On the attribution of changing pan evaporation in a nature reserve in SW China

HYDROLOGICAL PROCESSES Hydrol. Process. (2012) Published online in Wiley Online Library (wileyonlinelibrary.com).9394 On the attribution of changing pan evaporation in a nature reserve in SW China Guangyong You, 1,4 Yiping Zhang, 1,2,3 * Yuhong Liu, 1,2,3 Qinghai Song, 1,4 Zhiyun Lu, 1,2,3 Zhenghong Tan, 1 Chuansheng Wu 1,2,3 and Youneng Xie 5 1 Key Laboratory of Tropical Forest Ecology, Xishuangbanna Tropical Botanical Garden, Chinese Academy of Sciences, Menglun, Mengla, Yunnan, 666303, China 2 Ailaoshan Station for Subtropical Forest Ecosystem Studies (ASSFE), Chinese Ecosystem Research Networks, Jingdong, Yunnan, 676209, China 3 National Forest Ecosystem Research Station at Ailaoshan, Jingdong, Yunnan, 676209, China 4 Graduate School of the Chinese Academy of Sciences, Beijing 100049, China 5 Jingdong Bureau of National Nature Reserve, Jingdong, Yunnan, 676209, China Abstract: Negative trends of measured pan evaporation are widely reported. Studies of the factors that underlie this reduction in pan evaporation have not reached a consensus about the controlling factors. Most studies employ statistical analysis (correlation analysis or stepwise regression) to identify the controlling climatic variables; in contrast, few studies have employed physicalbased theories. In addition, observations of pan evaporation and related climatic variables are reported to be influenced by anthropogenic activities. Consequently, the observed trends of climatic variables in a nature reserve would be useful for understanding regional climate change. The present study site is located in Ailaoshan National Nature Reserve, SW China, which is free of anthropogenic activity. In this study, we firstly applied the adjusted PenPan model to estimate the pan evaporation. Then, using this physical-based model, we identified a positive trend in pan evaporation, with a much larger increase in the dry season than in the wet season. The model results indicate that the change in the aerodynamic component is larger than that in the radiative component. In contrast to the reduction in wind speed and sunshine hours that has been reported in previous studies at various sites, we found that wind speed and sunshine hours have increased in recent decades, thereby explaining the increase of the pan evaporation rate. Wind speed made the greatest contribution to the change in pan evaporation, followed by sunshine duration. This study indicates that the potential evaporation has increased at this site despite the widely reported reduction in measured pan evaporation. During the dry season, the availability of water for agriculture and agroforestry could be threatened. Copyright 2012 John Wiley & Sons, Ltd. KEY WORDS climate change; pan evaporation; adjusted PenPan model; Ailaoshan National Nature Reserve Received 17 November 2011; Accepted 20 April 2012 INTRODUCTION One of the expected consequences of global warming is an increase in the rate of evaporation as a consequence of increased air temperature (Stocker et al., 2001). However, negative trends in measured pan evaporation have been reported in the United States (Hobbins et al., 2004), the former Soviet Union (Peterson et al., 1995), the Canadian prairies (Burn and Hesch, 2007), Australia (Roderick et al., 2007), New Zealand (Roderick and Farquhar, 2005), northeast India (Jhajharia et al., 2009), Thailand (Tebakari et al., 2005), and China (Xu et al., 2006; Wang et al., 2007; Zhang et al., 2007). However, positive trends have been reported in west Iran (Tabari and Marofi, 2011), north-eastern Brazil (da Silva, 2004), Israel (Cohen and Stanhill, 2002), and West Africa (Oguntunde et al., 2006). Consequently, there is no consistency in the reported trends in pan evaporation measurements (Fu, 2009). Such contradictory trends in pan evaporation *Correspondence to: Yiping Zhang, Key Laboratory of Tropical Forest Ecology, Xishuangbanna Tropical Botanical Garden, Chinese Academy of Sciences, Menglun, Mengla, Yunnan 666303, China E-mail: yipingzh@xtbg.ac.cn trends can even coexist in the same region (Wang et al., 2007; Ji and Zhou, 2011). No consensus has been reached about the underlying causes of these changes in pan evaporation. The reduction in pan evaporation has predominantly been attributed to reduced solar irradiance in Russia (Peterson et al., 1995), the United States (Hobbins et al., 2004), China (Thomas, 2000; Liu et al., 2004; Gao et al., 2006; Wang et al., 2007), and Israel (Cohen and Stanhill, 2002), which is related to increased cloud coverage and enhanced aerosol concentrations (Roderick and Farquhar, 2002). Others have mainly attributed reductions in pan evaporation to changes in wind speed in Australia (Rayner, 2007; Roderick et al., 2007), the Tibetan Plateau (Chen et al., 2006; Zhang et al., 2007), and the Canadian prairies (Burn and Hesch, 2007); to changes in relative humidity in India (Chattopadhyay and Hulme, 1997) and the Chang Jiang Basin (Gong et al., 2006); and to changes in the maximum temperature in China (Cong and Yang, 2009). One key concern with current climate change is the impact of these changes on the hydro balance, which is largely determined by the potential evaporation (Cho et al., 2011; Kirono and Ken, 2011). Furthermore, changes in pan evaporation and the related climatic Copyright 2012 John Wiley & Sons, Ltd.

G. YOU ET AL. variables are considered to be influenced by anthropogenic activities, such as urbanization and irrigation (Alpert et al., 2005; Alpert and Kishcha, 2008; Ji and Zhou, 2011). Moreover, previous understandings of factors that influence pan evaporation change were based on statistical analyses, such as correlation and stepwise regression (Chattopadhyay and Hulme, 1997; Chen et al., 2006; Cong and Yang, 2009; Jhajharia et al., 2009; Ji and Zhou, 2011). Consequently, a physically based estimate of pan evaporation in a nature reserve should be useful in understanding the natural change in pan evaporation. This study was undertaken at a site in a nature reserve in south-west China, and the objective of the study was to answer the following questions: (1) What changes in pan evaporation are occurring in this nature reserve? and (2) What are the main factors contributing to these changes in pan evaporation? Study site MATERIALS AND METHODS The Ailaoshan Station for Subtropical Forest Ecosystem Studies (ASSFE; 24 32 0 N, 101 01 0 E; 2480 m a.s.l.) of the Chinese Academy of Sciences is located in Jingdong County, Yunnan Province (Figure 1). A standard meteorological station operated by ASSFE has been in operation at the site since 1981. Based on longterm meteorological observations from that station, the annual mean temperature is 11.0 C, with a monthly mean temperature of 5.2 C in the coldest month (January) and 15.2 C in the warmest month (July). The average annual rainfall is 1902 mm, with 1630 mm falling during the wet season (May October) and 272 mm during the dry season (November April). The average annual pan evaporation (Φ20 cm) is 1297 mm: 583 mm in the wet season and 714 mm in the dry season. Estimation of pan evaporation The PenPan model was derived from Penman s combination equation (Penman, 1948) and then developed by Roderick et al. (2007). The pan evaporation (E p )isseparated into radiative (E p,r ) and aerodynamic (E p,a ) components (Equation (1)). E p ¼ E p;r þ E p;a Δ R n ag ¼ þ Δ þ ag l Δ þ ag f qðu 2 ÞVPD Δ (k Pa C 1 ) refers to the slope of the saturation vapour pressure (e s, k Pa) with temperature (T, C) (Equation (2)), R n (J m 2 ) refers to the net radiation (Equation (7)), a (= 2.4 here) is the ratio of the effective surface area for heat and vapour transfer, l (= 2.45 MJ kg 1 ) is the latent heat of vaporization, g (k Pa C 1 ) is the psychrometric constant adjusted for elevation (Equation (3)), f q (U 2 )isa function based on the wind speed at a height of 2 m (Equation (4)), and U 2 (m s 1 ) is calculated from the wind speed at a height of 10 m and an empirical coefficient from the wind profile function (Equation (5)). VPD (k Pa) is the vapour saturation deficit (Equation (6)) between the actual vapour pressure (e a, k Pa) and the saturation vapour pressure (e s, k Pa). T a and T w in Equation (6) are the temperatures of the dry bulb and wet bulb, respectively. Δ ¼ h 4098 0:6108 exp 17:27T Tþ237:3 i (1) ðt þ 237:3Þ 2 (2) g ¼ C pp el ¼ 0:665 10 3 P (3) f q ðu 2 Þ ¼ 1:39 10 8 ð1 þ 1:35U 2 Þ (4) 4:87 U 2 ¼ U z lnð67:8z 5:42Þ 10 (5) VPD ¼ e s e a ¼ 0:611 exp 17:27T T þ 237:3 gðt a T w Þ (6) The net radiation (R n ) in Equation (1) was estimated with Equation (7), which differs from the process used by Roderick et al. (2007). S t (hours of sunshine) is the actual duration of sunshine, recorded with a dark-tube recorder. N (h) is the maximum possible duration of sunshine, and R a is the extra-terrestrial radiation (MJ m 2 day 1 ). a (= 0.25 here) and b (= 0.5 here) are empirical constants (Allen et al., 1998). s is the Stefan Boltzmann constant (4.903 10 9 MJ K 4 ). T max (K) and T min (K) are the maximum and minimum temperatures, respectively. R n ¼ 0:77 a þ b St R a s T max 4 4 þ T min N 2 0:34 0:14 ffiffiffiffi p ð e a Þ 1:35 a þ b St (7) N 0:35 a þ b Figure 1. The location of the study site To evaluate the estimated R n, the observed water surface R n in the evaporator was calculated with Equation (8). Global radiation (Q) was observed by a pyranometer

POSITIVE TREND OF PAN EVAPORATION (CM11, Kipp & Zonen, Delft, Holland). We used A sp = 0.14 as the albedo of the water surface in the evaporator (Linacre, 1992; Rotstayn et al., 2006). R l is the net long-wave irradiance of the pan. In this study, R l was simply considered a fixed value of 40 W m 2 during cloudless periods (Linacre, 1994). R n ¼ 1 A sp Q þ Rl (8) The performances of the estimated R n and E p were evaluated with Equation (9). x i refers to the observed records; ^x i refers to the modelled values. n is the sample months of the observations. In this study, the PenPan model was evaluated for the 36-month observation of E 601 (2008 2010) and the 23-year observation of the Φ20-cm pan (1983 1990, 1996 2010). sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi P n i¼1ð RSME ¼ x i ^x i Þ 2 (9) n Attribution of the change in pan evaporation The change in pan evaporation (de p /) was separated into the change in the radiation component (de p,r /) and the change in the aerodynamic component (de p,a /) (Equation (10)). The change in the radiation component (Equation (11)) could be attributed to the changes in the temperature (T # ), the minimum temperature (T # min ), the maximum temperature (T # max ), the sunshine hours (S # t ), and the actual vapour pressure (e # a ). The change in the aerodynamic component (Equation (12)) could be attributed to the changes in the wind speed (U # ), the saturation vapour pressure deficit (VPD # ), and the temperature (T # ). When calculating the contribution of each variable, the other variables were assumed to remain constant at their long-term means. de p;r @E p;r @Δ þ @E p;r @Rn de p ¼ de p;r þ de p;a @Δ dt @T þ @E p;r @R n @R n @St @Rn dt min @T min dst þ @E p;r @R n @R n dt max @T max (10) (11) slopes were calculated with the Kendall package (McLeod, 2009) and zyp package (Bronaugh, 2009), respectively, in the R environment (R Development Core Team, 2010). A positive tau value indicated an increasing trend, and a negative tau indicated a decreasing trend. The slope of each variable was computed by simple linear regression against the year. Model performance RESULTS We first estimated R n over 36 months (2008 2010) with Equation (7). The estimated R n was then compared with the observed value calculated with Equation (8). The agreement between the estimated R n and observed R n was excellent (Figure 2). We then extended the estimation of R n to the rest of the years (before 2008), and the modelled E p was calculated with the estimated R n. Figure 3 shows that the modelled E p had conversion coefficients of 0.691 and 1.046 for the simultaneous observations of the Φ20-cm pan and E 601,respectively. Attribution of changing pan evaporation The analysis of the trends in the climatic variables is shown in Table I. Positive trends were observed in temperature, sunshine hours, and wind speed, and their increases in the dry season were greater than those in the wet season. VPD did not show an obvious trend but increased slightly in the dry season and decreased slightly in the wet season. Table II shows the partitioning of E p in the dry season, the wet season, and the whole year. The modelled E p increased with a trend of 4.54 mm a 2 in the last 30 years, and the increase in E p was predominantly attributable to its increase in the dry season. Changes in the aerodynamic component accounted for 56.68% of the E p change in the dry season. However, in the wet season, the radiative component accounted for 89.02% of the change in E p. þ @E p;r @R n de a @R n @e a ¼ T # þ T # min þ T # max þ St # # þ e a (11) de p;a @E p;a du @u þ @E p;a dvpd þ @E p;a dδ dt a @VPD @Δ dt a ¼ U # þ VPD # þ T # (12) Trend analysis The trend for each meteorological variable was analysed with the Mann Kendall test (Mann, 1945). Mann Kendall s tau values and the non-parametric Sen s Figure 2. The linear relationship between the estimated and observed net radiation (R n )

G. YOU ET AL. Figure 3. Comparison of the modelled pan evaporation and the pan evaporation of (a) Φ20-cm and (b) E 601 Table I. Trend analysis of meteorological variables (S t, Φ20-cm E p and U had 5 years data gap (1991 1995)) Meteorological variables Trend Dry Wet Average MK tau 0.448*** 0.256* 0.523*** T ( Ca 1 ) Sen s slope 0.058*** 0.011* 0.041*** Linear trend 0.050*** 0.014* 0.032*** MK tau 0.417** 0.466** 0.450** T max ( Ca 1 ) Sen s slope 0.074** 0.054** 0.065** Linear trend 0.069** 0.043** 0.057*** MK tau 0.328* 0.147 0.393** T min ( Ca 1 ) Sen s slope 0.037* 0.011 0.026** Linear trend 0.039** 0.009 0.024** MK tau 0.148 0.099 0.113 Rainfall (mm a 1 ) Sen s slope 1.718 6.205 6.397 Linear trend 1.532 5.046 3.514 MK tau 0.273 0.020 0.217 St (h a 1 ) Sen s slope 7.562 1.409 7.652 Linear trend 7.542* 1.084 8.627 MK tau 0.115 0.044 0.130 Φ20-cm E p (mm a 1 ) Sen s slope 2.387 1.433 4.989 Linear trend 1.196 0.827 0.369 MK tau 0.559** 0.496** 0.513** U (m s 1 a 1 ) Sen s slope 0.061** 0.045** 0.052** Linear trend 0.062*** 0.041** 0.052*** MK tau 0.150 0.230* 0.057 VPD (kpa a 1 ) Sen s slope 0.001 0.001* 0.000 Linear trend 0.001 0.001 0.000 *P < 0.1; **P < 0.01; ***P < 0.001 The increase in E p,a was mainly contributed by an increase in the wind speed, and the increase in E p,r was mainly contributed by an increase in the number of sunshine hours. In summary, the change in E p was predominantly attributed to the change in wind speed, followed by the change in sunshine hours (Table II). DISCUSSION PenPan model performance and uncertainties Fu et al. (2004) calculated the conversion coefficients between various evaporation pans and a 20-m 2 evaporation tank, yielding coefficients of about 0.6 for the Φ20-cm pan and 1.07 for E 601. In addition, Xu et al. (2006)reporteda conversion coefficient of 0.7 for Φ20-cm pan evaporation compared with reference evapotranspiration in the Yangtze River catchment. Consequently, the PenPan model showed an excellent performance in this study because the conversion coefficients for the Φ20-cm pan and E 601 were similar to those in the studies listed above (Figure 3). However, several uncertainties in the PenPan model must be improved. For example, the albedo cannot be constant under field conditions (Roderick et al., 2007), and the current variables inadequately represent the actual variation in solar irradiance. Changes in atmospheric aerosols could not be detected by the current dark-tube recorder; however, such changes could have influenced the pan evaporation trend.

POSITIVE TREND OF PAN EVAPORATION Table II. The modelled pan evaporation rate (de p /, inmma 2 ) and its components Dry Wet Average Estimated E p de p / 4.12 0.42 4.54 Radiation de p,r / 1.78 0.38 2.16 Radiation partition T # 0.55 0.11 0.66 # T min 0.00 0.00 0.00 # T max 0.00 0.00 0.00 St # 1.23 0.26 1.50 ea # 0.00 0.00 0.00 Aerodynamic de p,a / 2.33 0.05 2.38 Aerodynamic partition T # 0.17 0.01 0.18 U # 1.87 0.68 2.55 VPD # 0.64 0.63 0.01 Moreover, the PenPan model does not consider the heat storage component; consequently, it is only applicable to weekly or monthly input data (Thom et al., 1981). The results of measured pan evaporation studies are inconsistent, partly because of differences in the measuring instruments used. For example, the Φ20-cm pan can harvest energy more efficiently than other pan evaporators (McVicar et al., 2007), and the fraction of vapour condensing from the air to the water surface in the evaporators, which occurs when the water temperature is below the dew point (mainly during the rainy season), cannot be detected by the current method. Therefore, the change in pan evaporation could be influenced by many factors, meaning that it may not accurately represent the changes in potential evaporation. The positive trend in pan evaporation estimated with the PenPan model in this study is consistent with the increasing pan evaporation and potential evaporation on the Tibetan Plateau and in south-west China (Thomas, 2000; Xu et al., 2005; Chen et al., 2006; Gao et al., 2006; Cong and Yang, 2009). In addition, Cong and Yang (2009) reported that most stations in China have shown positive trends in pan evaporation since 1986. Liu et al. (2011) reported an increase of pan evaporation in China from 1992 to 2007. Consequently, this study concludes that potential evaporation and water evaporation demand could have increased in the post-1981 period. Further studies at the regional level are needed to confirm this finding. According to the physical theory of water vaporization, potential evaporation is more sensitive to relative humidity than to other climatic variables, which means that a small change in relative humidity could result in a large change in potential evaporation (Gong et al., 2006; Yin et al., 2010). This could be the reason that the correlation coefficient for relative humidity is higher than that for any other climatic variable (Ji and Zhou, 2011). However, when the slope of relative humidity is considered, its contribution to pan evaporation may differ (because the contribution is obtained by multiplying the coefficient by the slope). In this study, the slope of VPD change was low (Table I), so the change in VPD made a limited contribution to the change of pan evaporation. Physically, T max, T min, and e a may have a strong influence on the water evaporation rate. These factors were used to estimate the net radiation in this study. Based on the coefficients for these factors in Equation (7), their contributions to the change in net radiation was very small. Consequently, the model results indicate that these factors made a minor contribution to the change in pan evaporation. Climate change at the nature reserve site Observed changes in climatic variables can be influenced by anthropogenic activities. For example, urbanization can change the surface roughness and consequently reduce the measured wind speed around the meteorological station (Ren et al., 2008; Jiang et al., 2010). The increase in aerosols in the atmosphere, deriving from the industrialization of recent decades, can reduce the availability of sunlight (Alpert et al., 2005). Irrigation around the meteorological station could have changed the vapour pressure deficit in rural area. As a result, observations in a nature reserve are more precise in representing the natural trends of climatic variables. Contrary to the widely reported negative trends in sunshine hours and wind speed, this nature reserve site showed positive trends in both sunshine hours and wind speed (Table I). This is possibly attributable to the absence of anthropogenic influences in this area. A slight increase in wind speed has recently been reported in south-west China (Jiang et al., 2010). Guo et al. (2010) also reported that the wind speed in Yunnan Province increased in spring and summer during the period 1991 2005. Given that previous studies have reported reduced precipitation (Wang and Ding, 2006) and vapour supply (Yao et al., 2008), it is postulated that more energy is partitioned to sensible heat, thereby increasing the air current momentum (McVicar et al., 2012). In addition, the positive trend in sunshine hours observed in this study is consistent with the study of Zheng et al. (2010), who reported a positive trend in sunshine hours in rural and highelevation mountain areas of the Yunnan Guizhou Plateau, andconsistentwithqianet al. (2006) who reported that cloud cover has decreased over China. CONCLUSIONS AND IMPLICATIONS Contrary to the widely reported reduction in pan evaporation, this study has shown a positive trend in pan evaporation, which is consistent with the increase in potential evaporation reported in south-west China (Thomas, 2000; Xu et al., 2005; Chen et al., 2006; Gao et al., 2006). The increase is much more pronounced in the dry season than in the wet season. Consequently, if there has been no simultaneous change in precipitation, the water available for agriculture and afforestation could be threatened. The present results show that change of wind speed has contributed most of the change in pan evaporation, followed by sunshine hours. In terms of seasonal patterns, the change of the aerodynamic component is dominant

G. YOU ET AL. over the change of E p in the dry season, and the change of the radiative component is dominant over the change of E p in the wet season. Consequently, water evaporation demand could be predicted more efficiently based on the current trends of climatic variables. These results were derived in a nature reserve, and the fact that the trends in pan evaporation, wind speed, and sunshine hours of the present study are inconsistent with those reported previously indicates that further study is required to consider the nature of anthropogenic effects on the widely reported negative trends in pan evaporation. ACKNOWLEDGEMENTS This study was supported by the National Key Project for Basic Research (No. 2010CB833501) and the National Natural Science Foundation of China (No. 31061140359). This is a research contribution from the Ailaoshan Station for Subtropical Forest Ecosystem Studies, Chinese Academy of Sciences, and the Ailaoshan National Ecosystem Observation Research Network Station, Chinese Ecological Research Network. We thank the two anonymous reviewers for their valuable comments. Dr Zhiheng Li, Dr Junbin Zhao, and Mr Jiafu Wu provided assistance with the data analysis and figure drawing. The kind replies from Dr Michael L. Roderick (Research School of Biological Sciences, Australian National University) were very helpful. We are grateful for the free software R (version 2.12.0), which allowed convenient and time-saving calculations. We thank the long-term manual observers, Mr Chengchang Luo, Mr Jinhua Qi, and Mr Xin Luo, for their careful and continuous work. 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: Rome. Alpert P, Kishcha P. 2008. Quantification of the effect of urbanization on solar dimming. Geophysical Research Letters 35: L08801. DOI: 10.1029/2007GL033012 Alpert P, Kishcha P, Kaufman YJ, Schwarzbard R. 2005. Global dimming or local dimming?: Effect of urbanization on sunlight availability. Geophysical Research Letters 32, L17802. DOI: 10.1029/ 2005GL023320 Bronaugh D. 2009. zyp: Zhang + Yue-Pilon trends package. R package version 2.1. http://cran.r-project.org/package=zyp Burn DH, Hesch NM. 2007. Trends in evaporation for the Canadian Prairies. Journal of Hydrology 336:61 73. DOI: 10.1016/j. jhydrol.2006.12.011 Chattopadhyay N, Hulme M. 1997. Evaporation and potential evapotranspiration in India under conditions of recent and future climate change. Agricultural and Forest Meteorology 87:55 73. DOI: 10.1016/ S0168-1923(97)00006-3 Chen S, Liu Y, Thomas A. 2006. Climatic change on the Tibetan Plateau: potential evapotranspiration trend from 1961 2006. Climatic Change 76:291 319. DOI: 10.1007/s10584-006-9080-z Cho J, Komatsu H, Pokhrel Y, Yeh PJF, Oki T, Kanae S. 2011. The effects of annual precipitation and mean air temperature on annual runoff in global forest regions. Climatic Change 108:401 410. DOI: 10.1007/s10584-011-0197-3 Cohen S, Stanhill G. 2002. Evaporative climate changes at Bet-Dagan Israel, 1964 1998. Agricultural and Forest Meteorology 111:83 91. DOI: 10.1016/S0168-1923(02)00016-3 Cong ZT, Yang DW. 2009. Does evaporation paradox exist in China? Hydrology and Earth System Sciences 13:357 366. DOI: 10.5194/ hessd-5-2111-2008 da Silva VPR. 2004. On climate variability in Northeast of Brazil. Journal of Arid Environments 58:575 596. DOI: 10.1016/j.jaridenv.2003.12.002 Fu GB, Liu C, Chen S, Hong J. 2004. Investigating the conversion coefficients for free water surface evaporation of different evaporation pans. Hydrological Processes 18:2247 2262..5526 Fu GB. 2009 A critical overview of pan evaporation trends over the last 50 years. Climatic Change 97:193 214. DOI: 10.1007/s10584-009-9579-1 Gao G, Chen DL, Ren GY, Chen Y, Liao YM. 2006. Spatial and temporal variations and controlling factors of potential evapotranspiration in China: 1956 2000. Journal of Geographical Sciences 16:3 12. DOI: 10.1007/s11442-006-0101-7 Gong LB, Xu CY, Chen DL, Halldin S, Chen YQD. 2006. Sensitivity of the Penman-Monteith reference evapotranspiration to key climatic variables in the Changjiang (Yangtze River) basin. Journal of Hydrology 329:620 629. DOI:10.1016/j.jhydrol.2006.03.027 Guo H, Xu M, Hu Q. 2010. Changes in near-surface wind speed in China: 1969 2005. International Journal of Climatology 31(3):349 358. DOI: 10.1002/joc.2091 Hobbins M, Ramirez JA, Brown TC. 2004. Trend in pan evaporation and actual evapotranspiration across the conterminous US: paradoxical or complementary? Geophysical Research Letters 31: L13503. DOI: 10.1029/2004GL019846 Jhajharia D, Shrivastava SK, Sarkar D, Sarkar S. 2009. Temporal characteristics of pan evaporation trends under the humid conditions of northeast India. Agricultural and Forest Meteorology 149:763 770. DOI: 10.1016/j.agrformet.2008.10.024 Ji YH, Zhou GS. 2011. Important factors governing the incompatible trends of annual pan evaporation: evidence from a small scale region. Climatic Change 106: 303 314. DOI: 10.1007/s10584-010-9900-z Jiang Y, Luo Y, Zhao ZC, Tao SW. 2010. Changes in wind speed over China during 1956 2004. Theoretical and Applied Climatology 99: 421 430. DOI: 10.1007/s00704-009-0152-7 Kirono DGC, Ken DM. 2011. Assessment of rainfall and potential evaporation from global climate models and its implications for Australian regional drought projection. International Journal of Climatology 31: 1295 1308 DOI: 10.1002/joc.2165 Linacre ET. 1992. Climate Data and Resources. Routledge: Boca Raton, FL; 384. Linacre ET. 1994. Estimating U.S. Class A pan evaporation from few climate data. Water International 19(1):5 14. DOI: 10.1080/02508069408686189 Liu XM, Luo YZ, Zhang D, Zhang MH, Liu CM. 2011. Recent changes in pan-evaporation dynamics in China. Geophysical Research Letters 38: L13404. DOI:10.1029/2011GL047929 Liu BH, Xu M, Henderson M, Gong WG. 2004. A spatial analysis of pan evaporation trends in China, 1955 2000. Journal of Geographical Sciences 109:D15102. DOI:15110.11029/12004JD004511 Mann HB. 1945. Nonparametric tests against trend. Econometrica 13(3):245 259 McLeod AI. 2009. Kendall: Kendall rank correlation and Mann-Kendall trend test. R package version 2.1. http://cran.r-project.org/package=kendall McVicar TR, Van Niel TG, Li LT, Hutchinson MF, Mu XM, Liu ZH. 2007. Spatially distributing monthly reference evapotranspiration and pan evaporation considering topographic influences. Journal of Hydrology 338: 196 220. DOI:10.1016/j.jhydrol.2007.02.018 McVicar TR, Roderick ML, Donohue RJ, Li LT, Van Niel TG, Thomas A, Grieser J, Jhajharia D, Himri Y, Mahowald NM, Mescherskaya AV, Kruger AC, Rehman S, Dinpashoh Y. 2012. Global review and synthesis of trends in observed terrestrial near-surface wind speeds: Implications for evaporation. Journal of Hydrology 416 417:182 205. DOI: 10.1016/j.jhydrol.2011.10.024 Oguntunde PG, Friesen J, van de Giesen N, Savenije HHG. 2006. Hydroclimatology of the Volta River Basin in West Africa: trends and variability from 1901 to 2002. Physics and Chemistry of the Earth 31:1180 1188. DOI: 10.1016/j.pce.2006.02.062 Penman HL. 1948. Natural evaporation from open water, bare soil and grass. Proceedings of the Royal Society of London. Series A: Mathematical and Physical Sciences 193(1032): 120 145. DOI: 10.1098/rspa.1948.0037 Peterson TC, Golubev VS, Groisman PY. 1995. Evaporation losing its strength. Nature 377:687 688. DOI:10.1038/377687b0 Qian Y, Kaiser DP, Leung LR, Xu M. 2006. More frequent cloud-free sky and less surface solar radiation in China from 1955 to 2000. Geophysical Research Letters 33: L01812. DOI:10.1029/2005GL024586

POSITIVE TREND OF PAN EVAPORATION R Development Core Team. 2010. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing: Vienna, Austria. ISBN 3-900051-07-0 Rayner DP. 2007. Wind run changes: the dominant factor affecting pan evaporation trends in Australia. Journal of Climate 20:3379 3394. DOI:10.1175/JCLI4181.1 Ren GY, Zhou YQ, Chu ZY, Zhou, JX, Zhang AY, Guo J, Liu XF. 2008. Urbanization effects on observed surface air temperature trends in North China. Journal of Climate 21(6): 1333 1348. DOI:10.1175/ 2007JCLI1348.1 Roderick ML, Farquhar GD. 2002. The cause of decreased pan evaporation over the past 50 years. Science 298: 1410 1411. DOI:10.1126/ science.1075390-a Roderick ML, Farquhar GD. 2005. Change in New Zealand pan evaporation since the 1970s. International Journal of Climatology 25:2031 2039. DOI:10.1002/joc.1262 Roderick ML, Rotstayn LD, Farquhar GD, Hobbins MT. 2007. On the attribution of changing pan evaporation. Geophysical Research Letters 34: L17403. DOI:17410.11029/12007GL031166 Rotstayn LD, Roderick ML, Farquhar GD. 2006. A simple pan-evaporation model for analysis of climate simulations: Evaluation over Australia. Geophysical Research Letters 33: L17715. DOI:10.1029/2006GL027114 Stocker TF, Clarke GKC, Le Treut H, Lindzen RS, Meleshko VP, Mugara RK, Palmer TN, Pierrehumbert RT, Sellers PJ, Trenberth KE, Willebrand J. 2001. Physical climate processes and feedbacks. In Climate Change 2001: The Scientific Basis, Houghton JT, Ding Y, Griggs DJ, Nouger M, van der Linden PJ, Dai X, Maskell K, Johnson CA (eds). Cambridge University Press: Cambridge, UK; 417 470. Tabari H, Marofi S. 2011. Changes of pan evaporation in the West of Iran. Water Resources Management 25:97 111. DOI: 10.1007/s11269-010- 9689-6 Tebakari T, Yoshitani J, Suvanpimol C. 2005. Time-space trend analysis in pan evaporation over Kingdom of Thailand. Journal of Hydrologic Engineering 10 (3):205 215. DOI: 10.1061/(Asce)1084-0699(2005) 10:3(205) Thom AS, Thony JL, Vauclin M. (1981) On the proper employment of evaporation pans and atmometers in estimating potential transpiration. Quarterly Journal of the Royal Meteorological Society 107 (453): 711 736. DOI: 10.1002/qj.49710745316 Thomas A. 2000. Spatial and temporal characteristics of potential evapotranspiration trends over China. International Journal of Climatology 20(4):381 396. DOI:10.1002/(SICI)1097-0088 (20000330)20:4<381::AID-JOC477>3.0.CO;2-K Wang B, Ding QH. 2006. Changes in global monsoon precipitation over the past 56 years. Geophysical Research Letters 33: L06711. DOI:10.1029/2005GL025347 Wang Y, Jiang T, Bothe O, Fraedrich K. 2007. Changes of pan evaporation and reference evapotranspiration in the Yangtze River basin. Theoretical and Applied Climatology 90:13 23. DOI:10.1007/s00704-006-0276-y Xu CY, 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. Journal of Hydrology 327:81 93. DOI: 10.1016/j.jhydrol.2005.11.029 Xu J, Haginoya S, Saito K, Motoya K. 2005. Surface heat balance and pan evaporation trends in Eastern Asia in the period 1971 2000. Hydrological Processes 19:2161 2186..5668 Yao C, Yang S, Qian WH, Lin ZM, Wen M. 2008. Regional summer precipitation events in Asia and their changes in the past decades. Journal of Geophysical Research 113: D17107. DOI:10.1029/2007JD009603 Yin YH, Wu SH, Chen G, Dai EF. 2010 Attribution analyses of potential evapotranspiration changes in China since the 1960s. Theoretical and Applied Climatology 101:19 28. DOI:10.1007/s00704-009-0197-7 Zhang YQ, Liu CM, Tang YH, Yang YH. 2007. Trends in pan evaporation and reference and actual evapotranspiration across the Tibetan Plateau. Journal of Geographical Sciences 112:D12110. DOI:12110.11029/ 12006JD008161 Zheng XB, Luo YX, Duan CC, Zhao TL, Chen J, Kang WM. 2010. Change trends and causes in sunshine duration and visibility over Yunnan-Guizhou plateau in recent 45 years [in Chinese with English abstract]. Plateau Meteorology 29(4):992 998.