There are very limited direct measurements of actual evapotranspiration over global land areas. Over oceans, estimates of evaporation depend on bulk flux estimates that contain large errors. Evaporation fields from the ERA-40 and NRA are not considered reliable because they are not well constrained by precipitation and radiation (Betts et al., 2003; Ruiz-Barradas and Nigam, 2005). The physical processes related to changes in evapotranspiration are discussed in Section 7.2 and Section 3.4, Box 3.2.
Decreasing trends during recent decades are found in sparse records of pan evaporation (measured evaporation from an open water surface in a pan) over the USA (Peterson et al., 1995; Golubev et al., 2001; Hobbins et al., 2004), India (Chattopadhyay and Hulme, 1997), Australia (Roderick and Farquhar, 2004), New Zealand (Roderick and Farquhar, 2005), China (Liu et al., 2004a; Qian et al., 2006b) and Thailand (Tebakari et al., 2005). Pan measurements do not represent actual evaporation (Brutsaert and Parlange, 1998), and any trend is more likely caused by decreasing surface solar radiation over the USA and parts of Europe and Russia (Abakumova et al., 1996; Liepert, 2002) and decreased sunshine duration over China (Kaiser and Qian, 2002) that may be related to increases in air pollution and atmospheric aerosols (Liepert et al., 2004; Qian et al., 2006a) and increases in cloud cover (Dai et al., 1999). Whether actual evapotranspiration decreases or not also depends on how surface wetness changes (see Section 3.4, Box 3.2). Changes in evapotranspiration are often calculated using empirical models as a function of precipitation, wind and surface net radiation (Milly and Dunne, 2001), or land surface models (LSMs; e.g., van den Dool et al., 2003; Qian et al., 2006a).
The TAR reported that actual evapotranspiration increased during the second half of the 20th century over most dry regions of the USA and Russia (Golubev et al., 2001), resulting from greater availability of surface moisture due to increased precipitation and larger atmospheric moisture demand due to higher temperature. One outcome is a larger surface latent heat flux (increased evapotranspiration) but decreased sensible heat flux (Trenberth and Shea, 2005). Using observed precipitation, temperature, cloudiness-based surface solar radiation and a comprehensive land surface model, Qian et al. (2006a) found that global land evapotranspiration closely follows variations in land precipitation. Global precipitation values (Figure 3.12) peaked in the early 1970s and then decreased somewhat, but reflect mainly tropical values, and precipitation has increased more generally over land at higher latitudes (Figures 3.13 and 3.14). Changes in evapotranspiration depend not only on moisture supply but also on energy availability and surface wind (see Section 3.4, Box 3.2).