Stoch Environ Res Risk Assess (2011) 25:139–150 DOI 10.1007/s00477-010-0428-6 ORIGINAL PAPER Comparison of evapotranspiration variations between the Yellow River and Pearl River basin, China Qiang Zhang • Chong-Yu Xu • Yongqin David Chen • Liliang Ren Published online: 8 July 2010 Ó Springer-Verlag 2010 Abstract Based on daily meteorological data at 43 gauging stations in the Pearl River basin and 65 gauging stations in the Yellow River basin, we analyze changing properties of actual evapotranspiration (ETa), reference evapotranspiration (ETref) and precipitation in these two river basins. In our study, Pearl River basin is taken as the ‘energy-limited’ system and the Yellow River basin as the ‘water-limited’ system. The results indicate decreasing ETa in the Pearl River and Yellow River basin. However, different changing properties are detected for ETref when compared to ETa. The middle and upper Yellow River basin are characterized by increasing ETref values, whereas the Pearl River basin is dominated by decreasing ETref values. This result demonstrates enhancing drying force in the Yellow River basin. ETa depends mainly on the Q. Zhang (&) Department of Water Resources and Environment, Sun Yat-sen University, Guangzhou 510275, China e-mail: zhangqnj@gmail.com C.-Y. Xu Department of Geosciences, University of Oslo, PO Box 1047, Blindern, 0316 Oslo, Norway Y. D. Chen Department of Geography and Resource Management, The Chinese University of Hong Kong, Hong Kong, Hong Kong, China Y. D. Chen Centre of Strategic Environmental Assessment for China, The Chinese University of Hong Kong, Hong Kong, Hong Kong, China L. Ren State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing 210098, China changes of precipitation amount in the Yellow River basin. In the Pearl River basin, however, ETa changes are similar to those of ETref, i.e. both are in decreasing trend and which may imply weakening hydrological cycle in the Pearl River basin. Different influencing factors are identified behind the ETa and ETref in the Pearl River and Yellow River basin: In the Pearl River basin, intensifying urbanization and increasing aerosol may contribute much to the evapotranspiration changes. Variations of precipitation amount may largely impact the spatial and temporal patterns of ETa in the Yellow River basin. The current study is practically and scientifically significant for regional assessment of water resource in the arid and humid regions of China under the changing climate. Keywords Reference evapotranspiration Actual evapotranspiration The humid and arid regions Hydrological cycle China 1 Introduction The increasing temperature and its influence on human society and the hydrologic cycle have caused considerable concern to investigators in recent years (e.g., Xu et al. 2005; Huntington 2006). Significant evidence (including precipitation, runoff and soil moisture variations; Alan et al. 2003; WMO 2003; Zhang et al. 2009) shows that the projected global climate changes have the potential to accelerate the global hydrologic cycle. Evaporation is the only hydrological variable that connects water balance and energy balance, and it provides a sound understanding of hydrologic cycle (Xu and Singh 2005). Xu et al. (2006) summarized the key parameters describing evaporation in the literatures: (1) free water evaporation, ETo, evaluates 123 140 the amount of evaporation from open/free water surface; (2) actual evapotranspiration, ETa, describes all the processes that liquid water at or near the land surface becomes atmospheric water vapor under natural conditions; (3) potential evapotranspiration introduced in the late 1940s and 1950s by Penman (1948, 1956) and is defined as ‘‘the amount of water transpired in a given time by a short green crop, completely shading the ground, of uniform height and with adequate water status in the soil profile’’. Reference evapotranspiration, ETref (Allen et al. 1998) is defined as ‘‘the rate of evapotranspiration from a hypothetical reference crop with an assumed crop height of 0.12 m, a fixed surface resistance of 70 s m-1 and an albedo of 0.23, closely resembling the evapotranspiration from an extensive surface of green grass of uniform height, actively growing, well-watered, and completely shading the ground’’. Evapotranspiration is perhaps the most difficult of all hydrological variables of the hydrologic cycle due to complex interactions among the variables of the land-plantatmosphere system (Xu and Singh 2005). It is generally expected that, as the average global temperature increases, the air will become drier and evaporation from terrestrial water bodies will increase. However, many studies showed decreased pan evaporation and ETref (e.g., Hobbins and Ramı́rez 2004). This is usually called the pan evaporation paradox. Peterson et al. (1995) demonstrated that the evaporation of water in the United States and the former Soviet Union has decreased. They attributed the decreased evaporation of water to decreasing diurnal temperature range (DTR). Roderick and Farquhar (2002) believed that the decrease in evaporation is the result of the observed large and widespread decreases in sunlight resulting from increasing cloud coverage and aerosol concentration. However, Brutsaert and Parlange (1998) indicated that, in nonhumid environments, measured pan evaporation was not a good measure of potential evaporation; moreover, in many situations, decreasing pan evaporation actually provides a strong indication of increasing terrestrial evaporation, implying intensifying hydrological cycle in large regions. Hobbins and Ramı́rez (2004) indicated that the ‘‘Pan Evaporation Paradox’’ is no more than a manifestation of the complementarity between ETa and potential evapotranspiration. They also pointed out that although pan evaporation is a very useful concept, it can be misleading if used by itself to indicate climatic trends. The previous studies only focused on evaporation processes in specific regions or specific evaporation variables such as reference evaporation (e.g. Xu et al. 2006). It should be noted here that evaporation changes depend heavily on the availability of water and energy. In this sense, comparisons of evaporation changes in humid and arid regions can be expected to improve and enhance 123 Stoch Environ Res Risk Assess (2011) 25:139–150 human understanding of hydrological cycle variations under the influences of climate changes, particularly the human understanding of ‘‘pan evaporation paradox’’. This is the major motivation of this current study. There are some studies available concerning evapotranspiration variations in China. Xu et al. (2006) investigated the spatial patterns of ETref and pan evaporation of the Yangtze River basin. Chen et al. (2005) explored differences in estimates of ETref by using Thornthwaite method, pan data and Penman–Monteith method. Gao et al. (2007) studied ETa trends by modified water balance methodology in China. Of course, there are also other similar studies and no complete list is provided here for the sake of limited space of this paper. The current study is different from the foregoing researches in that the current study focus on the evaporation changes in the humid and arid regions characterized by different water and energy conditions. Availability of water and energy is the key factor influencing the evaporation, ETref or ETa. Besides, major factors influencing the evaporation changes in the arid and humid regions are discussed. These points clarify the novelty of this current analysis. In this case, we analyze changes of ETref and ETa in the arid and humid regions of China, i.e. the Yellow River and the Pearl River with the major objectives as: (1) analysis of spatial patterns of ETref and ETa; (2) investigation of trends of ETref and ETa in the humid and arid regions of China; and (3) explore possible causes behind the changing patterns of evapotranspiration in the humid and arid regions respectively in China. 2 Study regions 2.1 The Pearl River basin The Pearl River basin (97°390 E–117°180 E; 3°410 N–29°150 N) (Fig. 1) is the third largest river in China with drainage area of 7.96 9 105 km2. The Pearl River basin constitutes three major tributaries: West River, North River and East River. The West River is the largest tributary accounting for 77.8% of the total drainage area of the basin. The North River is the second largest tributary with a drainage area of 46710 km2. The East River accounts for 6.6% of the total area of the Pearl River. The Pearl River basin is located in the tropical and sub-tropical climate zone with the annual mean temperature ranging between 14 and 22°C and the precipitation mainly concentrating during April–September (Zhang et al. 2009), accounting for 72–88% of the annual precipitation (PRWRC 1991). The average annual humidity is 71–80%. In this study, the Pearl River basin is regarded as the humid environment. Stoch Environ Res Risk Assess (2011) 25:139–150 141 Fig. 1 Location of the study river basins. The points in the river basins show the locations of the gauging stations 2.2 The Yellow River basin The Yellow River (Huang He) (95°530 E–119°50 E; 32°100 N– 41°500 ) (Fig. 1) is the second largest river in China and the fifth largest river in the world (Zhang et al. 2008). The mainstem of the Yellow River is 5464 km in length with an area of 752440 km2, flowing mainly through the arid and semi-arid region. The altitude of the river basin ranges from above 4000 m a.s.l. in the west to below 1000– 2000 m a.s.l. in the middle and lower reaches. The average annual precipitation is about 466 mm and the annual pan evaporation ranges from 700 mm to 1800 mm. The annual mean surface air temperature is 1–8°C in the upper Yellow River basin, 8–14°C in the middle Yellow River basin and 12–14°C in the lower Yellow River basin. The evaporation exerts significant impacts on availability of water resources in the Yellow River basin (Zhang et al. 2008). 3 Data and methodology 3.1 Data Daily meteorological data (such as temperature, precipitation, relative humidity, sunshine hours, wind speed, etc.) covering 1960–2005 at 65 meteorological stations in the Yellow River basin and at 42 stations in the Pearl River basin have been analyzed in the current study. The missing data account for\0.001% of the total data and are filled by using the neighboring stations through the simple linear regressive method (the R2 is [0.8). If the relation is not good, e.g. the R2 is \0.8, the missing data will be replaced by long-term average value. The 1-day missing data are replaced by the average of the neighboring days. The series with missing data of more than 1 year will be excluded from the analysis. We believe that the process of the missing data satisfies the analysis of the current study. Furthermore, the data consistency is checked by the double-mass method and the result reveals that all the data series used in the study are consistent. 3.2 Methodology There exist some methods proposed for computation of the ETa (Xu and Singh 2005): (1) Penman–Monteith method, and (2) complementary relationship. The first method requires data on aerodynamic resistance and surface resistance with are not readily available, which limits its use in practice. The second method is usually preferred because it requires only standard meteorological variables and does not require local parameter calibration (Xu and Singh 2005). Thus, the complementary relationship is used in the current study. The computation steps of this study involve: (I) Computation of long-term annual average ETa by using traditional long-term water balance methods. Xu and Singh (2004) summarized the models for computation of ETa with the long-term water balance 123 Stoch Environ Res Risk Assess (2011) 25:139–150 where AE denotes the long-term average ETa, P denotes the long-term average precipitation, and PE denotes the long-term average ETref. After this step, one long-term average ETa value is computed for each station. (II) Xu and Singh (2005) evaluated and compared the performance of three evapotranspiration models in three study regions, i.e. Central Sweden (seasonally snow-covered boreal region), Eastern China (subtropical humid monsoon region) and Northwestern Cyprus (semiarid region) and the models are: (1) advection-aridity (AA) model (Brutsaert and Stricker 1979); (2) the complementary relationship areal evapotranspiration (CRAE) model (Morton 1978, 1983); and (3) the complementary relationship model proposed by Granger and Gray (1989) using the concept of relative evapotranspiration. Xu and Singh (2005) named this model as GG model in their study. More detailed information about these three models can be referred to Xu and Singh (2005). Xu and Singh (2005), after thorough analysis, indicated that in all the three study regions, the CRAE model produced slightly better results when the recalibrated model parameters are used. In this case, CRAE model is used in this current study. (III) The original parameters of CRAE model are calibrated based on the long-term average ETa by Eq. 1 for each station. Then the CRAE model with calibrated parameters is used to compute the daily ETa for each station. We analyze the daily ETa and the ETref data by Penman–Monteith method with aim to investigate changing properties of evapotranspiration in the arid region (the Yellow River basin) and the humid region (the Pearl River basin) of China (Fig. 2). To detect trends within the evapotranspiration series, we use non-parametric Mann–Kendall trend test since that non-parametric trend detection methods are less sensitive to outliers than are parametric statistics (Kendall 1975; Mann 1945). Moreover, the rank-based nonparametric Mann–Kendall test can test trends in a time series without requiring normality or linearity (Wang et al. 2008), and is therefore highly recommended for general use by the World Meteorological Organization (Mitchell et al. 1966). It is widely used in 123 200 100 0 30 Temperature (oC) ð1Þ 300 2 4 6 8 10 12 20 10 −10 3 2 4 6 8 10 12 2 4 6 8 10 12 3 2 1 0 Pearl River Yellow River 0 Wind speed (m/s) P AE ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi P 2ffi 1 þ PE Vapor pressure (0.1 hPa) methods. The water balance analysis results indicated that the formula by Pike (1964) can well describe the long-term annual average ETa. The formula is written as: Precipitation (mm) 142 2 4 6 8 10 12 2.5 2 1.5 1 Fig. 2 Comparison of the mean monthly climatological variables for the two study regions detection of trends in hydrological series (e.g., Gao et al. 2007; Zhang et al. 2009). The 95% confidence level is used to detect significance of trends. The observed trends are spatially interpolated by applying the Inverse Distance Weighted (IDW) interpolation method. IDW implies that each station has a local influence that decreases with distance (De By 2001). It is not meaningful to analyze the regional structure of trends of hydrological components in this study by applying and comparing different interpolation methods. Applying other interpolation methods (e.g. kriging) results in similar output maps (Gemmer et al. 2004). 4 Results 4.1 Relations between precipitation, ETa and ETref Figure 3 demonstrates spatial variations of ETa (Fig. 3a), precipitation (Fig. 3b) and ETref (Fig. 3c). In the Pearl River basin, precipitation amount is much larger than ETref and ETa (Fig. 3). Figure 2 also illustrates larger precipitation, temperature and vapor pressure in the Pearl River than in the Yellow River basin, and it is particularly the case for the vapor pressure, except the wind speed. Figure 3 also shows that spatial patterns of ETa are similar to those of ETref. Large difference can be observed between spatial distributions of precipitation and those of ETa and ETref. In the Yellow River basin, however, spatial properties of ETa are similar to those of precipitation, and comparatively, the spatial patterns of ETref are distinctly different in comparison with those of ETa and precipitation (Fig. 4). For further insight into these viewpoints, we randomly select two stations in the Pearl River basin (Anshun and Stoch Environ Res Risk Assess (2011) 25:139–150 A 143 A B B C C Fig. 3 Long-term average of annual actual evaporation (a), precipitation (b) and reference evapotranspiration (c) of the Pearl River basin Fengshan) and in the Yellow River basin (Xixian and Jingtai) with aim to show relations of precipitation (P), ETa and ETref. Figure 5 indicates that ETa is similar to ETref in terms of changing characteristics in the Pearl River basin. The precipitation amount is much higher than ETa and ETref. In the Yellow River basin, however, changes of ETa are close to precipitation, showing considerable influences of precipitation on ETa. These results are in good agreement with ‘water limited’ and ‘energy limited’ system (Xu and Singh 2004). Besides, Chi-square analysis of above hydro-meteorological variables also shows significant difference of humidity conditions of the Yellow River and the Pearl River basin respectively. In this sense, it is reasonable to accept that the Pearl River basin is an ‘energy limited’ system and the Yellow River basin a ‘water limited’ system. 4.2 Trends of precipitation, ETa and ETref Annual trends of precipitation, ETa and ETref for the Pearl River basin are demonstrated in Fig. 6. The Pearl River basin is dominated by decreasing ETa, ETref and Fig. 4 Long-term average of annual actual evaporation (a), precipitation (b) and reference evapotranspiration (c) of the Yellow River basin precipitation. The difference is that the decrease of precipitation in the Pearl River basin is not significant. Most stations show significant decreasing ETa. However, stations characterized by significant decreasing ETref are located mainly in the Pearl River basin between 106E and 110E. In terms of seasonal precipitation trends (Fig. 7), generally speaking, seasonal precipitation changes of the Pearl River basin are characterized mainly by decreasing trends, particularly in spring, summer and autumn. Precipitation in winter is increasing but is not significant. Decreasing ETref can be observed in most parts of the Pearl River basin (Fig. 8) in spring (Fig. 8a), summer (Fig. 8b) and autumn (Fig. 8c). Difference of ETref changes in winter when compared to spring, summer and autumn is that some stations in the upper and lower Pearl River basin show increasing ETref (Fig. 8d). ETa changes are in similar patterns when compared to ETref. Results by Gao et al. (2007) 123 144 Stoch Environ Res Risk Assess (2011) 25:139–150 300 250 250 Anshun station (26.25 N, 105.96E) Fengshan station (24.55 N, 107.03E) P AE ET ET, AE and P (mm) ET, AE and P (mm) Fig. 5 Long-term monthly mean reference evapotranspiration (ET), actual evaporation (AE) and precipitation (P) of two gauging stations randomly selected from the Pearl River basin (upper two panels) and the Yellow River basin (lower two panels) 200 150 100 150 100 50 50 0 200 2 4 6 8 10 0 12 Xixian station (36.7N, 110.95E) 100 50 0 2 4 6 8 Time (months) A B C Fig. 6 Trends of actual evaporation (a), precipitation (b) and reference evapotranspiration (c) over the Pearl River basin. Solid lines show positive trends and dashed lines denote negative trends 123 10 12 4 6 8 10 12 8 10 12 Jingtai station (37.18N, 104.05E) ET, AE and P (mm) ET, AE and P (mm) 150 2 150 100 P AE ET 50 0 2 4 6 Time (months) also indicated decreasing ETa in the Pearl River basin in terms of annual changes. Our results further show that seasonal ETa is also decreasing (Fig. 9). Annual trends of precipitation, ETa and ETref are illustrated in Fig. 10. ETa and precipitation are in decreasing trends and significant decrease can be identified in the middle and upper Yellow River basin. These observations are in good agreement with the results by Gao et al. (2007). ETref shows different changing properties when compared to those of precipitation and ETa (Fig. 10c). A majority of stations in the upper Yellow River basin are characterized by increasing ETref and significant increasing ETref is observe in upper Yellow River basin. These results are also in good line with those by Gao et al. (2007). Changes of seasonal precipitation, ETa and ETref are demonstrated in Fig. 11. Figure 11 shows different properties of precipitation in diverse seasons. The Yellow River basin is dominated by decreasing precipitation in spring, summer and autumn (Fig. 11a–c). Precipitation in winter is in different changing properties in comparison with those in other seasons, i.e. the west and north parts of the Yellow River basin are characterized by increasing trends and the increasing trends at some stations are significant statistically (Fig. 11d). ETa are in similar spatial patterns with those of precipitation (Fig. 12a), showing good relations between precipitation and ETa in the Yellow River basin. Comparison between Figs. 11c and 12c indicates that regions with stations characterized by significant decreasing precipitation match well the places with stations featured by significant decreasing ETa. As for precipitation Stoch Environ Res Risk Assess (2011) 25:139–150 145 A B C D Fig. 7 Trends of seasonal precipitation in spring (a), summer (b), autumn (c) and winter (d) in the Pearl River basin. Solid lines show positive trends and dashed lines denote negative trends A C B D Fig. 8 Trends of reference evapotranspiration in the Pearl River basin. Solid lines show positive trends and dashed lines denote negative trends changes in winter (Fig. 11d), majority of stations are dominated by increasing precipitation. Precipitation at some stations in west parts of the Yellow River basin is in significant increasing trends. Stations with increasing ETa distribute sporadically in the Yellow River basin (Fig. 12d). Even so, stations having increasing ETa are also located within the area dominated by increasing precipitation. The above-mentioned analysis told such a story that ETa in the Yellow River basin largely but not exactly hinges on spatial and temporal changes of precipitation. It is reasonable since that ETa variations are the results of more than meteorological factors, such as wind speed, temperature, soil moisture conditions, vapor pressure and so forth. However, the results mentioned above still well support the viewpoint that the Yellow River basin is a ‘water limited system’. ETref variations present distinctly different spatial patterns when compared to precipitation and ETa changes (Fig. 13). ETref in spring, autumn and winter is increasing in the regions between 102E and 110E of the Yellow River basin. Increasing ETref in summer is observed mainly in the west corner of the Yellow River basin. 5 Discussions Xu et al. (2006) and Gong et al. (2006) analyzed influencing factors of ETref, indicating that the most sensitive 123 146 Stoch Environ Res Risk Assess (2011) 25:139–150 A B C D Fig. 9 Trends of actual evaporation in the Pearl River basin. Solid lines show positive trends and dashed lines denote negative trends A B C Fig. 10 Trends of actual evaporation (a), precipitation (b) and reference evapotranspiration (c) over the Yellow River basin. Solid lines show positive trends and dashed lines denote negative trends 123 variable for the ETref is the relative humidity followed by the net total radiation, air temperature and wind speed. Due to considerable dependence of ETa on precipitation, we also analyze precipitation changes. In this case, we analyzed changes of precipitation, air temperature, wind speed and sunshine hours for the Pearl River (Fig. 14a–d) and the Yellow River basin (Fig. 14e–h). Seasonal variations of these meteorological factors show similar characteristics when compared to annual changes, therefore, we only analyzed the annual variations of these meteorological factors in this study. Figure 14a–d show that, in the Pearl River basin, precipitation and air temperature are increasing statistically. However, wind speed and sunshine hours are decreasing statistically. Increasing temperature may have the potential to enhance evaporation. However, some scholars indicated that a warmer atmosphere does not necessarily increase evaporation due to the fact that hemispheric evaporation is much more substantial in winter than in summer under the present climate (Ohmura and Wild 2002). Furthermore, analysis results of temperature extremes indicated significant increasing extreme low temperature but no significant increasing extreme high temperature (Zhang et al. 2008). Decreasing wind speed and sunshine hours may be mainly responsible for decreasing ETa and ETref. As for wind speed, we analyzed areal average wind speed at 200, 500, 850 and 1000 hPa (figures not shown here). The results indicated increasing wind speed at 200 hPa, and downward tendency from 200 to 1000 hPa. Therefore, we speculated that urbanization may be one of the causes behind decreasing wind speed near the ground surface. Decreasing sunshine hours may be due to increasing cloud cover. However, our analysis (figures not shown here) indicated that the areal average cloud cover is decreasing, Stoch Environ Res Risk Assess (2011) 25:139–150 147 A B C D Fig. 11 Trends of precipitation in spring (a), summer (b), autumn (c) and winter (d) in the Yellow River basin. Solid lines show positive trends and dashed lines denote negative trends A B C D Fig. 12 Trends of actual evaporation in spring (a), summer (b), autumn (c) and winter (d) in the Yellow River basin. Solid lines show positive trends and dashed lines denote negative trends thus the decreasing sunshine hours could be due to increasing aerosol across the Pearl River basin (Xu et al. 2006). All these factors combined to cause decreasing ETa and ETref over the Pearl River basin. Similarly, decreasing wind speed and sunshine hours are also observed in the Yellow River basin. Except these factors, evapotranspiration changes are also heavily influenced by precipitation amount in the Yellow River basin, a ‘water-limited’ system. Decreasing ETa across the Yellow River basin should be attributed to areal average precipitation of the Yellow River basin. Increasing ETref in the Yellow River basin implies enhancing hydrological cycle in the Yellow River basin (Xu et al. 2006; Gao et al. 2007). In this study, the Pearl River basin is taken as the wet environment and the Yellow River basin the dry environment. Figure 2 well corroborates these assumptions: the 123 148 Stoch Environ Res Risk Assess (2011) 25:139–150 A B C D Fig. 13 Trends of reference evapotranspiration in spring (a), summer (b), autumn (c) and winter (d) in the Yellow River basin. Solid lines show positive trends and dashed lines denote negative trends Fig. 14 Changes of standardized a temperature, b precipitation, c wind speed, and d sunshine hours for the Pearl River basin; and e temperature, f precipitation, g wind speed, and h sunshine hours for the Yellow River basin ETa and ETref approach each other in the Pearl River basin; precipitation and ETa match well in the Yellow River basin. Therefore, we can say that the ETa reaches the evaporation demand (or energy available for evaporation) in the Pearl River basin, i.e. an ‘‘energy limited’’ system (Xu and Singh 2004). In this Yellow River basin however, ETa are controlled mainly by the precipitation amount even the ETref is much higher than precipitation in the Yellow River basin. This is called as a ‘‘water limited’’ system (Xu and Singh 2004). As for the evapotranspiration models we used in this study, we accepted the CRAE model based on the previous studies (e.g. Xu et al. 2005). Hobbins et al. (2001) evaluated the performance of the Complementary 123 Relationship Areal Evapotranspiration (CRAE) model and the Advection-Aridity (AA) model against independent estimates of regional evapotranspiration derived from longterm, large-scale water balances (1962–1988) for 120 minimally impacted basins in the conterminous United States, demonstrating that the CRAE model accurately predicts monthly regional evapotranspiration. All these previous analysis further confirms the rationality of the application of CRAE model in the current study. It should be noted here uncertainty of the models are common in the hydrological study and the discussion of modeling uncertainty is out of the scope of this study. Even so, the results of this study still provide scientifically meaningful Stoch Environ Res Risk Assess (2011) 25:139–150 conclusions for the further development of human understanding of hydrological cycle in arid and humid regions in China, and also contribute much to the similar studies of the world. 6 Conclusions The evapotranspiration is one of the key components of hydrological cycle. ETa is attached considerable importance in sound understanding of the way the changing climate influence the hydrological cycle on a regional and global scale (e.g. Ohmura and Wild 2002). In the current study, we analyzed changing properties of ETa, ETref and precipitation in the arid region (the Yellow River basin) and humid region (the Pearl River basin). We also investigate possible causes for these changing properties by investigation changing features of wind speed, sunshine hours, precipitation and air temperature. Some interesting conclusions are achieved as the follows: (1) (2) (3) In the arid and humid regions, decreasing ETa is observed. However, ETref presents different changing properties in the arid and humid regions of China. Majority parts of the Pearl River basin are dominated by decreasing ETref and some parts of the Yellow River basin are characterized by increasing ETref, which may imply enhancing hydrological cycle in the Yellow River basin when compared to that of the Pearl River basin. In the Yellow River basin, spatial patterns of ETa match well those of precipitation changes, implying that the changes of ETa in the arid regions largely hinges on precipitation amount. In the humid regions, changes of the ETa follow the similar patterns when compared to those of ETref. Decreasing actual and ETref may imply weakening hydrological cycle in the Pearl River basin when compared to the Yellow River basin. This conclusion has the potential to contribute to better understanding of influences of climate changes on the hydrological cycle in the arid and humid regions respectively in China. Decreasing wind speed in the Pearl River and Yellow River basin may be partly due to intensifying urbanization in that wind speed tends to be increasing from 1000 to 200 hPa. Decreasing sunshine hours should be attributed to increasing aerosol because the cloud cover is increasing over the Pearl River basin. Therefore, human activities contributed much to the decreasing ETa and ETref in the Pearl River basin. In the Yellow River basin, evapotranspiration largely depends on precipitation amount. Decreasing precipitation should be responsible for decreasing ETa in the 149 Yellow River basin. The current study is practically and scientifically significant for regional assessment of water resource in the arid and humid regions of China under the changing climate. It should be noted here that in no way does it mean that the current report perfectly settled down the subject of evapotranspiration. However the current study provides an important clue for further study of hydrological cycle in the arid and humid regions of China under the changing climate, and which is the major contribution of this study to the hydrological science. Acknowledgements This work is financially supported by the Program for Outstanding Young Teachers of the Sun Yat-sen University (Grant No.: 2009-37000-1132381), the ‘985 Project’ (Grant No.: 37000-3171315), the Key National Natural Science Foundation of China (Grant No.: 50839005), a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (Project No. CUHK405308) and by Program of Introducing Talents of Discipline to Universities—the 111 Project of Hohai University. Cordial thanks should be extended to two anonymous reviewers and the editor, Prof. Dr. Jiaping Wu, for their constructive and pertinent comments which greatly helped to improve the quality of this manuscript. References Alan DZ, Justin S, Edwin PM, Bart N, Eric FW, Dennis PL (2003) Detection of intensification in global- and continental-scale hydrological cycles: temporal scale of evaluation. J Clim 16: 535–547 Allen RG, Pereira LS, Raes D, Smith M (1998) Crop evapotranspiration—guidelines for computing crop water requirements— FAO Irrigation and Drainge Paper 56. FAO. ISBN 92-5-104219-5 Brutsaert W, Parlange MB (1998) Hydrologic cycle explains the evaporation paradox. Nature 396:30 Brutsaert W, Stricker H (1979) An advection-aridity approach to estimate actual regional evapotranspiration. Water Resour Res 15(2):443–450 Chen DL, Gao G, Xu C-Y, Guo J, Ren GY (2005) Comparison of the Thornthwaite method and pan data with the standard Penman– Monteith estimates of reference evapotranspiration in China. Clim Res 28:123–132 De By RA (2001) Principles of geographic information systems. ITC Educational Textbook Series 1. ITC, Enschede, 234 pp Gao G, Chen DL, Xu C-Y, Simelton E (2007) Trend of estimated actual evapotranspiration over China during 1960–2002. J Geophys Res 112:D11120. doi:10.1029/2006JD008010 Gemmer M, Becker S, Jiang T (2004) Observed monthly precipitation trends in China 1951–2002. Theor Appl Climatol 77:39–45 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 Granger RJ, Gray DM (1989) Evaporation from natural nonsaturated surface. J Hydrol 111:21–29 Hobbins MT, Ramı́rez JA (2004) Trends in pan evaporation and actual evapotranspiration across the conterminous U.S.: Paradoxical or complementary? Geophys Res Lett 31. doi: 10.1029/2004GL019846 123 150 Hobbins MT, Ramı́rez JA, Brown TC, Claessens LHJM (2001) The complementary relationship in estimation of regional evapotranspiration: the Complementary Relationship Areal Evapotranspiration and Advection-Aridity models. Water Resour Res 37(5):1367–1387 Huntington TG (2006) Evidence for intensification of the global water cycle: review and synthesis. J Hydrol 319:83–95 Kendall MG (1975) Rank correlation methods. Griffin, London, UK Mann HB (1945) Nonparametric tests against trend. Econometrica 13:245–259 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, 79 pp Morton FI (1978) Estimating evapotranspiration from potential evaporation: practicality of an iconoclastic approach. J Hydrol 38:1–32 Morton FI (1983) Operational estimates of areal evapotranspiration and their significance to the science and practice of hydrology. J Hydrol 66:1–76 Ohmura A, Wild M (2002) Is the hydrological cycle accelerating? Science 298:1345–1346 Pearl River Water Resources Committee (PRWRC) (1991) Pearl River Water Resources Committee (PRWRC), The Zhujiang Archive, vol 1. Guangdong Science and Technology Press, Guangzhou (in Chinese) Penman HL (1948) Natural evaporation from open water, bare soil and grass. Proc R Soc Lond 193:120–145 Penman HL (1956) Evaporation: an introductory survey. Neth J Agric Sci 4:9–29 Peterson TC, Golubev VS, Groisman PY (1995) Evaporation losing its strength. Nature 337:687–688 Pike JG (1964) The estimation of annual runoff from meteorological data in a tropical climate. J Hydrol 2:116–123 123 Stoch Environ Res Risk Assess (2011) 25:139–150 Roderick ML, Farquhar GD (2002) The cause of decreased pan evaporation over the past 50 years. Science 298(15):1410–1411 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 Xu C-Y, Singh VP (2004) Review on regional water resources assessment models under stationary and changing climate. Water Resour Manage 18:591–612 Xu C-Y, Singh VP (2005) Evaluation of three complementary relationship evapotranspiration models by water balance approach to estimate actual evapotranspiration in different climatic regions. J Hydrol 308:105–121 Xu C-Y, Widén E, Halldin S (2005) Modeling hydrological consequences of climate change-progress and challenge. Advances in Atmospheric Sciences 22(6):789–797 Xu C-Y, Gong L, Jiang T, Chen D, 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 Zhang Q, Xu C, Zhang Z, Ren G (2008) Climate change or variability? The case of Yellow river as indicated by extreme maximum and minimum air temperature during 1960–2004. Theor Appl Climatol 93:35–43 Zhang Q, Xu C-Y, Gemmer M, Chen YD, Liu C-L (2009) Changing properties of precipitation concentration in the Pearl River basin, China. Stoch Environ Res Risk Assess 23:377–385