Evaluation of 18 models for calculating potential evapotranspiration in different climatic zones of China

The accurate calculation of potential evapotranspiration (PET) is a critical step in researching evapotranspiration, hydrology and many other fields. Numerous models have been developed to quantify PET from standard meteorological observations, and combination methods are usually considered the most physically reasonable. However, adequate observations of meteorological variables for combination methods are unavailable in many locations, making it necessary to select alternative PET models with fewer data requirements. Here, a set of 18 models including four aerodynamic methods, five temperature-based methods, six radiation-based methods, and three combination methods were evaluated using meteorological measurements from 789 stations in four climatic zones in China: the mountain plateau zone (MPZ), temperate monsoon zone (TMZ), temperate continental zone (TCZ), and subtropical monsoon zone (SMZ). The annual PET calculated in each of the four climatic zones showed large discrepancies among the 18 models, and the largest disparity nationwide was 2.95-fold. The combination models performed best for calculating PET in all four climatic zones, followed by the radiation-based models, and both categories outperformed the aerodynamic and temperature-based methods. The Rohwer model was the only recommended aerodynamic method, and the Romanenko model was the only recommended temperature-based method for calculating PET in China. The Turc model was marginally the best radiation-based model in the SMZ, TMZ and TCZ, and the Hargreaves model in the MPZ, but both should be applied with caution in cold months. The Penman model was the recommended combination method in all four zones. Further comparison of the best models from each category showed that the Rohwer model might overestimate PET in the TMZ and TCZ, and underestimate it in the MPZ and SMZ. The Romanenko model overestimated PET, and the Turc and Hargreaves models both underestimated PET in all four zones, especially in the MPZ. The empirical coefficients of the five recommended models were regional calibrated to meet the requirements of PET calculation in different climatic zones. •Combination models performed best for calculating potential evapotranspiration, and the Penman model was optimal.•The Rohwer model was the only recommended aerodynamic method.•The Romanenko model was the only recommended temperature-based method.•The Hargreaves model was the best radiation-based method in the mountain plateau zone, an