Pan evaporation modelling and changing attribution analysis on the Tibetan Plateau (1970-2012)

Pan evaporation (E_p) is an important indicator of water and energy and the decline of E_p has been reported in many regions over the last decades. The climate and E_p are dependent on each other. In this study, the temporal trends of E_p and main E_p drivers, namely mean air temperature (T_a), wind speed (u), global solar radiation (R_s), net long-wave radiation (R_nl) and vapour pressure deficit (D) from 1970 to 2012, were calculated on the basis of 26 meteorological stations on the Tibetan Plateau. The arithmetic average of Ep from 26 stations decreased with the rate of -11.91mm a^-2; the trends of R_s, R_nl, T_a, u and D were -1.434wm^-2 decade^-1, 0.2511wm^-2 decade^-1, 0.3590°Cdecade^-1, -0.2376ms^-1 decade^-1 and 9.523 Pa decade^-1, respectively. The diffuse irradiance is an essential parameter to model E_p and quantify the contribution of climatic factors to changing E_p. 60724 observations of R_s and diffuse solar irradiance (R_d) from seven of the 26 stations were used to develop the correlation between the diffuse fraction (R_d/R_s), and the clearness index (R_s/R_o). On the basis of the estimation of the diffuse component of R_s and climatic data, we modified the PenPan model to estimate Chinese micro-pan evaporation (E_p) and assess the attribution of E_p dynamics using partial derivatives. The results showed that there was a good agreement between the observed and calculated daily E_p values. The observed decrease in E_p was mostly due to declining wind speed (-13.7mm a^-2) with some contributions from decreasing solar irradiance (-3.1mm a^-2); and the increase of temperature had a large positive effect (4.55mm a^-2) in total whilst the increase of Rnl had insignificant effect (0.35mm a^-2) on E_p rates. The change of E_p is the net result of all the climatic variables.