comparison of six potential evapotranspiration methods for regional use in the Southeastern United States

Six commonly used potential evapotranspiration (PET) methods were compared that have the potential to be incorporated into regional-scale hydrologic modeling, and PET was quantified across the climatic gradient of the southeastern US. The methods, which are tabulated, included three temperature-based methods and three radiation-based methods. Long-term annual watershed-scale actual ET (AET) was estimated by the water balance equation to compare the PET values. Results showed that the Thornthwaite and Hamon PET methods had the highest Pearson correlation coefficient values, while the Hargreaves-Samani PET method had the lowest values. The Thornthwaite method yielded the lowest long-term averaged annual PET, while the Hargreaves-Samani method predicted the highest values. Among the 36 southeastern US sites, greater differences were found among the temperature-based PET methods than among the radiation-based methods, and the relative differences in predicted PET among methods varied greatly both between watersheds and between years. Overall, the Priestley-Taylor, Turc, and Hamon PET methods were determined to be more accurate than the Thornthwaite, Makkink, and Hargreaves-Samani PET methods for watershed-scale applications in the southeastern US.