementary relationship evapotranspiration models by water balance approach to estimate actual regional evapotranspiration in different climatic regions. J Hydrol 308:105–121 Xu C-Y, Widén E, Halldin S (2005) Modelling hydrological consequences of climate change—progress and challenges. Adv Atmos Sci 22(6):789–797 Xu C-Y, Gong L-B, Jiang T, Chen DL, Singh VP (2006) Analysis of spatial distribution and temporal trend of reference evapotranspiration and pan evaporation in Changjiang (Yangtze River) catchment. J Hydrol 327:81–93 Zhai PM, Ren FM, Zhang Q (1999) Detection of trends in China’s precipitation extremes. Acta Meteorol Sin 57(2):208–216

Q. Zhang et al. Zhang YL, Qin BQ, Chen WM (2004) Analysis of 40 year records of solar radiation data in Shanghai, Nanjing and Hangzhou in Eastern China. Theor Appl Climatol 78:217–227 Zhang Q, Liu C-L, Xu C-Y, Xu YP, Jiang T (2006) Observed trends of annual maximum water level and streamflow during past 130 years in the Yangtze River basin, China. J Hydrol 324:255–265 Zhang Q, Xu C-Y, Zhang Z, Chen YD, Liu C-Y, Lin H (2008) Spatial and temporal variability of precipitation maxima during 1960– 2005 in the Yangtze River basin and possible association with large-scale circulation. J Hydrol 353:215–227 Zhang Q, Xu C-Y, Gemmer M, Chen YD, Liu C-Y (2009a) Changing properties of precipitation concentration in the Pearl River Basin, China. Stoch Environ Res Risk Assess 23(3):377–385 Zhang Q, Xu C-Y, Zhang Z, Chen YD (2009b) Changes of temperature extremes for 1960–2004 in Far-West China. Stoch Environ Res Risk Assess. doi:10.1007/s00477-008-0252-4 Zhang Q, Xu C-Y, Zhang Z, Chen YD, Liu C-L (2009c) Spatial and temporal variability of precipitation over China, 1951–2005. Theor Appl Climatol 95:53–68