Assessing climate change induced modification of Penman potential evaporation and runoff sensitivity in a large water-limited basin Qiang Liu a,b,⇑ , Tim R. McVicar c a Key Laboratory for Water and Sediment Sciences, Ministry of Education, School of Environment, Beijing Normal University, Beijing 100875, China b State Key Laboratory of Water Environment Simulation, School of Environment, Beijing Normal University, Beijing 100875, China c CSIRO Land and Water, GPO Box 1666, Canberra, 2601 ACT, Australia article info Article history: Received 23 March 2012 Received in revised form 17 July 2012 Accepted 19 July 2012 Available online 3 August 2012 This manuscript was handled by Geoff Syme, Editor-in-Chief Keywords: Attribution analysis Potential evaporation Net radiation Ecohydrological processes Yellow River Basin summary Potential evaporation (E p ) reflects the combined effects of four key meteorological variables: (i) net radi- ation (R n ); (ii) wind speed (u); (iii) relative humidity (rh); and (iv) air temperature (T a ). Here, attribution analysis was conducted to investigate the contribution of the four key meteorological variables to changes of a physically-based E p in a large water-limited basin, the Yellow River Basin (YRB), China. Then the influences of these changes, and precipitation (P) changes, on streamflow (Q) were explored analyt- ically. Results show that: (i) E p presented different temporal trends for the water yielding region (WYR) and water consuming region (WCR) with a overall changes of +0.16 mm a 2 and 0.66 mm a 2 during 1961–2010, respectively; (ii) trend analysis of E p and the four key meteorological variables at the basin scale showed that increasing trend in T a increased E p during 1961–2010, while changes in R n and u increased the 1961–1979 E p rate and reduced it during 1980–1994 and 1995–2010; (iii) revealed by attri- bution analysis, E p increased by changes in T a and rh and reduced by changes of R n and u in both WYR and WCR, in all, E p rate presented positive and negative trends in the WYR and WCR, respectively; (iv) the changes of Q and actual evaporation (E) are much more sensitive to changes in P than the changes in E p ; and (v) of critical importance for water resource management of the YRB changes in Q are mainly attributed to changes in catchment-specific parameter (n) and P, while E p reduced Q in WYR and increased Q in WCR. These results indicated that the causes of trend of E p rates, influenced by combined effects of radiative and aerodynamic variables should be explicitly explained using fully physically based E p formulations. Additionally, in the water-limited YRB, changes of Q are primarily controlled by the changes in catchment conditions, and secondarily by hydroclimatic factors where the available water (P) rather than energy condition (E p ) is more important. Better understanding all of these relationships and how they have varied will help water resource management in a changing climate. Ó 2012 Published by Elsevier B.V. 1. Introduction Decreasing trends in pan evaporation (McVicar et al., 2012; their Table 5 and the references therein) and potential evaporation (E p ) have been reported to be occurring simultaneously in many regions with increasing trends of air temperature, which has been denoted the ‘‘evaporation paradox’’ (Roderick and Farquhar, 2002). Possible reasons for this include: (i) complementary relationship between the actual evaporation (E) and E p (Hobbins et al., 2004; Ramírez et al., 2005; Yang et al., 2006); (ii) reducing trends in irradiance due to increased cloudiness and aerosol concentrations (Roderick and Farquhar, 2002); and (iii) decreasing trends in wind speed (Ray- ner, 2007; Roderick et al., 2007; McVicar et al., 2012). E p varies with the changes of aerodynamic and radiative variables that are influ- enced by both climatic changes and terrestrial conditions (e.g., Peel et al., 2010; Thanapakpawin et al., 2006; McVicar et al., 2007a; Rod- erick et al., 2009a, 2009b; Liang et al., 2010). E p and precipitation (P) are regarded as available energy and water respectively, and clima- tologically their different amounts controls the partitioning of P into E and streamflow (Q) (e.g., Budyko, 1974; Donohue et al., 2007, 2010b; Liu and Yang, 2010; Roderick and Farquhar, 2011). E p can be defined as ‘‘the quantity of water evaporated per unit area, per unit time from an idealized, extensive free water surface under existing atmospheric conditions. This is a conceptual entity which measures the meteorological control on evaporation from an open water surface’’ (Shuttleworth, 1993, p. 4.2). This reflects the combined effects of the four key meteorological variables pri- marily governing they evaporative process, they are: (i) net radiation (R n ); (ii) wind speed (u); (iii) relative humidity (rh); and (iv) air temperature (T a ). In order to calculate E p many different methods, using one, or more, of these four variables have been 0022-1694/$ - see front matter Ó 2012 Published by Elsevier B.V. http://dx.doi.org/10.1016/j.jhydrol.2012.07.032 ⇑ Corresponding author at: Key Laboratory for Water and Sediment Sciences, Ministry of Education, School of Environment, Beijing Normal University, Beijing 100875, China. Tel.: +86 10 58802771. E-mail address: liuqiangbnu@163.com (Q. Liu). Journal of Hydrology 464–465 (2012) 352–362 Contents lists available at SciVerse ScienceDirect Journal of Hydrology journal homepage: www.elsevier.com/locate/jhydrol