For high frequency irrigation management, based on climate data, it is desirable to estimate reference evapotranspiration at small time steps. This work aimed at to evaluate the performance of daily reference evapotranspiration estimated by the following methods: FAO-24 Radiation, Priestley-Taylor, Hargreaves-Samani, Camargo-71 and Blaney-Criddle, for Garanhuns-PE micro-region, having FAO Penman-Monteith method as standard. Considering the mean absolute error (MAE), among other statistics, it was verified that the best performance was obtained by the Priestley-Taylor method. Regarding the methods applicable when only temperature data are available, the Hargreaves-Samani showed the best performance, having been calibrated based on the minimization of the MAE.
model assessment; mean absolute error; reliability index
