Evaluating the complementary relationship of evapotranspiration in the alpine steppe of the Tibetan Plateau

Abstract The complementary relationship (CR) of evapotranspiration allows the estimation of the actual evapotranspiration rate ( ET a ) of the land surface using only routine meteorological data, which is of great importance in the Tibetan Plateau (TP) due to its sparse observation network. With the highest in situ automatic climate observation system in a typical semiarid alpine steppe region of the TP, the wind function of Penman was replaced by one based on the Monin‐Obukhov Similarity theory for calculating the potential evapotranspiration rate ( ET p ); the Priestley‐Taylor coefficient, α , was estimated using observations in wet days; and the slope of the saturation vapor pressure curve was evaluated at an estimate of the wet surface temperature, provided the latter was smaller than the actual air temperature. A symmetric CR was obtained between the observed daily actual and potential evapotranspiration. Local calibration of the parameter value (in this order) is key to obtaining a symmetric CR: α , wet environment air temperature ( T wea ), and wind function. Also, present symmetric CR contradicts previous research that used default parameter values for claiming an asymmetric CR in arid and semiarid regions of the TP. The effectiveness of estimating the daily ET a via symmetric CR was greatly improved when local calibrations were implemented. At the same time, an asymmetric CR was found between the observed daily ET a and pan evaporation rates ( E pan ), both for D20 aboveground and E601B sunken pans. The daily ET a could also be estimated by coupling the E pan of D20 aboveground and/or E601B sunken pan through CR. The former provided good descriptors for observed ET a , while the latter still tended to overestimate it to some extent.